Battery‐Free and Wireless Technologies for Cardiovascular Implantable Medical Devices

Cardiovascular disease continues to be one of the dominant causes of global mortality. One effective treatment is to utilize cardiovascular implantable devices (cIMDs) with multi‐functional cell sensing and monitoring features that have the potential to manipulate cardiovascular hyperplasia disorders as well as provide therapy. However, batteries with a fixed capacity entail high‐risk surgeries for battery‐replacement, which causes health hazards and imposes significant costs to patients. This review accesses comprehensive power solutions for cIMDs, from conventional batteries to state‐of‐the‐art energy harvesters and wireless power transfer (WPT) schemes. In particular, WPT has great potential to eliminate the percutaneous wires and overcome frequent battery removal. Here, the fundamentals, power transfer efficiency, antenna design and miniaturization, and operating frequencies in various WPT schemes are presented. Moreover, the power loss attenuation and bio‐safety standard (specific absorption rate) for implants are also considered in WPT design envelope. In addition, wireless data transmission of implantable devices from external to internal milieu (and vice versa) along with different modulation and demodulation techniques are investigated. The last advanced power solutions for cIMDs in in‐vivo and in vitro research are illustrated throughout. Finally, specifications and future potential of WPT systems in cIMDs are highlighted.


Introduction
Advanced microelectronics have increasingly improved the intelligence and miniaturization of cardiovascular devices, which is critical for monitoring, diagnosing, and treating to replace the battery and charge cardiovascular implants, but they are currently hindered by unstable power generation and percutaneous wires.
The electromagnetic (EM) wireless transfer system has drawn extensive attention for noninvasive powering and monitoring of implantable devices by the electromagnetic energy transfer mechanism. This scheme can simultaneously transmit power and bidirectional data from an adjustable external source to an internal implantable device. Over the last few decades, various wireless power transfer (WPT) links along with their characteristics, such as power transfer efficiency (PTE), transmission distance, and operating frequency have been significantly investigated, from the most established inductive coupling (IC) power transfer and far-field radio frequency (RF) to the emerging magnetic resonant coupling and mid-field WPT. Although WPT is preferable power candidates for the implantable devices, there are still several constraints that require further research and optimization. Particularly, the PTE of inductive coupling WPT is vulnerably affected by the misalignment and transmission distance between the coils, resulting in weak coupling. In addition, the power transfer is suppressed by the body-safety constraints (called specific absorption rate (SAR)) to prevent adverse thermal effects in the tissues. Furthermore, in order to match the ultraminiaturized implants, various antenna types and geometry modifications are investigated to achieve small and efficient antennas. In addition to power strategies for cardiovascular implants, various modulation and demodulation techniques adopted in data transmission system are also explored to detect and control biological parameters in the body. This paper clarifies power strategies and data transmission schemes of cardiovascular implants, with special emphasis on the wireless power and data transmission system and its future potential. Section 2 explores the specifications of several cardiovascular implants and their relevant medical applications, which can be observed in Figure 1. Moreover, the battery evolution and power requirement in cardiovascular implants are also illustrated in this section. In Section 3, we consider various energy transducers from multiple aspects such as working principles, essential properties, and advanced materials. Finally, the basic principles of wireless power and data transfer system are presented in the Sections 4 and 5, including essential specifications, optimization suggestions, and advanced research in cardiovascular implants. Particularly, the mid-field WPT with patterned metal plates has a great potential to power miniaturized cIMDs in deep tissue among numerous WPT schemes.

cIMDs
It is estimated that 17.9 million people worldwide die from cardiovascular disease every year. [6] The cardiovascular system consists of a heart and associated blood vessels that convey oxygenated blood and nutrients to perfuse the body with all substrates required for metabolism. The cardiac conduction system of a heart stimulates and coordinates the contractions and relaxations of all four chambers of the heart to transport blood away and around the body and then to the lungs for gas exchange. The coronary arteries and their branches supply the blood to the entire heart muscle but are susceptible to disease such as fibrofatty blockage caused by atherosclerosis. The failures of the cardiac conduction system or vascular can bring occlusion inducement of heart attacks and/or coronary artery disease which can undoubtedly prove fatal. Early intervention in cardiac pathologies offers patients an early-detection and fast-response therapy. cIMDs such as cardiac pacemakers, defibrillators, ventricular assist devices, and novel smart stents have been pervasively regarded as promising treatments which are likely to bring positive and effective solutions to cardiovascular patients. Figure 2 depicts numerous cIMDs and power schemes along with their development. For instance, the pacemaker was first implanted in 1958, [7] is the most established cardiac implant and whose technologies continue to improve. Approximately 1.25 million pacemakers are implanted every year, with this number increasing every year. [8] The pacemaker is specially designed to help detect abnormal heartbeats and adjust them back to normal levels. They can generate the electrical stimulus required to coordinate and restore the heart rhythm by mimicking the sinoatrial node (SAN). [9] Assisted by several components such as the pulse generator and an array of sensors like accelerometers, piezoelectric crystals and electrocardiogram (ECG) sensors, the pacemaker can detect cardiac-relevant parameters (e.g., minute ventilation, peak endocardial acceleration, and respiratory changes), further regulating the heart rate when its workload exceeds the set threshold. [10] Similarly, the implantable cardioverter defibrillators (ICDs) are distinguished by delivering high-level shocks against sudden cardiac death caused by ventricular arrhythmias. VADs are commonly used for end-stage heart failure patients who are waiting for heart transplantation. VAD is an implantable blood pump that transfers blood from the failure ventricle to the aorta and flows through all the body. [11] However, conventional VAD runs with a fixed blood flow speed, which cannot adjust the blood flow according to the cardiac demand. Overpumping or underpumping of blood supply can cause pressure imbalance at the interfaces between the device and the heart, and even be fatal. Recently, an adaptative tuning VAD was proposed, which integrated a pressure sensor and a flow sensor to detect and alter blood flow and pressure in real-time. [12] Interventional coronary stent placement, called percutaneous (through the skin) coronary intervention (PCI), is a surgical procedure for combatting coronary artery diseases (CADs). CAD frequently occurs on narrow or blocked blood vessels, which is caused by the progressive build-up of plaque, termed atherosclerosis, which can further occlude the blood vessels and reduce oxygen supply to the heart. Current coronary stents are delivered through the femoral or radial arteries and serve as passive mechanical support. By placing the stent on an inflatable balloon, the stent can be directed to the exact point of occlusion. Once the balloon inflated with saline, the stent expands in the artery to reopen the full caliber of the vessel and restore blood flow. However, a postoperative complication in-stent restenosis (ISR) frequently occurs and is attributable to the formation of thrombosis, the proliferation of smooth muscle cells, and neointimal hyperplasia. [13] A novel, selfreported stent incorporates a pressure flow sensor and a cell detector to provide a predictive solution to impending vascular complications. More advanced reasons propose protein molecular sensors to detect the smooth muscle cells and clots in the vascular system, with the aim of overcoming in-stent restenosis from early detection [14] and even deliver appropriate drugs [15,16] In addition, such smart stents enable continuous monitoring and appropriate treatments like hyperthermia therapy [17] and drug delivery. [18]

Continuous Energy Requirement and Power Management in cIMDs
Successful operation of cardiovascular implants relies on a continuous power supply. Extensive efforts are now being devoted to new power strategies and the development of inexhaustible sources. Figure 2A-N presents the milestones in the evolution of cardiovascular implants and corresponding feasible power strategies over the decades. [19][20][21][22] Emerging cardiovascular Figure 2. Milestones in the cardiovascular implantable devices and corresponding power strategies: Cardiovascular implants: A) first successful pacemaker implantation in 1958; [7] B) the first successful implantation of a left ventricular assist device lasted for 10 days until the patient completed the heart transplant; [19] C) the first ICD was implanted in a patient in 1980; [20] D) the first coronary stent implanted in a patient coronary artery by Sigwart et al. in 1986; [21] E) the first leadless pacemaker was implanted in 2012. [22] Power strategies: F) Nikola Tesla achieved inductive power transfer by using a pair of coils in 1889; G) the first far-field wireless microwave transfer was developed in 1964; [191] H) magnetic resonant coupling; [159] I) capacitive coupling; [173] J) mid-field WPT. [217] K) The first research work of piezoelectric nanogenerator was developed by Wang and Song in 2006; [70] L) triboelectric nanogenerator was first invented in 2012 and used to collect mechanical energy in-body and convert it into electrical energy; [86] M) the nuclear battery is the early form of TENGs; [26] N) the first biofuel cell implanted in rats in 2010. [59] O) The first implanted pacemaker was powered by a rechargeable nickel-cadmium battery; [23] P) battery mercury-zinc battery get involved in powering pacemaker in 1966; [25] Q) nuclear battery prolonged the lifespan over 15 years and employed in pacemaker from 1970; [26] R) since 1972, lithium/iodine battery have been the power solution for cardias implants until today. [27] implants with novel power solutions are moving toward smart and fully implanted devices. However, state-of-the-art energy harvesting technologies are still at the level of investigation or animal testing. Batteries remain the dominant power source of cardiovascular implants for clinical and commercial purposes.
Batteries have been used for serving cardiac implants since the first implantable pacemaker, which used a nickel-cadmium battery. This rechargeable battery was depleted after 7 weeks [23] of implantation and, meanwhile, encountered toxicity and cumbersome charging problems. Subsequently, primary mercuryzinc batteries were employed to powering pacemakers, which can prolong the lifespan of the battery by up to 3 years. [24,25] In 1970, [26] the nuclear battery replaced the mercury-zinc battery to support pacemaker deployment. It is an early thermoelectric generator whose electricity is derived from the heat released by decaying radioactive isotopes. Although the lifespan of the mercury-zinc battery is extended to 15 years, this toxic and expensive power source may pose safety hazards due to radiation exposure. The breakthrough of battery application in cardiac implants happened in 1972, [27] lithium/iodine cells had been successful used in clinical and achieved over 10 years lifespan. Since then, lithium and lithium-ion batteries have become the dominant power strategy to supply implantable devices. Table 1 summarizes the characteristics of several cIMDs and their power requirements. [20,[28][29][30][31][32][33][34][35][36][37][38][39][40] For instance, when defibrillation occurs, the ICD needs to delivery of high voltage electric shocks (provide at least 5 W pulse power for 10 s), so it demands a higher power capability than a pacemaker. [41] ICDs are frequently powered by lithium/silver vanadium oxide (Li/ SVO) batteries, which encounters the problem of delay of shock delivery due to the increasing resistance during discharging. [42] Recently, other lithium-ion batteries like the lithium manganese dioxide (MDX) battery, the lithium carbon monofluoride have been employed to power the cardiac implants. In particular, silver vanadium oxide (Ag 2 VO 2 PO 4, SVPO) has little effect on the delay of electric shock than the Li/SVO when the resistance increased within the cell. [43] Currently, the lithium-ion battery is a pervasive clinical power approach with fixed energy density, limited lifespan, and large size, which is not a permanent and comfortable solution. Repeated battery replacement surgery to vulnerable to high risks patients can induce serious adverse effects. Moreover, the power capability and energy density of the battery are proportional to its volume that cannot satisfy the novel miniaturized implants. Hence, it is imperative to study alternative power solutions for cIMDs.

Energy Harvesting Strategies in Powering cIMDs
An effective approach to overcome the drawbacks of the battery is to harvest energy from the body so that implantable devices can be self-powered. To drive these devices, abundant chemical, thermal, and mechanical energy from the human body can be harnessed as sustainable renewable energy sources extracted via various energy transducers. In particular, four types of energy transducer: biofuel cells, piezoelectric generators, triboelectric generators, and thermoelectric generators are attractive for powering implantable devices, as presented in Figure 3A-D. In order to determine the power feasibility of in-body energy in the field of cardiovascular implantable devices supply, the aforementioned energy transducers have been investigated in this paper based on their fundamentals, power generation, and in vivo animal experiments.

Biofuel Cell
Biofuel cells have received considerable attention since their significant progress in in vivo animal tests. The concept of biofuel cells was developed in the early 1960s [44] until the breakthrough of biofuel cells made in the 20th century provided an encouraging alternative power solution for implantable devices. The working principle of the biofuel cells is the same as that of electrochemical batteries based on redox reactions. Biofuel cells employ renewable enzymes or microbes as the catalyst rather than using metals as catalyst in electrochemical batteries. Moreover, biofuel cells extract chemical energy from continuous biofuel in blood flow (e.g., glucose, lactose) that can be replenished, while electrochemical batteries are limited by fixed fuel, and need to be replaced when it is depleted.
Biofuel cells can be classified into microbial fuel cells and enzymatic fuel cells (EFCs) according to the biocatalyst type. Microbial fuel cells exploit living microorganisms (e.g., bacteria) as the biocatalyst, where the power generation is significantly influenced by the microbial metabolism [45] and culture condition (e.g., temperature and potential of hydrogen (PH) [46,47] ). Although microbial fuel cells offer a durable operating period (more than 5 years) due to the long lifespan of living microorganisms, [48] this type of cell exhibits low power density owing to the resistance of cell membranes and low catalytic activity. [49] In contrast, enzyme-based fuel cells have been  [29] 60 g c) [30] 130 g d) 145 g e) [31] 270 g f) [32] 8-38 mm g) , 2.5-4 mm h) [33] Lifespan 7.2 years [34] 4.9 ± 1.6 years [35] 29.7 ± 14.9 days [36] -Power level 1-10 μW [37] 10-100 μW [20,38] 5-25 W [39] mW level [40] Power extensively studied for energy conversion due to their higher power density (expressed in mW cm −2 ) [50] than microbial fuel cells. However, enzymes are ease of degraded, resulting in an enzyme fuel cell's lifespan usually of 7-10 days. [48] A typical enzymatic biofuel cell consists of an anode and cathode, which are separated by a proton exchange membrane, as shown in Figure 3A. At the anode, the enzyme catalyzes the oxidation reaction of a biofuel while releasing electrons to reach the electrode surface and then traveling to the cathode via external wires. Meanwhile, protons produced by the anode migrate to the cathode and complete the reduction reaction with an oxidant and electrons. The performance of enzyme fuel cells is mainly determined by electron transfer between the enzyme and electrode. [51] Two major transfer mechanisms are adopted to establish effective electrical communication in enzyme fuel cells: mediated electron transfer (MET) and direct electron transfer (DET). [52] DET carries out the electrons from the active site of the enzyme directly to the electrode surface, while MET requires a mediated molecule (mediator) as an electron transfer relay to shuttle the electrons to the electrode. In general, redox-active modules and polymers act as electron donors and acceptors between the enzyme and electrode surface, making the redox potential close to the active center of the enzyme. [53] In addition, MET has a faster electron transfer speed than DET due to its shorter electron mobile distance (within 2 nm). However, the MET-based biofuel cells may cause safety issues and low open-circuit voltages induced by the potential difference between the active site of the enzyme and the mediator. [54] Therefore, the DET method is widely employed in biofuel cells due to its simple structure.
The significant breakthrough in biofuel cells in animal implementation offers a potential solution for cardiovascular implants. A glucose enzyme-based biofuel cell was implanted in a lobster and generated an open-circuit voltage (V oc ) of 0.54 V. [55] To enhance the V oc , they found that two series-connection biofuels in two separate lobsters doubled the V oc up to 1.2 V. Moreover, they designed a biofuel cell array that consists of 5 individual biofuel cells connected in series in human serum solutions filled with concentration of glucose (for mimicking the human blood circulatory system) to power the pacemaker. As expected, a 2.8 V of V oc was obtained, and the pacemaker worked as normal after being connected with the biofuel cell array. Another study employed carbon nanotubes (CNTs) as electrodes of the biofuel cell and implanted in a snail. In particular, the researchers adopted the DET mechanism in the biofuel cell, which is conducive to the effective immobilization of enzymes on electrodes, resulting in a power of 7.45 μW and a V oc of 0.53 V. [56] Subsequently, the same team [57] developed a comparative study in which implanted three biofuel cells in three independent clams, connecting in series and parallel, respectively. Three series array biofuel cells delivered peak power of 5.2 μW with a V oc of 0.8 V, while the parallel one obtained a 0.36 V of V oc and 37 μW of maximum power, respectively. However, the feasibility of biofuel cells for powering biomedical implants cannot be fully demonstrated by aforementioned experiment in invertebrates. Therefore, a biofuel cell was implanted in a mammal (in a rat cremaster tissue) to harvest the electricity. [58] The CNTs were applied at enzyme-modified electrodes and obtained a stable V oc of 140 ± 30 mV and a 0.35 μW of maximum power. Another biofuel cell [59] was also implanted in a rat but placed in retroperitoneal space, in which both electrodes were based on the compacted graphite disc (containing ubiquinone and glucose oxidase at the anode, while polyphenol oxidase and quinone at the cathode). The biofuel cell was measured in vivo with a power production of 6.5 μW under 0.13 V and then dropped to a stable power of 2 μW after several hours. Later on, a study on biofuel cells was carried out in the blood vessel of a rabbit ear. [60] In particular, to prevent the blood clot from covering the electrode surface, they coated 2-methacryloyloxyethyl phosphorylcholine (MPC) polymer on electrodes to achieve biocompatibility in vivo and generated power of 0.42 μW at 0.56 V, while the control group without the MPC coating presented a 40% reduction of output power. To verify the sustainability of biofuel cells for operating biomedical implants in vivo, El Ichi-Ribault et al. [61] implemented a wireless tele-transmission system in a rabbit to monitor the variation in the biofuel cell for two months. During the first week of implantation, the V oc remained at a range of 0.65-0.68 V and then obtained a stable V oc of 0.42 V after 18 days. Moreover, the inflammatory action and biofouling caused a sharp drop in voltage of implanted biofuel cell in the first week and subsequently became steady.
While several biofuel cells have been explored in vivo experiments in animals, the output level (sub μW, V oc < 1 V) produced by biofuel cells can not satisfy the power consumption of cIMDs except for the ultralow-power pacemakers. Moreover, there remain hurdles in the practical deployment of enzyme biofuel cells. In particular, the efficient electron transfer between the enzyme active site-electrode interface and the enzyme immobilization are the primary issues for improving the overall power generation. To this end, the nanostructured conducting CNTs have been developed to facilitate the electron transfer rate, stability, and concentrations of enzymes. CNTs possess superior biocompatibility and large electroactive surface (1-10 nm [62] ) which benefits the enzyme-electrode interface. Furthermore, the operation period and the inflammatory response of biofuel cells in vivo impede their application in implantable devices.

Piezoelectric Transducer
Piezoelectric transducer is a prominent candidate for converting scavenging vibrational energy into electricity to drive implantable devices. Piezoelectric transducers are characterized by higher energy density (35.4 mJ cm −3 ) compared to electrostatic (4 mJ cm −3 ) and electromagnetic (24.8 mJ cm −3 ) transducers. [63] The piezoelectric transducers benefit from their smart piezoelectric materials with inherent transduction capacity, while the other transducers require additional conditions to generate electricity, such as external source (electrostatic transducers), induction magnetic field (electromagnetic harvesters), and frictional contact (triboelectric generators). [64] Furthermore, piezoelectric materials can generate the opposite electric charges in response to the mechanical strain, known as the piezoelectric effect. In the exploration of vibration energy, piezoelectric transducers have drawn considerable attention in academia. In particular, the cardiovascular system can supply mechanical energy to cardiovascular implants from heart contraction/relaxation, blood circulation, and arterial pulsation. However, piezoelectric generators require sensitive transducer structures and stretchable piezoelectric materials due to slight vibrations of heart motion (vibrational frequency of 1-1.34 Hz [65] ) and the vulnerability of the heart.

Piezoelectric Materials
Piezoelectric materials have been extensively explored to generate power from kinetic since the piezoelectric effect was first discovered by Pierre and Jacques Curie in 1880, [66] which has a significant impact on the performance and applications of piezoelectric transducers.
Ceramic materials such as lead zirconium titanate (PZT) and barium titanate (BaTiO 3 ) are the most broadly used piezoelectric materials in piezoelectric transducers, which possess high dielectric constant (PZT: 200-5000, BaTiO 3 : 1700) and superior piezoelectric coefficient (PZT: d 33 = 100-1000 pC N −1 , BaTiO 3 : d 33 = 190 pC N −1 ). [67,68] However, the brittleness and toxicity properties of PZT are unacceptable for implantable devices. On the contrary, polymeric materials are characterized by superior flexibility and biocompatibility, but suffering from low sensitivity and efficiency. In particular, the typical polymer material polyvinylidene fluoride (PVDF) possesses a 30 pC N −1 piezoelectric coefficient (d 33 ), [69] which is ten times lower than that of ceramics. In order to realize high flexibility and piezoelectricity on piezoelectric materials, various types of composites and nanomaterials have been investigated. In 2006, Wang and Song [70] initially proposed a piezoelectric nanogenerator (PENG) based on zinc oxide (ZnO) nanowires. When the ZnO nanowire is deformed by an atomic force microscope (AFM) tip, it will generate a strain field and the charge separation, as illustrated in Figure 3B. In addition, novel composite piezoelectric materials (mix polymeric piezoelectric materials with inorganic materials) can overcome the shortcomings of ceramic materials and polymeric materials have also attracted great attention. For instance, Lu et al. prepared a PZT in form of layer stacks and printed it onto a flexible polyimide substrate (Kapton), packaging with polydimethylsiloxane (PMDS) encapsulation. [71] This ultraflexible piezoelectric transducer extracted the cardiac motion of a swine and generated a 3 V peak-to-peak voltage. Alternatively, another type of composite uses flexible polymers filled with biocompatible ZnO nanoparticles. For instance, by adding ZnO nanoparticles and multiwall carbon nanotubes (MWCNTs) in PDMS-infilled PVDF-TrFE composite film, Xu et al. [72] developed a nanomaterial-mediated polyvinylidene fluoride-trifluoroethylene (PVDF-TrFE) film. This promising film combines the merits of ZnO nanoparticles and PDMS, enhancing the piezoelectricity and sensitivity, respectively. Moreover, they designed a multibeam PEG to support a pacemaker based on this film and generated an output of 3.22 ± 0.24 V. Recently, Azimi et al. [73] proposed a composite PVDF nanofiber containing hybrid nanofillers of ZnO and reduced graphene oxide (rGO). In order to evaluate the feasibility of the polymer composite-based PENG in powering pacemakers, an in vivo implantation has been carried out in a dog and yielded 0.487 μJ per heartbeat. In addition, the recent emergence of single crystals is regarded as an alternative piezoelectric material of PZT ceramics due to their exceptional piezoelectric performance. For instance, the piezoelectric coefficient (d 33 > 2000 pC N −1 ) and electrotechnical coupling coefficient (k 33 > 0.9) of lead magnesium niobate-lead titanate (PMN-PT) exhibits are several times higher than that of PZT. [74] In a paper by Hwang et al., [75] a self-powered cardiac pacemaker in a living rat was powered by a PMN-PN energy transducer. This PENG used the PMN-PT single crystal on a polyethylene terephthalate (PET) substrate, and successfully collected an opencircuit voltage of 8.2 V and a short-circuit current of 0.223 mA.

Piezoelectric Transducer Structure
Cantilever beam piezoelectric transducers are the most established and extensively used transducers for vibration harvesting.
Since its higher strain generation for a given force input compares with other types of piezoelectric transducers (e.g., circular diaphragm, cymbal type, and stack type). [76] This simple type of transducer can generate significant power under excitation of vibration frequencies in the body. However, extra-low vibration frequencies from cardiovascular systems hinders the amount power generation of piezoelectric transducers. To overcome the resonance matching and miniaturization problems of piezoelectric transducers, plenty of studies on the configuration of cantilever beams have been carried out. Compared with conventional cantilever beams, Amin Karami and Inman [77] initially proposed and proved that flat zigzag-shaped transducers can significantly increase the power density and reduce the natural frequency of piezoelectric transducers. This solution was achieved by reducing the stiffness of beams and eliminating torsional effects in the structural design of multiple cantilever beams, generating 10 μW power under the excitation of heartbeats. Since then, the advent of other cantilever shapes (including fan-fold shape [78] and elephant shape [79] ) have improved the performance of piezoelectric transducers at low vibrational frequencies.
Recently, advanced PENGs with nanostructures exhibit superior flexibility and stretchability under an off-resonance mechanism. A PENG consist of a pair of top and bottom electrodes in a sandwich structure. [80] When the nanowires are bent by an AFM tip, PENG generates electric charge polarization and induces an electric field, thereby forming a Schottky barrier between nanowires and metal contacts. [70] Therefore, PENGs can provide great potential in improving energy conversion efficiency and arouse significant interest in self-powered micro/ nanoscale implantable devices Table 2 summarizes several bulk cantilever-type piezoelectric transducers and PENGs in applications of using cardiovascular vibrations to power cIMDs, as well as their design parameters and output performance. In addition, piezoelectric transducers need to be attached to the heart to harvest heartbeats vibration, but the movement of heart is complicated and random. The placement of piezoelectric transducers on the heart affects their output performance. Particularly, Lu et al. [71] implanted a PENG in a swine and verified that the output signal is strongly related to the location and orientation of the PENG placement. They found that the optimal position is between the left apex of the ventricle and the right ventricle of the heart, where the maximum power can be output. Moreover, Li et al. [3] compared 4 sites on the pig's heart and identified the maximum output is at the apex of the heart.

Triboelectric Nanogenerator (TENG)
The kinetic-electricity conversion can also be realized by TENGs, which generate electricity based on the coupled triboelectrification and electrostatic induction. [81] Since almost every material has a triboelectrification effect at different levels (referring to triboelectric series table [82] ), TENGs have a wider range of materials to select from than PENGs. However, the TENG requires two contacting or sliding materials to create the triboelectric charge by rubbing between their surfaces, while PENGs simply generate charges through deformation of piezoelectric materials. TENGs work upon the principle of triboelectric effects, in which certain materials are in frictional contact and then separated from each other, driving electrons to transfer between different materials, and further generating current. [83] On the basis of this Heartbeat 2014 [75] Pacemaker (trigger energy of 1.1 μJ) Heartbeat 2015 [71] Pacemaker PZT ceramics Nanogenerator 3 V d) 400-600 nm (thickness) Heartbeat 2017 [78] Pacemaker PZT ceramics Cantilever-type (fan-folded) 16 Heartbeat 2020 [72] Pacemaker PDMS-infilled microporous P(VDF-TrFE) Cantilever-type (multibeam) 3.22 ± 0.24 V a) 6 mm (thickness) Heartbeat 2021 [73] Pacemaker Nanofibers of PVDF and hybrid nanofillers Ascending aorta pulse 2015 [4] Pacemaker/ICD/cardiac monitor principle, four fundamental working modes of TENGs: vertical contact-separation mode, lateral sliding mode, single-electrode mode, and freestanding triboelectric layer mode [84] have been developed for different target applications, which are presented in detail in Figure 3D-i-iv and paper. [85] The first TENG was proposed by Wang and co-workers [86] in 2012 as an effective energy harvesting technology. It possesses numerous merits including broad material availability, lightweight, and flexibility, which makes TENG considered an ideal power solution in the applications of low-consumption implantable devices.
Currently, the main challenge is to generate high output through effective TENGs. In this regard, triboelectric material pairs matching with strong electron affinity has a significant impact on the output of TENGs. In general, certain triboelectric materials (tendency to gain electrons) such as polytetrafluoroethylene (PTFE), PDMS, Kapton, terephthalate (PET), and Fluorinated ethylene propylene (FEP) are commonly coupled with other specific materials (tendency to lose electrons) such as aluminum (Al), gold (Au), copper (Cu), and sliver (Au).
In addition, the surface charge density on the triboelectric layers also determines the electric output of TENGs. In order to further improve the output performance, numerous studies have explored surface-modified TENGs, which can be divide into four categories: physical, chemical, biological, and hybrid modification. [87] Among them, physical modification (controlling surface morphology and interlayer interaction) and chemical modification (adjusting the surface-charging tendency and modifying the interfacial potential) are widely employed to increase the electric output of TENGs. [88] In particular, the physical surface nanostructure modification is the most common and efficient way. This method introduces diverse surface nanostructures with the aim of enhancing the surface roughness and effective contact area, thereby increasing the charge density between the triboelectric layers. [89] Paper [88,90] reviewed exhaustive surface engineering techniques and novel modified materials in TENGs, we will not elaborate them here.
Since Zheng et al. [91] realized the first implantable TENG, flexible and lightweight TENGs have been significantly encouraging in powering cardiovascular devices. They utilized the TENG to harvest the periodic breathing of a rat and successfully drive a pacemaker, as shown in Figure 4A. A PDMS film with patterned pyramid arrays and nanosurface modified Al foils are employed as a pair of triboelectric layers, in which Au and Al film are served as electrodes. The introduced flexible PDMS and Kapton layers exhibit high flexibility and biosafety. The TENG tested 12 V of V oc and 0.25 μA of I sc output in vitro and harvested 3.73 V of V oc and 0.14 μA of I sc in vivo, respectively. Subsequently, this team proposed a new TENG that was continuously powering the cardiac monitoring system in a porcine in vivo over 72 h, [92] as illustrated in Figure 4B. Notably, the friction layers and electrodes materials are the same as before, but a highly resilient titanium strip was added to the TENG structure to ensure effective contact B) The implanted TENG was driven by the heartbeat of a porcine for achieving a self-powered wireless cardiac monitoring system. a) Schematic structure of the TENG, b) different implant sites on the cardiac: OTR, the right ventricular; ALA, the right ventricular; ALA, the auricle of the left atrium; CB, the cardiac base; LWL, the lateral wall of the left ventricular; IWL, the inferior wall of the left ventricular; and c) the output V oc at different implant sites. C) The TENG collected the energy from blood flow to support the endocardial pressure sensor in a porcine model. a) Schematic structure of the TENG, b) the implant site in the swine's heart, and c) the figure of in vivo experiment. A-a,b) Reproduced with permission. [91] Copyright 2014, Wiley-VCH. B-a) Reproduced with permission. [92] Copyright 2016, American Chemical Society. C-a,c) Reproduced with permission. [93] Copyright 2019, Wiley-VCH. and separation in vivo. After being encapsulated by PDMS and parylene C, the TENG measured 45 V of V oc and a 7.5 μA of I sc in vitro and collected 14 V V oc and 5 μA I sc from contraction/ relaxation movement of the heart in vivo. In addition, tests were located differently on the heart to validate the placement with optimal output performance. Consequently, they found that place the TENG on the lateral wall of the left ventricular can produce the best output. As depicted in Figure 4C, Liu et al. [93] devoted to monitoring severe heart failure through a self-powered endocardial pressure sensor, which was inserted into the left ventricle of a porcine. This research demonstrates that TENGs can collect mechanical energy from the blood flow in the heart chamber to power the endocardial pressure sensor. They used a corona discharge method to modify the surface of the nano-PTFE film (acting as one triboelectric layer) to increase the surface charge density, and the Al foil was employed as the other triboelectric layer. Furthermore, Au layer and Al foil were acted as a pair of back electrodes, and a 3D ethylene-vinyl acetate (EVA) copolymer film was used as a spacer between them. The result shows that the output voltage enhanced from 1.2 to 6.2 V after the surface modification. In 2019, Ouyang et al. [94] demonstrated a symbiotic pacemaker powered by a multilayer-structure TENG. In order to improve the output performance, this study adopted several methods including: employing a nanostructured PTFE triboelectric layer, adding a spacer layer to increase the effective contact area, and introducing the corona charge method to improve surface charge density. In addition, this TENG was fully packaged with flexible Teflon and PDMS films to achieve leak-free structure and stretchability. As a result, the TENG was mounted between the heart and pericardium of a porcine, generating 0.495 μJ energy during each cardiac motion cycle. In summary, various explorations of TENGs, including advanced materials, effective multilayer structures, and placements in vivo have greatly promoted their output performance and applications in cardiovascular implants.

Thermoelectric Generator
Human body waste heat can be directly converted into electricity through TEGs, which also launch a substitute for powering biomedical implants. In the battery section of this review, the TEG has been mentioned as an early transducer for powering the pacemaker by extracting heat from the radioactive isotope decay of nuclear batteries. With the remarkable development of microscale and nanoscale generators, the emerging TEGs cam harvest energy from temperature gradients in human tissue based on the Seebeck effect. [95] The basic module of TEGs is a thermocouple, which consists of a pair of n-type and p-type semiconductor legs connected to electrodes, as shown in Figure 3C. TEGs arrange multiple thermocouples electrically in series and thermally in parallel in a sandwich structure of two insulator layers (or ceramic substrate). Besides, doped semiconductors contain a large number of free carriers, which are holes in p-type materials and electrons in n-type materials. [96] When a thermocouple is subjected to a temperature difference, the Seebeck effect occurs, which allows carriers to migrate from the hot side to the cold side, thereby generating a voltage. The output voltage produced by a TEG is proportional to the temperature gradient, and the formula is shown as follow [96] · out 1 2 where n is the number of thermocouples, ΔT is the temperature difference across the TEG, and α 1 , α 2 are the Seebeck coefficients.
The power transfer efficiency of a TEG is determined by the temperature difference, thermoelectric material features, where the equation can be expressed as Equation (2) [97] T T where T is the average temperature, and Τ h and T c are the temperatures of the hot plate and the cold plate, respectively. The figure of merit (ZT) is a significant factor to evaluate the performance of thermoelectric materials, which can be defined as follow [98] ZT where S is the Seebeck conductivity, σ is the electrical conductivity, and k is the thermal conductivity.
Since semiconductors materials (such as bismuth telluride (Bi 2 Te 3 ), antimony telluride (Sb 2 Te 3 ), and lead telluride (PbTe) alloys) possess a high value of ZT (ZT > 2 [99] ), these materials have been widely employed. However, these inorganic materials have some disadvantages, such as toxicity and rigidity, which are inappropriate for implantable devices. Therefore, conductive polymer materials such as polyaniline (PANI), polyimide (PI), and poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS) are attractive due to the flexibility, biocompatibility, and inherent low thermal conductivity. Nevertheless, these organic materials have a low value of ZT (between 0.01 and 0.25 [100] ), which is 2-3 orders lower than that of semiconductor materials. [101] Therefore, organic-inorganic hybrid materials combine the merits of the two and show bright prospects in thermoelectric materials. Effective methods to develop flexible and high thermoelectric performance materials include nanostructure engineering technology and the semiconductor doping methods (e.g., PANI/graphene thermoelectric composite films, [102] and tellurium-PEDOT:PSS hybrid composites [103] ). Additionally, the innovation in hybrid materials and more detailed features are reported in papers [101,104] In addition to the investigation of materials, temperature difference is another decisive factor for improving efficiency (as presented in Equations (2) and (3)). Several academics have studied the maximum heat gradient across the body and found that the largest difference (ΔT) occurs in the fat layer (near the skin layer) [105,106] A TEG made of doped bismuth telluride materials was reported by Bhatia et al., [107] which employed nearly 4000 thermocouples in series to charge pacemakers and/or defibrillators. The surface area of the TEG is 6 cm 2 and requires a 2 °C of ΔT to drive. This is easy to fulfil due to the 5 °C temperature difference in multiple parts of the body. The TEG generates 4 V of V oc and 100 μW of output power, which presents great potential for powering low-consumption cardiac implants. In addition, Yang et al. [108] presented a prototype of TEG equipped with a specific boost circuit for powering a pacemaker. Moreover, they verified the energy harvesting circuit by utilizing a higher power consumption clock circuit (145 μW) instead of a real pacemaker (with the same pulse interval). When the temperature gradient stabilized at 0.5 °C in a rabbit, the output voltage was measured as 20 mV and then rose to 3.3 V after passing through the boost circuit.
Although previous studies have demonstrated the feasibility of converting body heat into useable electricity that can be used to power cardiac implantable devices, there are few reports in the field of implantable TEGs according to the latest research on energy transducers area. In general, the development of TEG in biomedical applications is hindered by the following basic limitations: A) The design and layout of thermoelectric transducers with hundreds or even thousands of thermocouples are hindered by the ultrasmall surface area. B) The amount of power generation is limited by temperature gradient in body. C) The time-variant temperature difference needs to be maintained all the time.

Wireless Power Transfer Solutions in cIMDs
The rapid growth of research in energy harvesting technology has led to significant improvements in power solutions for low-consumption cardiovascular implants such as pacemakers and defibrillators. However, the in vivo output collected from energy harvesters is too insufficient to be utilized as an alternative power solution for other higher power level cardiovascular implants. Electromagnetic wireless power transfer is an actively adjustable power strategy that can regulate the input voltage to provide stable and adequate support for the load. This is a promising power solution that uses EM energy between the internal receiver and external transmitter to support various wearable and implantable devices without percutaneous wires and batteries (e.g., contact lens, [109] brain, [110] neural, [111] and cardiovascular medical devices). When the wireless implants are initially implanted in the patients, the cIMDs should be disconnected from the external controller. [112] The external programmer will send a modulated wake-up signal to the implantable receiver through a wireless link. The receiver detects the signal and then activates the rest of the chip. In detail, a low dropout (LDO) regulator is needed to generate a stable voltage to support implantable devices. The voltage regulator output is disconnected from the implanted capacitor during startup, which is controlled by a power-on reset (POR) circuit. When the power supply voltage reaches the POR activation threshold, the POR circuit sends out a power-on signal to initialize the chip and start working. [113] It is expected that WPT technology will lead to fully self-reported implantable devices in the future. To evaluate and optimize the performance of WPT in powering cardiovascular implants, this review investigates three significant aspects: A) Miniaturization and compactness of advanced antennas for cardiovascular implants and their radiation performance. B) Human body safety guidelines when exposed to electromagnetic fields and the investigation of changes in dielectric properties under different frequencies. C) The basic principles and power transmission efficiency of various WPT links.
In this section, the specific characteristics and potentialities of different WPT schemes and their state-of-the-art applications in cardiovascular implants are discussed.

Antenna Design and Miniaturization of cIMDs
An antenna is one of the most vital components in WPT, used to transmit and receive signals. In particular, the antenna design is the basic and indispensable work to guarantee efficient power transmission. Herein, we review the types, geometries, and performance of miniaturized and compact antennas used in cIMDs. Advanced WPT technology for powering implantable devices requires simultaneous miniaturization and high-performance implantable antennas. Planar antennas, [114] in particular the rectangular microstrip antenna [115] circular patch antenna, [116,117] PIFA antenna [118] are preferred for small size antenna. These antennas benefit from their simple design and fabrication, are also easily integrated with other electronic components.
In order to further meet the requirement of ultrasmall implantable antennas, the antenna size can be reduced by following methods: A) Geometry configuration: Folded configuration can shrink the size of antenna by increasing the current flowing path, such as meander-shape, [119] spiral-shape, [120,121] and fractal structure. [122] The effective dimensions of an antenna will increase in the form of the meander and/or spiral configuration, which allows increasing the physical path for surface current flow, and ultimately leads to the decrease of resonant frequency compared to the conventional antenna. [123] Moreover, the antenna size reduction factor β depends critically on the number of meander elements per wavelength and the width of the rectangular loops, which is defined as in Equation (4) [124,125] l L β = / where L is the length of the conventional monopole and l is the length of the meander antenna with the same resonant frequency.
B) High dielectric substrate materials: As previously mentioned, the propagation of EM waves is affected by dielectric properties and frequencies. Therefore, the introduction of a high dielectric constant is capable of decreasing the guided wavelength and the length of a patch, which gives the following equation [126] 2 2 where L is the length of the patch antenna, λ g is the guided wavelength, C is the speed of light, and ε eff is the effective dielectric constant. However, the drawback of this approach is that it narrows the bandwidth of the antenna and losses surface-wave. [127] C) Superstrates: The introduction of a superstrate on a microstrip antenna is beneficial to simultaneously reduce the size and enhance the gain and bandwidth of the antenna by forming a larger in-phase electric field on the superstrate surface. [128] D) Shorting pins: By deploying shorting pins connecting the patch and the ground plane of the microstrip patch antennas can be utilized for compressing the size of antennas. [123] The shorting pin can be equivalently modeled as an inductance and added in parallel to the resonant LC circuit of the patch. [129] This makes the patch antenna resonate at a lower frequency and the sized reduction for the antenna can be obtained at a fixed frequency. Table 3 illustrates various antenna types with their special geometry modifications, antenna characteristics and their simulation results, as well as the SAR values in the cardiac tissue. [120,[130][131][132][133][134] In addition to the miniaturized planar-type antennas, significant efforts have been made in designing 3D conformal antennas that can be used as metal stents, thereby saving the additional volume of the antenna used for coronary stents. Figure 5A-J illustrates the planar antenna design of cardiovascular implants and the stent-based 3D conformal antenna design. For instance, Takahata et al. [15] proposed a concept of a stent-based antenna (called "Stentenna"), which uses a stent as an antenna (or an inductor). It incorporates a capacitive pressure sensor to form an inductive-capacitor (LC) tank, which can be inductively coupled with an external antenna. This structure allows highly compact integration of antennas, stents, and biosensors. An extended study of the "Stentenna" [16] led by this team designed a zigzag pattern without bridge struts in a stainless-steel tube, and then formed a helical structure when the tube is expanded, resulting in a significant increase of stent inductance (multiplied by a factor of 3.2). Moreover, a copper layer was covered on the stent owing to its high conductivity, thereby achieving a good resonance of the L-C circuit with a higher Q factor (35 at 150 MHz). In addition, the stent was further coated with a gold layer and a conformal parylene-C layer to guarantee the biocompatibility and electrical insulation of the stent antenna. This "Stentenna" promotes innovation from passive metal stents to compact and smart stents. In particular, Chen et al. [135] reported the first attempt of a smart stent in the clinical setting, where the stent was deployed on a pig through a balloon catheter. The helical-structure stent is composed of an inductive stent and a capacitive pressure sensor, which provides an encouraging demonstration for wireless monitoring blood pressure in the vessel (a resolution of 12.4 mmHg) and also meets the PCI standard procedure. In 2019, Liu et al. [136] presented a stent antenna with good radiation performance. It consists of multiple rings connected by a single connector rather than a conventional multiconnector. Due to the random current distribution of the multiconnector stent antenna, the multiconnector stent antenna showed a higher resonant frequency than the single-connector antenna under the same dimension. This single-connector antenna achieved a S 11 of −17 dB, a gain value of 1.38 dBi, and a radiation efficiency 74.5% at a resonant frequency of 2.07 GHz. In addition to EM-based antennas, acoustic wave-mediated magnetoelectric (ME) antennas have also emerged in recent years. Acoustic-based WPT is another branch of WPT, and it has also drawn considerable attention in powering implantable devices because it can realize ME antennas in nanoscale. Compared with EM-based WPT, which works on EM induction or EM radiation, the acoustic power transfer uses ultrasound waves or vibration to transmit energy. Particularly, acoustically driven nanomechanical ME antennas employ ferromagnetic/piezoelectric thin-film materials, [137,138] and transmit energy through direct and converse piezoelectric effect and magnetostriction effect at the acoustic resonance frequency. Since the wavelength of the acoustic wave is shorter than the EM wave at the same frequency, these ME antennas enable to shrink 1-2 orders [139,140] of magnitude miniaturization than the conventional laminate antennas, which greatly encourages the antenna miniaturization. However, the fabrication difficulty, low power levels, and high demands of materials (e.g., high piezomagnetic coefficient) [141,142] are the main constraints currently faced for using ME antennas. Furthermore, the acoustic-based WPT is not the focus of this review paper, and we will introduce the EM-based WPT in detail in the rest of this section.

Specific Absorption Rate and Dielectric Properties in Human Tissue
The electromagnetic WPT system is based on the EM field to propagate energy from an external source to an internal implant. The attenuation of EM waves caused by tissue absorption will lead to non-negligible power loss and thermal effects. Therefore, the safety of human EM radiation is the primary concern of wireless power and data transmission in implantable devices. The power dissipation in WPT systems must be comply with the safety regulations, in terms of SAR (W kg −1 ) value, to avoid overheating of tissues/adverse health effect. SAR is defined as the EM energy absorbed by tissue per unit mass of tissue when exposed to EM radiation, which resulted in the tissue heat up [143] where E is the induced electric field, and σ and ρ are the conductivity (kg m −3 ) and the mass density (S m −1 ) of the tissue. In order to prevent the tissue from localized overheating, the SAR value must comply with the standards and guidelines of Federal Communications Commission (SAR 1g ≤ 1.6 W kg −1 [144] and IEEE SAR safety standard (SAR 10g ≤ 2 W kg −1 [145] ). Moreover, the dielectric properties (permittivity and conductivity) of biological medium are significantly different from those in vacuum, resulting in a mismatch in EM waves propagation between the free space and biological medium. In addition, the SAR value is heavily affected by frequency. [146] To select a proper operational frequency to minimize the power loss caused by the antennatissue interaction and simultaneously meet the SAR regulation,  [120] Copyright 2011, Hindawi Publishing. B) Reproduced with permission. [134] Copyright 2018, IEEE. C) Reproduced with permission. [132] Copyright 2019, IEEE. D) Reproduced with permission. [133] Copyright 2019, IEEE. E) Reproduced with permission. [130] Copyright 2019, IEEE. F) Reproduced with permission. [131] Copyright 2020, IEEE. G) Reproduced with permission. [15] Copyright 2006, IEEE. H) Reproduced with permission. [16] Copyright 2013, IOP Publishing. I) Reproduced under the terms of the CC-BY 4.0 license (https://creativecommons.org/licenses/by/4.0/). [135] J) Reproduced with permission. [136] Copyright 2019, Multidisciplinary Digital Publishing Institute.
the influence of EM radiation on the dielectric properties of biological tissue is worth investigating.
Permittivity, ε (F m −1 ) (also known as dielectric constant), measures the ability of a substance to store electrical charges and is relates to how well molecules of polarization under a given applied electric field. [147] In order to reduce the interference to the antenna radiation performance, a dissipative medium with a high dielectric constant is preferred. Relative permittivity determines as the ratio of a certain material's permittivity to the permittivity of vacuum (ε 0 = 8.85 × 10 −12 F m −1 ) ε r = ε/ε 0 . Conductivity, σ, represents the ability of a substance to carry charges, and it significantly affects the radiation efficiency, absorption cross-section, and stability of an antenna. [148] Significant efforts have been made to estimate the dielectric properties in human phantom models under different operating frequencies. Table 4 summarizes the electrical properties with frequency changes in skin, fat, and muscle layers. [149][150][151][152][153][154][155] This table indicates that the overall decreasing and growing trends of relative permittivity and conductivity change with increasing frequencies, separately. Therefore, it is pervasively agreed to select the optimal operational frequency at relatively low frequencies to maintain safe tissue absorption and minimal power loss. Furthermore, a charging control unit for reliable and safe charging of fully implantable medical systems is essential. In general, the control unit in WPT systems connects to charging circuit and the implantable device to coordinate the transmitted power and data. In particular, the power transmitted from the transmitter depends on the feedback of the receiver. For instance, when the transmitted power exceeds the power level required by the receiver, the transmitter will stop supplying power and enter a low-power standby mode. [156] In addition to the above discussion, this section also introduces the working principle of various WPT power links, antenna design, PTE, and advanced applications in cardiovascular implants. In general, WPT technology can be categorized into near-field, mid-field and far-field WPT according to its transmission distance and radiation mode. Figure 6 illustrates different field regions of an antenna. The near-field region can be further divided into reactive near-field and radiative where d is the radius of transmission region, D is the largest dimension of antenna, and λ is the wavelength.

Near-Field WPT
Near-field WPT has been widely used to transmit energy over short distances (limited to the range of millimeters (mm) and centimeters (cm)). Near-field WPT technology can be divided into three categories according to the working principle: inductive coupling, magnetic resonant coupling (MRC), and capacitive coupling (CC). Each scheme has its own merits and demerits, which will be discussed in detail. For instance, inductive coupling exhibits the highest power transfer efficiency, while magnetic resonant coupling possesses the largest transmission distance.

IC-WPT
Inductive coupling is the most established power scheme and has been extensively used in various wireless charging scenarios. It is a nonradiative technology that works in the nearfield region based on the EM induction. A stable source supplies power and generates current in the transmitter coil (Tx), which induces a time-varying magnetic field between two coils (in the range of millimeter (mm) and centimeter (cm)), and then generates a voltage in the receiver coil (Rx). The IC-WPT exhibits remarkable high efficiency under the short distance and perfect alignment between Tx and Rx. The main factors that determine the PTE are the coupling coefficient "k" and the quality factor "Q," as shown in Equations (10)-(12) [158] where M is mutual inductance, and L 1 , L 2 are the self-inductance of a pair of coils separately. In the derived equation, d 1, d 2 are the diameter of the transmitter and receiver antenna, and D is the separation distance. In Equations (11) and (12), L is the inductance and R effective is the self-resistance of antenna, Q 1 , Q 2 are the quality factor of two coils, Q 2L = Q 2 Q L /Q 2 + Q L , and Q L = R LOAD /2πfL 2 . The PTE of this power link is highly related to the size of the coils and the distance between the Tx and Rx coils. Although the inductive coupling presents a high PTE within the limited distance, the PTE is extremely sensitive to the distance limitation and precise alignment between Tx and Rx in the nonresonant state, which is known as weakly coupled (or loosely coupling). In particular, the intensity of the magnetic field is inversely proportional to the distance (1/D 3 ). Obviously, the PTE of inductive coupling attenuates significantly with the separation increase.

MRC-WPT
Inductive coupling exhibits a relatively high-power efficiency but hindered by short transmission distance and perfect coils alignment. If the separation distance between Tx and Rx exceeds this range or the coils are misaligned, the PTE will drop dramatically. Therefore, a strongly coupled MRC was initially investigated by Kurs et al. [159] in 2007 to extend the transmission distance up to 2 m by introducing a four-coil WPT system. As illustrated in Figure 2H, the four-coil WPT system consists of a source coil, a transmitter coil, a receiver coil, and a load coil. In addition, external compensation capacitors are introduced to tune the resonant frequency of the resonant coils (Tx and Rx). [160] Therefore, adjacent coils are coupled to each other through magnetic resonance coupling, which allows compensation of magnetic flux leakage to improve PTE with a longer transmission distance compared to the IC-WPT.
To guarantee the high PTE of MRC-WPT system, the consistency of resonance frequency is essential, as expressed in Equation (13) [159] f LC π = 1 2 r (13) where L and C are the self-inductance and parasitic capacitance of the resonant transmitter or receiver coils. In addition, the coils can be modeled by a series RLC circuit and the quality factor (Q) is given by Equation (14) [161] where R is the parasitic resistance of the coils. Furthermore, the PTE is calculated based on the equivalent circuit, which is illustrated in the paper. [162] The MRC-WPT system can be considered as a two-port network and the PTE of the wires power transfer system can be analyzed by scattering parameters S 11 and S 21 , as presented in Equations (15) and (16) [162,163] / Re 1 / where S 11 is input reflection coefficient, Z 0 = 50 Ω is the characteristic impendence, Z in is the input impedance, and S 21 is the forward transmission coefficient can be expressed as follows

S k k k Q Q Q Q k Q Q k Q Q k Q Q k k Q Q Q Q
where k 12 (source to Tx), k 23 (Tx to Rx), and k 34 (Rx to load) are the coupling coefficient of adjacent coils, which can be found in Equation (10).
In general, two-coil and four-coil configurations are widely adopted by MRC-WPT systems. [164,165] In the two-coil topology, the power source and the load directly connected to the Tx and Rx, eliminating the additional source and load coils. However, it has been demonstrated that the two-coil system exhibits lower tolerance with the distance and rotation variations compared to the four-coil configuration. [166,167] For instance, a twocoil MRC system reported by Monti et al. [168] delivered up to 1 mW power at a resonant frequency of 403 MHz to drive a pacemaker. In this study, a planar spiral-shaped resonator (SSR) and a square split ring-shaped resonator (SRR) were designed on Arlon 880 substrates served as the Tx and Rx, separately. Moreover, they also tested the sensitivity of the MCR link in terms of distance and misalignment. It shows that the PTE decreases rapidly as the distance exceeds 5 mm and the rotation angle is greater than ±30°. In addition, Yellappa et al. [169] reported a four-coil MRC-WPT system for biomedical implantation. They designed the circular and spherical shape coils and compared their performance in the case of lateral and angular misalignment. The external coils and implantable coils are fixed at an initial distance of 10 mm initial distance, the PTE is 78.5%. Then the coils moved to 40 mm and the rotation changed from 0° to 90°. The PTE of the four-coil configuration is robust within 15 mm and up to 50° of angular misalignment. In general, the results verified that the spherical shape coil presents a higher tolerance with overall misalignment than the circular coil Although aligning the coils can maximize the coupling between the coils, it is usually not possible in biomedical implants. For example, blood flow and body motion can affect the position of internal coil. Compared with the inductive coupling, the MRC-WPT scheme can mitigate the interference induced by the misalignment over a longer transmission distance. The four-coil structure MRC is preferable by adding the extra high Q resonant coils, which is strongly coupled to the source and load coils. However, the four-coil configuration contributes to the volume of the WPT system, causing the burden for implantation in the body. In addition, the phenomenon of frequency splitting of both two-coil and four-coil structures can lead to a high PTE but low load power at the resonant frequency. [161,170]

CC-WPT
Capacitive coupling WPT uses one or two pairs of parallel conductor plates to instead of coils as the Tx and Rx. [171] This scheme uses induced displacement current as the carrier, which is derived from the electric field (compared to the magnetic field of the inductive coupling) between the conductor plate pairs to achieve wireless power delivery. [172] Therefore, CC-WPT profits from a lower EM interference and a higher misalignment tolerance compared with the inductive coupling. [171,173] The most common CC-WPT system requires two metal plates as Tx, and the other two served as Rx to form a coupling capacitor separating by a medium to provide a power flow loop (to form receive and return paths of the displacement current), and then connect to an alternating current (AC) source and a load, separately. In this approach, the coupling capacitor (similar to the Q factor in the inductive coupling) plays a vital role in the amount of power transfer. The equations of coupling capacitance and displacement current are defined as Equations (17) and (18) [172,174] where ε 0 , ε r (ω) are the free space permittivity and frequency variant permittivity, A is the effective area, and D is the separation between the plates, and t E ∂→ ∂ is the rate of the electric field delay. The power loss in capacitive coupling includes tissue loss, conduction loss, and source loss. Among them, the conductive loss and source loss are much smaller than the tissue loss and can be ignored when operating in MHz to GHz range. Therefore, the tissue loss can be quantified as equivalent to resistance, and the PTE of a capacitive coupling link is defined as Equation (20) [175] R Real jD A r ωε ε ω ( ) where R L is the load resistor and Γ is the reflection coefficient. However, the value of coupling capacitance is constrained by the size of conductor plates and the small ε 0 value (ε 0 = 8.85 × 10 −12 F m −1 ). Moreover, when designing the CC-WPT link to realize the optimal PTE, a trade-off between parameters A, D, and operational frequency need to be considered. Due to the low coupling capacitance and short transmission range (in the millimeter range), a resonant compensation network (including capacitors and inductors) is commonly required to resonate with the coupler and boost the power transmission. [176,177] Due to the limited conditions and performance of CC-WPT, it has not been widely investigated in the application of biomedical implants. Early research on capacitive coupling in implantable medical systems started in 2009. [173] It listed several advantages of capacitive coupling over inductive coupling (e.g., low EM interference and high immunity with misalignment [178] ) in biomedical applications, and also presented a design of power and data telemetry link for biomedical implants. However, this scheme is constrained by the bulky volume (e.g., four-plate structure) and low power output. For instance, Aldaoud et al. [40] tested both inductive and capacitive power links for powering stents in vitro. The result shows that the overall PTE of inductive coupling is higher than its capacitive coupling counterpart.

Mid-Field WPT
The mid-field WPT offers an innovative solution for powering over a longer range at a higher frequency (low GHz range) than the reactive near field coupling methods through a hybrid induction mode (external) and radiation mode (internal). Midfield WPT benefits from a patterned metal plate with four ports to generate an ideal field pattern. By manipulating the ports, the metal plate can induce concentrated and adaptative energy transmission in deep tissue through the exponential decay of the magnetic field in the propagation mode. [179,180] Particularly, the four ports of the patterned metal plate are excited by four independent phased RF signal, which is used to operate the focused region of the field and further improve the efficiency. As depicted in Figure 7A, the patterned metal plate is placed near the air-skin interface, and then the output field can be concentrated in small ranges (less than the vacuum wavelength), thereby generating a high energy density region. [180] The overview of SAR, dielectric properties and frequency in the previous section indicates a general concept that low frequencies are more suitable for WPT propagation in biological tissues. However, by employing a full wave analysis of WPT in implantable devices, it has been proved that the optimal frequency of miniaturized coils (in the range of millimeter or micrometer) is in the GHz range, as presented in Equation (21) [181] 1 2 where c is speed of light, ε 0, ε ∞, τ are the dielectric properties of tissue in the Debye relaxation model, d is the distance between Tx and Rx. Compared with the most common WPT in the KHz or MHz range, this method can dramatically shrink the size of the receiver coil by 10 4 times due to its higher operating frequency (midfield). [181] Mid-field WPT is preferrable for powering deep tissue (>5 cm) implants, especially for cardiovascular implants, which has motivated many researchers to contribute to this promising approach. For instance, Ho et al. [180] demonstrated the feasibility of mid-field WPT in powering implantable devices, as shown in Figure 7B-a,b,c. The small receiver coil (2 mm in diameter) in the heart received 195 μW of power from 500 mW input power and spaced at least 5 cm apart. Later on, Das et al. [182] applied the mid-field WPT technique in powering a leadless pacemaker at the optimal frequency 1.5 GHz (derived by calculating formula Figure 7. The working principle of mid-field wireless link and its applications in cardiovascular implants: A) work principle of mid-field WPT. B-a) Scheme of mid-field source (metal plate), b) magnetic field distribution in multitissue layers, and c) its application in powering electrostimulator implanted in a rabbit. Reproduced with permission. [180] Copyright 2014, Proceedings of the National Academy of Sciences of the United States of America. C-a) Mid-field transmitting antenna, b) spiral type implanted antenna, and c) antenna measurement set up. Reproduced with permission. [218] Copyright 2015, IEEE. D-a) Mid-field transmitter, b) quad-band implantable antenna, and c) The SAR distribution of Tx at 1470 MHz. Reproduced with permission. [185] Copyright 2020, IEEE. E-a) Transmitting antenna, b) PIFA implanted antenna, and c) simulation result of return loss. Reproduced with permission. [184] Copyright 2020, IEEE.
(21) by using heart tissue). The mid-field transmitter antenna was designed as a slot array-like metal plate based on a Teflon substrate, and a spiral-shaped PIFA antenna (Rx) was fabricated on a flexible polyamide substrate, as shown in Figure 7C-a,b,c. Under the midfield WPT system, they delivered 500 mW power at Tx and received 2.9 mW power at Rx, which is sufficient for most cardiovascular implants. Recently, Basir and Yoo [183] proposed an efficient WPT system based on the mid-field technique for deepbody implants. The proposed configurations of Tx and Rx are shown in Figure 7D-a,b,c. The mid-field transmitter constructed a slotted structure and four ports on Teflon substrate to efficiently concentrate the power in the tissue through appropriate phase control. Meanwhile, the authors developed a compact implantable antenna with a spiral-shaped structure that supports quadband operation. From their measurement results, the system can deliver 6.7 mW power for deep tissue implants over 5 cm, which shows great potential of powering the cardiovascular implants. Subsequently, Nguyen et al. [184] reported a compact mid-field Tx configuration to power miniature implants, as presented in Figure 7E-a,b,c. A planar PIFA type antenna with a Z-shaped slot (on the ground layer) and three short pins was used for the Rx (with a size of 9 × 13 mm 2 ) to achieve the maximum gain value. This mid-field WPT system operates at an optimal frequency of 1.5 GHz and shows a highly concentrated magnetic field in the tissue by observing the current density and the magnetic field. The Rx was placed in a heart tissue layer 55 mm away from the Tx, and 5.6 mW of output power was measured at the Rx from 1 W power coupled to the Tx. In addition, they tested the performance of the midfield WPT under various misalignment conditions (lateral and angular misalignment). Consequently, the interference caused by the misalignment can be basically elimi-nated under the mid-field WPT according to the stable S 21 value and PTE with all misalignment cases, which is an encouraging solution for implantable devices.
Mid-field WPT provides a prominent strategy that is more suitable for powering cardiovascular implants than other conventional near-field WPT schemes. It surpasses the limitations of short transmission distance and misalignment. Meanwhile, the controllable metal plate can adjust the magnetic field pattern in tissues to increase the PTE. In addition, this approach promotes the miniaturization of implantable receivers in the low GHz range.

Far-Field WPT
Far-field WPT/far-field RF refers to the power that can be transmitted by laser beams or microwaves [185] based on EM radiation. The propagation of EM radiation from the Tx to the Rx in a far field WPT system is depicted in Figure 8A. [186] In contrast, the transmission power attenuates rapidly as the distance between Tx and Rx increases in the near-field coupling, while the beam of far-field WPT can radiate a long distance (in the range of meters). Nevertheless, a long transmission distance will cause significant path loss, resulting in low power transfer efficiency. This is the major drawback of farfield WPT due to the energy transmitted by EM waves radiates and dissipates in all directions. [187] The architecture of an end-to-end far-field power link is presented in Figure 8B, [188] which includes several energyconversion processes: direct current (DC) to RF conversion (P Tx ), effective radiative power from Tx (P in ), free-space transmission (P inc ), incident power from Rx (P Rx ), and RF to DC Figure 8. The working principle of far-field radiation and its application in cardiovascular implants: A) the work principle of far-field radiation. Adapted with permission. [1] Copyright 2020, Wiley-VCH. B) End-to-end power link of far-field WPT, where η 1 × η 2 is the transmit efficiency, η 4 × η 5 is the receive efficiency, and η 3 is the free-space propagation efficiency. C) RF-powered leadless pacing system with a wearable transmit-antenna array, an implantable rectenna, and its animal experiment in an ovine. Adapted with permission. [196] Copyright 2018, IEEE. conversion (P out ). Therefore, the overall power transfer efficiency is presented as Equation (22) [188]  The PTE of far-field WPT is hindered by multipath loss. In particular, the transmit efficiency (η 1 × η 2 ) and the receive efficiency (η 4 × η 5 ) are of great significance to the optimization of PTE in the far-field WPT. Regarding the transmit efficiency, directional antennas or antenna arrays can be employed to mitigate the path loss by accurately concentrating the EM radiation to the targeted receiver. [189] The antenna array with beam steering characteristics can manipulate the beam pattern to achieve a directional main lobe with low side lobes, thereby enhancing the directivity. [190] When receiver antenna receives the radiated RF power, it will pass through the rectifier circuit and coverts it into the DC power to the load. Therefore, an efficient "rectenna" (combined antenna and rectifier circuit) is also crucial for improving the receive efficiency, especially for converting RF to DC.
The first far-field wireless microwave transfer was demonstrated by Brown in 1964. [191] It utilized a rectenna to receive microwaves to drive a model helicopter. In the past few years, extensive work has been done in investigating and designing high-performance rectenna. [192,193] Particularly, it is recommended to use circularly polarized (CP) antennas to solve the polarization mismatch and multipath reduction issues. [194] When the rectenna rotates, the output DC voltage will not be affected too much due to the circular polarization allows the rotation of the circuit. [195] Therefore, the CP antenna overcomes the misalignment issue between Tx and Rx.
In recent years, RF-based far-field power strategies have aroused great interest in wearable and implantable devices with the Internet of things (IoT) medical systems. For instance, the paper [196] proposed an RF-powered leadless pacing system with a wearable transmit-antenna array and an implantable rectenna (Rx), as shown in Figure 8C. They adopted a planar dipole antenna with a fractal geometry as the receiver and fabricated it on a Rogers TMM10i substrate (tan δ = 0.002, ε r = 9.8), and then matched it with a modified-Greinacher rectifier circuit. The efficient transmit array based on the RT/Duroid 6010 (thickness of 0.254 mm) was designed with two semioval shaped truncations in the ground plane to realize the directional beam of the antenna. As a result, the overall wireless pacing system was successfully verified in an ovine model in vivo, and finally achieved 65% of PTE at operating frequency of 954 MHz. Another work in the study of far-field RF powered leadless pacemaker was developed by Abdi and Aliakbarian. [197] The rectenna system consists of a CP spiral PIFA antenna and a single-diode detector based on Schottky diodes, and then measured in a three-layer tissue model. This system achieved a 40% RF-DC power conversion efficiency and measured a 0.2 V DC output voltage from an input power level of −20 dBm.
Compared with near-field WPT, far-field WPT significantly enhances the flexibility in Tx/Rx distance and movement. However, safety issues caused by RF exposure and the low efficiency caused by the polarization mismatch and multipath are the main concerns of far-field WPT.
For safety reasons, the RF-based far-field WPT should operate in the low frequency range with high PTE. In addition, the PTE of far field RF can be improved by the DC to RF, RF to RF, and RF to DC subsystems. In particular, the CP beam-steering antenna array and efficient rectenna circuit are the directions that should be contribute to the future. In summary, a comprehensive comparison of the state-of-the-art WPT system along with the characteristics of transceivers, power specification, and transfer distance, is presented in Table 5.

Bidirectional Wireless Data Transfer and Communications
The wireless power link is the premise of continuously performing the functions of biomedical implants. In addition, patients with chronic cardiovascular diseases require long-time physiological data monitoring and therapy. Hence, the wireless link is also essential in carrying bidirectional data for the communication between internal implants and external systems. The aforementioned wireless links for powering are available for data transmission as well. Presently, near-field communication is dominant in wireless data transfer (e.g., near field communication (NFC)) due to its lower power loss and higher safety when interacting with the human tissue.
Wireless data communication in IMDs has been developed to control and collect the biological data from the implantable devices to external systems, a bidirectional wireless data link for cardiovascular implants is presented in Figure 9. [198] Implantable biosensors can detect and collect specific biological signals through sensing electrodes. However, the collected signal is weak and noisy, so it needs to be postprocessed through an amplifier and a filter to adjust its amplitude and remove the noise from the biological signal. The amplified signal is then interfaced to an analog-to-digital converter (ADC), which is used to convert the signal to a digital signal. The digital signal is further sent to a DSP processor, and then transmitted to the external system by using an appropriate modulation scheme.
In general, the data link can be divided into uplink (back telemetry) and downlink (forward telemetry). In implantable biomedical systems, the physiological data feedback from internal implants to the external unit is defined as uplink, and the data sent from outside system to internal implants to control the device is defined as downlink. In particular, the carrier data of uplink and downlink are transmitted through modulation and demodulation technique.
With the rapid advancement in wireless links and biomedical implants, wireless power and data systems are expected to be applied to the monitoring and treatment of cardiovascular implants. However, achieving simultaneous high PTE and high data rates over a single link is a major challenge of the near field scheme. In general, low operating frequencies (MHz range) are preferable since the fewer losses and higher efficiency in the near-field link. Nevertheless, compared to higher frequencies, lower frequencies generally have a narrower bandwidth, which limits the channel capacity and data rate. In addition, high data rates usually require high power consumption, which rapidly depletes the energy of the entire wireless link. Alternatively, multiple dedicated wireless links (multicarrier) have been proposed to separate the power and data transmission links to achieve their respective goal. [199] However, the multicarrier will introduce multiple coils, leading to complicated link design, space occupation, and crosstalk between links. [200] From this perspective, compared with the multicarrier, the single carrier profits from better robust coupling thus realizing more reliable data transmission. [201] Moreover, appropriate modulation/demodulation technique over a single link may resolve the conflict between power and data link requirements. In particular, a suitable modulation/demodulation scheme can offer high reliability and low power consumption on the datalink by reducing the data error rate and power dissipation. [202] In this section, we will introduce and review the fundamentals of various modulation/demodulation schemes for uplink and downlink communications and their applications in cardiovascular implants.

Downlink Data Transmission
Modulation is the process of impressing information from an input signal into a carrier signal by modifying the carrier signal parameters (such as frequency, amplitude, and phase), while demodulation is used to extract information from the carrier signal or recover the original signal. [202] Presently, various modulation and demodulation techniques have been widely used to perform downlink data transmission from external devices to wireless implants (such as amplitude-shift keying (ASK), frequency-shift keying (FSK), phase-shift keying (PSK), and their derivatives). An ideal modulation scheme for wireless implants should have the characteristics of high data rate, low power consumption, low bit error rate (BER), and narrow bandwidth. [203] However, each type of modulation has its advantages and disadvantages, it is necessary to select the appropriate modulation method according to the specific requirements. ASK or its variant on/off keying (OOK) modulation/demodulation scheme is most commonly used for wireless implants because of its simple architecture and low power consumption. [200] Nevertheless, the evident drawback of this linear scheme is vulnerable to noise due to the carrier amplitude is easily corrupt by the interference. [204] Therefore, the probability of BER in OOK is higher than that of FSK and PSK, [203] where BER is defined as the ratio of the number of bit errors to the total received bits. The noise immunity of a data link is measured by BER versus signal-tonoise ratio (SNR). FSK technique is suitable for sending binary data with two carrier frequencies. [205] In particular, FSK benefits from better noise immunity, but it subjected to limited data rate and larger bandwidth occupation. [206] Alternatively, PSK also exhibits less susceptibility to noise for a given SNR and is more power-efficient than ASK. [203] In PSK modulation, the phase of the carrier is modulated to represent the binary data. Among various variants of PSK, binary-PSK (BPSK) is the simplest scheme. For instance, Hu and Sawan [207] employed a fully integrated BPSK as a demodulator in the inductive coupling link of IMDs, with a power consumption of 0.7 mW and a data transmission rate of 1.12 Mbps. To achieve higher data rates, other PSK-variants (such as quaternary-PSK(QPSK), [208] offset-QPSK(OQPSK), [209] differential PSK (DPSK), [210] and highorder PSK methods) have been investigated in biomedical applications.

Uplink Data Transmission
Uplink data transmission is used to record the parameters inbody and transmit data to the external unit. In particular, the most extensively employed passive impedance modulation for uplink data transmission is load shift keying (LSK). The LSK uses an on-off switch to change the load impedance of the implant to transmit data. The change of load impedance will induce the current changes in the implantable circuit, and further cause the current change of the external circuit on the inductive link. [211] Nevertheless, when the weak coupling occurs (coupling coefficient (k) = 0.01 in general), [212] this modulation exhibits low reliability. In addition, the mistuning on/off state of LSK [213] will reduce the PTE of power transmission. Owing to that, an advanced alternative solution called passive phase shift keying (PPSK) has been proposed [214] to address the above shortcomings of LSK modulation. The PPSK uses a similar circuit design to the LSK, but is distinguished by its more accurate state (on/off) operation than the LSK, which enables a rapid transient response according to impedance change. [215] In particular, Jiang et al. [215] reported a PPSK modulator for biomedical implants, using a 13.56 MHz carrier frequency under the weak coupling condition to achieve a data rate of 1.35 Mb s −1 and a BER of less than 1 × 10 −5 for uplink data transmission. They also verified the effect of PTE with the PPSK and showed that the PTE was only reduced by 6% PTE when PPSK modulation was applied. In addition, cyclic on-off keying (COOK) modulation is another substitute that presents a higher data rate than LSK and PPSK. The COOK scheme is prominent in introducing a shorting switch on the LC tank, and it only takes one single cycle to close the switch, thereby achieving a high data rate. [216] In addition, the COOK synchronous short circuit does not affect the resonance effect or reduce power loss. For example, Ha et al. [216] verified the performance of COOK modulation to simultaneously achieve efficient power transfer (maximum 89.2%) and high data rate (6.78 Mb s −1 ) on a single 13.56 MHz link. Considering the reliability, simplicity and size constraints, the single link is preferable for simultaneous power and data transmission. However, achieving high effectively bidirectional transmission is the main challenge in developing and exploiting appropriate modulation/demodulation technique.

Conclusion and Future Direction
Long-term implantation of biomedical devices is indispensable for chronic cardiovascular patients to regulate or replace the abnormal or missing functions in the body. Wireless and battery-free power technologies allow the support for noninvasive devices for diagnostic and therapeutic purposes without repeated surgical procedures, a comprehensive comparison of all the battery-less power strategies for cIMDs is presented in Table 6. Particularly, energy harvesting technologies have emerged to collect the dissipated energy in the human body or ambient to realize self-powered implants. These energy transducers based on advanced flexible materials have been tested in animals to prove the feasibility of powering low power cIMDs. In particular, energy from the cardiovascular system is considered to be an ideal source that can directly power cardiovascular implants through PENGs or TENGs. However, the output of biochemical and mechanical energy in the body cannot satisfy the power requirement of the cardiovascular implants with relatively high-power dissipation, such as smart stents. Moreover, an additional data transmission system needs to be constructed to record the internal conditions of the cardiac, resulting in additional circuits and unwanted percutaneous wires.
Instead of passively harvesting energy, EM-based wireless transfer systems are expected to eliminate batteries and percutaneous wires over a single link that can guarantee a continuous and noninvasive operation of the cardiovascular implants. In addition, WPT is designed to provide a stable and controllable wireless power supply from the external transmitter unit. However, the inductive coupling WPT suffers from the short transmission distance and poor misalignment tolerance, resulting in weak coupling. Although magnetic resonant coupling exhibits higher immunity against the misalignment and achieves a longer distance by introducing tuning resonant circuits, the complex circuitry design and large occupation of implantable circuits hindered its application in biomedical implants. Furthermore, far-field RF suffers from side effects of low efficiency and localized heating in the body caused by the multipath absorption and scattering.
Consequently, mid-field WPT has marvelous potential in power solutions for deep cardiovascular implants. It manipulates the EM field to be concentrated in the tissue through a patterned metal plate, thereby generating a high energy density region and further enhancing the PTE. In addition, compared with the near field WPT, the mid-field WPT addresses the interference induced by misalignment and extends the transmission distance. Furthermore, due to its optimal frequency in biological tissues in the sub-GHz range, this solution also promotes the miniaturization of the antenna.
In terms of data transmission, advanced modulation/demodulation technology is a breakthrough to realize high power and high data rates with low BER over a single link. In particular, the COOK scheme has the potential to allow high data rates and minimize power loss. By further research on wireless power and data transfer technique under safety regulations, fully wireless and battery-free cardiovascular implants are thriving and can be realized in the near future.