Bioelectronic Therapeutics: A Revolutionary Medical Practice in Health Care

Abstract

The emerging field of bioelectronic therapeutics unfolds great opportunities for treating numerous neurological and inflammatory conditions by utilizing the amalgamation of molecular medicine, neuroscience, engineering, and computing. These innovative treatments leverage advanced technology to precisely identify, design, and regulate electrical signaling patterns in the nervous system, addressing multiple diseases. Modifying neural signaling patterns to produce therapeutic effects at a particular organ may blur the lines between conventional medical practices. These modify the neurological behavior using electrical, magnetic, optical, and ultrasonic pulses through closed-loop systems to optimize neural behavior. The Food and Drug Administration (FDA) has approved numerous invasive and noninvasive bioelectronic devices, in the treatment of various neuronal diseases and non-neuronal diseases. Furthermore, the FDA has approved many devices for clinical studies. The field of bioelectronics encounters challenges in integrating with the health care system, including incomplete understanding of human nervous anatomy, neuronal function, membrane potential, and technological limitations. This review aims to explore bioelectronics therapeutics, their role or action in challenges to growth and their solutions, and the prospects of bioelectronic therapeutics.

Introduction

The human brain and the nervous system control the body’s functions by communicating through neural impulses. In today’s health care, treatments often rely on drugs that act through neural receptors, but their inability to target specific areas precisely can lead to side effects. Therefore, there is a probability that medical therapy may evolve around the use of electrical impulses, enabling more precise treatments that interact with the body’s stimuli during disease states.^1,2 Bioelectronics has emerged as a viable alternative for pharmaceutical companies, complementing the traditional healthcare system within the medical device industry. The concept of bioelectronics began in the 18th century when Sir Luigi Galvani connected a voltage source to a pair of detached frog legs, causing them to move.³ This discovery paved the way for the development of bioelectronic technologies, which gained significant momentum with the invention of the pacemaker. Glucose monitoring devices, electrocardiographs, cardiac pacemakers and defibrillators, blood pressure and flow monitors, and medical imaging systems, are all examples of bioelectronics.^4,5

Bioelectronics is a field that integrates biology with electronics to develop devices that interact with living organisms to monitor, diagnose, and treat medical conditions. These devices interface with biological systems, such as the nervous system, muscles, or tissues, using electrical signals or impulses to collect data and deliver therapeutic interventions. Bioelectronics is revolutionizing health care by providing innovative, precise, and personalized solutions that improve on traditional methods by minimizing side effects, enhancing monitoring capabilities, and enabling real-time interventions.^6,7 Bioelectronic systems are electrifying multidisciplinary approaches compared to tiny implantable devices that deliver electrical pulses to control the peripheral nervous system (PNS) commotion and attain therapeutic effects. Electrical pulses have been used by many devices known as neural prostheses to connect with the neurons to advance health and save lives.^8,9 The goings-on of a dysfunctional, damaged nerve and a neural circuit can be modified and substituted by these devices.¹⁰ Bioelectronic devices empowered by advancements in engineering and micro- and nanotechnologies can revolutionize physiological responses through advancements in prosthetics, artificial implants, and hybrid electronic-biological systems. These devices utilize microchip-controlled electrode arrays that harness electromagnetic, mechanical, and chemical energy directly from the body.¹¹ Some of the scopes of bioelectronics are summarized in Table 1.

Table 1.

Subfields/Scope of Bioelectronic Therapeutics^12–14

Bioelectronic in therapeutics	Subcategory	Comments	Related disease
Cardiac cycle and rhythm	Cardiac rhythm	They are used as pacemakers.	Bradycardia, tachycardia, and heart failureb.
	Electrophysiology	Catheter-based therapy is done to learn about the workings of the heart so that the collected data can be used as a neural biomarker.	Atrial fibrillation and ventricular tachycardia.
Auditory and ocular	Cochlear implants	For capturing, digitalizing, and transmitting sound use the data as a neural biomarker.	Hearing and body balance loss/impairment.
	Retinal implants	Capture, digitize, and transmit visual information to use the data as a neural biomarker.	Retinal degenerative disorder and loss of visionb.
Brain and CNS	Spinal cord		Chronic paina
	Brain		Treatment-resistant epilepsy, Parkinson’sa, depression, Obsessive Compulsive Disorder, and Post-traumatic Stress Disorder.
Peripheral nervous system	Vagus nerve		Depressiona, Epilepsya, Rheumatoid arthritis, migraine, inflammatory bowel disease, obesity, and heart failure.
	Sacral nerve		In overactive bladder and urinary incontinencea.

The table explains the various sub-fields of Bioelectronic therapeutics. Some of the fields mentioned are under clinical study and will be the future of personalized bioelectronic medicines, whereas ^aare FDA-approved ones, and ^bare also FDA-approved though further studies are still needed (all discussed later in the review).

CNS, central nervous system.

Structural Featuring of Bioelectronics

Bioelectronic devices can be categorized based on their working mechanism and loop system. Open-loop systems operate through tonic stimulation based on predetermined parameters rather than feedback from stimulation consequences. However, changes in tissue surrounding implants and patient movement can render them ineffective (Fig. 1). However, a closed-loop system proceeds whereby the devices obtain patient input that modifies the stimulation setting. Such a closed-loop system relies on the body’s neural signals as biomarkers to indicate when, where, and how to administer stimulation, that is, acting as a feedback system with comparatively better results than open-loop systems (Fig. 2).^15,16

FIG. 1.

Systemic presentation of the working of closed-loop bioelectronics therapeutics.

FIG. 2.

A comparison of open- and closed-loop systems. The red arrow explains the open-loop system explaining that the pulse generator consistently applies the stimulations. The orange arrows depict the closed-loop system where the system completely works on the feedback system. Here brain signals are recorded as EEG and ECoG and the point of neuromodulation is determined that is used to trigger for injury prevention, advice on taking drugs, and stimulating drug delivery. EEG, electroencephalogram; ECoG, electrocorticogram; LFP, local field potential; AI, artificial intelligence.

Bioelectronics is composed of several components that enable it to modulate neuronal impulses. Briefly, the components are as follows.

Electrode

The component is linked and positioned close to the nerve for impulse modulation. These are made of inorganic materials such as silicone and platinum, lacking carbon in their core structure, have low voltages, and are capable of stimulation and recording. The size of the electrodes cannot be further reduced, as doing so would diminish the charge capacity of platinum and titanium nitride electrodes.¹⁷ New electrodes can be miniaturized with boost charge injection capacity through the transduction of electronic-ionic coupling signals and intrinsic injection capacity by using active iridium oxide electrodes.^18,19

Another problem of oxide layer formation is prevented by using organic electrodes, which facilitate charge transport, size reduction, and refining spatial resolution. Both electronic and ionic transport can be supported by poly(3,4-ethylenedioxythiophene) doped with poly(styrene sulfonate). Organic electrodes can transduce ionic signals originating from neurons into electrode signals that are recorded by microelectronic devices, and vice versa.³

Recently, polymers such as polyimide and parylene-based polymers have advanced the development of flexible peripheral nerve interfaces.^20,21 Innovations in flexible electronics are driven by materials like polyimides and parylenes, which can integrate microscale compounds and produce features suitable for bioelectronics.^22–24 Furthermore, silicone offers desirable mechanical properties with relatively low and tunable values for Young’s modulus,²⁰ high extensibility, and bulk hydrophobicity for electrode formation.^25,26

High electronics cuff electrode

Such electrodes feature a cuff designed for implantation on the external surface of the nerve, using platinum foil. For example, a monopolar nerve cuff electrode has a platinum-coated wire connected to the contact surface, which is joined to a junction and attached to a stainless-steel lead crimped to a gold pin, all encapsulated in a silicon elastomer. A window can be cut on the contact to establish an interface with the nerve.²⁷ With the use of a nonpenetrating nerve cuff electrode, electrical stimulation can be applied to the nerve, providing adequate contact density to activate different populations of nerve fascicles consistently. This enables independent control of various motor functions governed by the same nerve.^28–30

The electrode of dynamic materials

For health monitoring, flexible hybrid electronics have been utilized to study electrical signals experiencing large strains.³¹ These electronics consist of a network of conductive flakes and particles embedded in a flexible material matrix, typically thermoplastic polymers.^30–32 Together, they form conductive inks made from metal flakes such as copper, gold, and silver, which act as compliant electrical conduits.^32–34 The dispersion of the conductive flakes and particles within a polymer matrix forms a flexible and conductive percolating network.³⁵

Stretchable electrodes on elastomeric substrates

Elastomers, particularly silicon-based ones such as polydimethylsiloxane, are essential for advancing stretchable electronics in bioelectronics, which offer precision for implantable medical devices and neural recordings. These materials are biocompatible, chemically inert, and suitable for physiological environments.³⁶

3D electrodes

3D bioelectronics is becoming increasingly important as it allows for access to deeper regions of various tissues. Current research on 3D bioelectronic devices involves electrical stimulation and the recording of neural signals from biological subjects, with 3D printing enabling the creation of micro- and nanostructured electrodes.³⁷

Pulse generator

Pulse generators drive continuous stimulation and interventions, triggering data capture at precise intervals before and after the stimulus. These generators contain both the battery and stimulation system. Currently, stand-alone devices use analog dials to set stimulus frequency, leading to inaccurate intervals. While digital timers provide precise intervals, they lack smooth transitions between frequencies. The programmable stimulation, based on computers, offers accuracy and flexibility but is relatively bulky.³⁸

Connectors

Connectors are the implants that connect the electrode to the pulse generators. Short-term implants such as sutures, trachea tubes, lead wires (flexible wires), and sensor probe tips have all been made of stainless steel, type 304 is used as connectors in bioelectronics devices depending on the site of implantation. This completes the circuit and carries the electrical current from the electrode to the pulse generator.³⁹

The components of bioelectronics are shown diagrammatically through the vagus nerve stimulator (VNS) as shown in Figure 3(a).

FIG. 3.

Anatomical location and origin of 12 cranial nerves and 31 pairs of spinal nerves.

Bioelectronics and the Human Nervous System

The PNS acts as the communication bridge between the central nervous system (CNS) and the rest of the body. In recent years, modulation of the PNS has appeared as a potential field of neuroscience that provides novel approaches in the treatment of motor disorders, neurological diseases, and psychiatric disorders.⁴⁰ Neural engineering may offer the possibility of modulating the autonomic nervous system (ANS) activity, influencing the body’s internal conditions, and helping to compensate for malfunctioning regulatory circuits that contribute to illnesses, making it a promising approach for using bioelectronics in therapeutics.⁴¹

The human nervous system consists of the brain, which consists of information-processing cells called neurons that control movements, sensations, emotions, language, and task-evoked responses.⁴² These neurons are widespread in the cerebrum, cerebral cortex, brainstem, hindbrain, spinal cord, and plexus.⁴³ The analysis of the changes in electrical activity resulting from the application of sensory inputs has been the main focus of studies on brain function.⁴⁴ Glial cells such as astrocytes, oligodendrocytes, and microglial cells preserve the ionic medium of neurons, control nerve signals relay center, control absorption of neurotransmitters, offer a scaffold for some elements of neural development, and assist in the healing process following neural injury.^45–47 All these are the primary focus to develop the feedback-regulated bioelectronics.

The spinal cord acts as a major relay center of neuronal signaling and descends inferiorly from the brainstem into the lower back with a bundle of nerve roots known as the cauda equine located in the spinal canal underneath the medullary cone.⁴⁸ The spinal cord’s H-shaped gray matter consists of cell bodies of motor and sensory neurons, interneurons, neuropils, and unmyelinated axons. The white matter consists of interconnected myelinated motor and sensory fiber tracts. White matter ascending tracts carry information from the sensory receptors to higher levels of the CNS, and descending tracts carry information from the CNS to the periphery. The spinal cord gives rise to 31 pairs of spinal nerves that further originate from various plexus systems as shown in Figure 4.^48,49

FIG. 4.

Mechanism of neuronal impulse and synaptic transmission.

Cervical plexus (anterior rami of C₁–C₅)

It supplies nerves to the scalp, neck, chest, and axilla; its sensory fibers provide proprioceptive innervation of the same area through the transverse cervical nerve, the supraclavicular nerve, the lesser occipital nerve, and the great auricular nerve. The cervical plexus’s motor branches facilitate diaphragm innervation and neck movement. Additionally, the anterior rami of the C₃ and C₄ loops combine with outputs from C₅ to innervate the phrenic nerve and regulate respiration.^50,51

Brachial plexus (ventral roots C₅–C₈ and T₁)

Approximately 50 muscles and skin in the pectoral area and upper extremities are innervated by the five nerve roots that combine to form trunks, divisions, cords, and branches. The brachial plexus is the origin of several important mixed nerves, such as the ulnar, radial, axillary, musculocutaneous, and median nerves.^52,53

Thoracic plexus

These are 12 pairs of spinal nerves with cutaneous innervation of the skin, musculoskeletal system, and viscera. Additionally, peripheral and visceral motor fibers innervate the thorax, deep back, abdominal wall, and gut muscles. Pre- and postganglionic sympathetic fibers are also discovered running alongside spinal nerves in the thoracic spine, as this area gives origin to a large portion of the sympathetic trunk.⁵⁴

Lumbosacral plexus

The lumbar and sacral plexuses are frequently referred to as the lumbosacral plexus because of their shared nerve root overlap. The axon innervates the abdomen wall in addition to the lower extremities for all sensory and motor functions.⁵²

Besides these nerves and their plexus, cranial nerves can also be targeted for impulse modulation. Cranial nerves are a set of 12 paired nerves that originate from the backside region of the brain. These send electrical signals between the brain, face, neck, and torso. The details are mentioned in Table 2 and shown in Figure 4.

Table 2.

Theory of Cranial Nerves^42,55,56

Cranial nerve	Point of origin	Comments
Olfactory [S]	Olfactory epithelium	Sense
Optic [S]	Retina	Vision
Oculomotor [M]	Midbrain	Ability to move, blinking of eyes
Trochlear [M]	Midbrain	Ability to move eyes down
Trigeminal [Mi]	Pons	Sensations in face and cheeks, jaw movements
Abducens [M]	Pons	Ability to move eyes
Facial [Mi]	Pons	Facial expressions, taste
Auditory [S]	Cochlea and semicircular canal	Hearing, balance
Glossopharyngeal [Mi]	Medulla	Tongue, pharynx, parotid glands, tonsils, pharynx
Vagus [Mi]	Medulla	Digestion, heart rate, breathing, exocrine pancreas, liver, spleen, kidneys, ureter, blood vessels
Spinal accessory [M]	Medulla	Shoulder and neck muscle movement
Hypoglossal [M]	Medulla	Tongue movement

S, sensory; M, motor; Mi, mixed.

Understanding the nerve conduction patterns in nerve cells is necessary to help both neuronal decoding and the development of closed-loop bioelectronics. The nerve’s function is to conduct information all over the body either through electrical, neural, or electrochemical signaling pathways.^57,58 Nerve impulses are conveyed and transmitted as action potentials, which are said to be transitory variations in membrane potential. Neurons have a resting membrane potential of around −60 mV, which develops as a result of ionic movement across the cell through the Na⁺-K⁺ pump (3Na⁺ out and 2K⁺ in).^59,60 Upon stimulation, the Na⁺ channel permeability changes, making voltage-gated Na⁺ channel allow Na⁺ current to enter the cell, which results in depolarization of the membrane that briefs a millisecond. As the process continues, the membrane potential increases dramatically, which results in the opening of voltage-gated K⁺, allowing K⁺ to flow outside of the neuron, bringing the repolarization.^59,61

In unmyelinated axons, the process is repeated at every single point of the axon membrane through a series of events until the signal reaches the synaptic region. In the myelinated axons, only mature action potentials at small gaps in myelin, called nodes of Ranvier, jump to reach the synaptic region.^61,62 Hasty and precise neuronal communication is achieved through the exceedingly synchronous discharge of signaling molecules and neurotransmitters within milliseconds of the action potential. It is classified as chemical and electrical, depending on the transmitting mechanism as shown in Figure 5.^63,64 Understanding the neuronal function, impulse generation, and action potential in normal and disease states is necessary for developing bioelectronics.

FIG. 5.

(a) Systemic presentation of bioelectronic therapeutic device for the Vagus Nerve Stimulation. (b) FDA approved invasive bioelectronic therapeutic targets and their application.

Spinal cord stimulation

The spinal cord’s neural circuit plasticity and neurochemical characteristics create a solid foundation for the occurrence of chronic pain, making the spinal cord a potential target for treating chronic pain.⁶⁵ Following the FDA’s 1989 approval of spinal cord stimulation (SCS) for the treatment of chronic pain resulting from nerve injury in the arms, legs, and trunk, the idea of using bioelectronics to treat chronic pain gained attraction.⁶⁶ SCS neuromodulation technique acts at the spinal dorsal columns and modulates the pain signals that ascend to the brain. According to the gate control theory proposed by Melzack and Wall, non-nociceptive stimuli are transmitted by myelinated Aβ fibers, while painful stimuli are transmitted by unmyelinated C fibers.⁶⁷

Sacral nerve stimulation for bladder impairment and urinary incontinence

The urinary bladder is controlled by the sympathetic nervous system, which innervates the trigone, while the parasympathetic sacral nerves innervate the detrusor muscle. During a contracting wave, the detrusor muscle pulls on fibers around the vesical neck, opening the outlet and releasing bladder contents. The voiding is controlled by both sympathetic and parasympathetic preganglionic fibers.⁶⁸ Sacral neuromodulation utilizes low-intensity electrical stimulation to restore bladder function in impaired bladders and urinary incontinence. This stimulation is believed to result in spinal circuitry reorganizations, improving bladder control regulation. The preliminary electrical signal likely voyages through the large fibers of the somatic system, but visceral nerve fibers cannot be activated by current intensities.⁶⁹

SNS for constipation and fecal incontinence

The colon, rectum, and anus work together to facilitate the process of defecation.⁷⁰ Preganglionic neurons of the sacral spinal cord scheme through the pelvic nerve and plexus to neurons positioned in the myenteric plexus to control the process of defecation.^71–73 The sacral nerve stimulation (SNS) involves direct, long-term, low-voltage electrical stimulation of the sacral nerve roots through an implanted pulse generator buried in the buttock’s subcutaneous fat, coupled to a quadripolar electrode lead system that is placed percutaneously.⁷⁴ Research has revealed that fecal incontinence increases the frequency of retrograde propagated sequences throughout the colon, while slow-transit constipation increases colonic propagating sequences.⁷⁰

VNS in epilepsy

Epilepsy is a chronic neurological condition that is characterized by an imbalance between brain excitability and inhibition leading to frequent, unprovoked seizures.⁷⁵ The parasympathetic efferent vagus nerve originates in the jugular and nodose ganglia that innervate the nucleus tractus solitarius bilaterally, sending projections to seizure-associated regions. Vagal nerve stimulation uses stimulation of these afferent fibers to exert antiseizure effects on intracranial structures remotely.⁷⁶ The animal models suggest that the activation of the right cervical vagal nerve lowers seizure activity; there is some indication that this also applies to humans.⁷⁷ There are mainly three methods for VNS: left and right cranial VNS and transcutaneous VNS. The most popular technique is the subcutaneous subimplantation in the left upper chest or axillary border. The most common symptoms of left mid-cervical VNS include voice changes, coughing, dyspnea, dysphagia, and paresthesias or discomfort in the neck.⁷⁸ For VNS, the generator is placed beneath the layers of the skin of the chest, and using a lead the generator is connected to the left vagus nerve. The device works by sending mild pulses to the left vagus nerve at regular intervals throughout the day to prevent the occurrence of seizures.⁷⁹

VNS in depression

According to the monoamine hypothesis for depression, decreased neurotransmission and availability of important monoamine neurotransmitters including 5-Hydroxytryptamine, Norepinephrine, and Dopamine cause depression.⁸⁰ The Hypothalamic-Pituitary-Adrenal (HPA) axis central model suggests involvement of cytokines in depression.⁸¹ VNS works in depression by inhibiting proinflammatory cytokines production and increasing peripheral production of anti-inflammatory circulating cytokines. Furthermore, VNS is also associated with alternate Corticotropin-Releasing Hormone secretion, thus preventing overdrive of the HPA axis. It has also been seen that VNS inhibits the peripheral blood formation of Tumor necrosis factor-α (higher in depression patients).⁸²

Deep brain stimulator for Parkinson

The label of deep brain stimulation has been trademarked by Medtronic, Inc., Minneapolis, MN, USA, and the first commercially marketed device was introduced in the mid-1970s.⁸³ Parkinson’s is the damage of striatal dopaminergic neurons. Deep brain stimulator (DBS) is a symptomatic intervention for movement disorders and other neurological and psychiatric conditions.^84,85 Early use of electrostimulation of deep brain structures was for the treatment of chronic pain and psychiatric illness.^86,87 Also, during the early period of stereotactic lesional surgery in Parkinson’s disease, there were some attempts at using DBS instead of making lesions.⁸⁸ In Parkinson’s disease, DBS of the internal globus pallidus and subthalamic nucleus was a safe and effective target. As it is associated with surgical lesioning procedures, chronic DBS is used with the standard stimulation parameter for Parkinson’s disease and causes minimal tissue damage.^89–92 This reversible and adaptable neuromodulation therapy has made innovations in the stimulation of the brain and the application of structural and functional neuroimaging. Notably, tractography of the corticothalamic and corticobasal ganglia connections has been produced using MRI with Deep Tissue Injury, and this information can be used to predict the improved outcome of DBS for neurological and psychiatric problems.^93–97 Currently, the nucleus subthalamic is the chief target for DBS in Parkinson’s disease.^98,99 Evidence suggests that the stimulation of the zona incerta also shows promising results.¹⁰⁰ Typically, subthalamic nucleus DBS should be carried out bilaterally to reduce motor symptoms on both sides and enable the best possible medication to decrease any kind of adverse event.¹⁰¹ Figure 3 shows the systemic organization of the VNS and FDA-approved invasive bioelectronic system activating location.

For sinusitis

Symptoms like dense nasal mucus, tissue swelling, nasal congestion or obstruction, and facial pain and pressure are caused by inflammation of the sinuses and nasal mucosa.^102,103 The nasal dorsum, bilateral supraorbital, and infraorbital nerve areas are the target sites for treating sinus externally (mentioned in Table 3).

Table 3.

Various Invasive and Non-Invasive Bioelectronic Devices

Device	Comments	References
Invasive bioelectronics
SCS
Senza SCS by Nevro Corporation:	FDA approval on January 18, 2022. The device is implantable, rechargeable, and designed to treat chronic, intractable pain in the trunk or limbs. Clinical trials demonstrated significant improvement and sustained outcomes using 10 kHz SCS as in diabetic neuropathy patients. Devices encounter allergies, migration, discomfort, and unpleasant stimulation, and loss of pain alleviation postimplantation.	104–107
Saluda Medical Evoke® SCS System by Saluda Medical Pty Ltd	FDA approval on February 28, 2022. The system can be operated as either open-loop (remote-controlled) or closed-loop stimulation (healthcare professional-controlled).	108
Prodigy, Proclaim, and Proclaim XR SCS Systems by Abbott Medical	FDA approval on February 15, 2023. For treating diabetic peripheral neuropathy through tonic stimulation mode that targets specific nerves.	109
Prospera SCS, by BIOTRONIK NRO, Inc.	FDA approval on March 31, 2023.	110,111
The devices listed above comprised an implanted signal generator linked to one or more implanted leads and an external remote controller that sends radio signals to the impulse generator, directing the device’s stimulation.
SNS
eCoin® Peripheral Neurostimulator, by Valencia Technologies Corporation	FDA approval on March 1, 2022. The device is implanted under the skin near the ankle and controlled by a healthcare professional via a remote control. The eCoin® system stimulates the bladder controlling the tibial nerve. The device generates electrical pulses to stimulate bladder control in patients with urgency urinary incontinence.	112
InterStim XTM system by Medtronic	FDA approval was announced on February 22, 2022. Recharge-free neurostimulator for bowel and bladder control. It is used to treat urinary retention, overactive bladder, and chronic fecal incontinence symptoms.	113,114
Axonics R20TM by Axonics, Inc.	FDA approval on January 31, 2023. A miniaturized, rechargeable device. A tiny implant that gently stimulates the nerves controlling the bladder and bowel. Help to restore regular connections between the brain and the bladder and bowel.	115,116
VNS
Product series by LivaNova	In 1997, U.S. FDA approved an implantable VNS device intended for the treatment of refractory epilepsy which was followed by a series of focused clinical trials. Between 1997 and 2020, the company developed a series of devices that were FDA-approved FDA for commercial use, like the NPH 100 in 1997 as the first approved product and the current SenTive Duo M1000-D* in 2020. Later, in 2005, the FDA approved the use of the same device to treat treatment-resistant chronic depression.	78,117,118
SenTiva M 1000 VNS Therapy System by LivaNova	FDA approved in 2017. Comprises a modified VNS Therapy Programming System with a wireless wand and an implanted pulse generator.	115
NEMOS	Transcutaneous VNS (t-VNS) is used to treat epilepsy. Approved in 2010 by the European Union for epilepsy and depression.	78
DBS
Vercise™ PC, Vercise Gevia™, and Vercise Genus™ DBS systems of Boston Scientific Corporation	FDA approved in 2021. For Parkinsonian tremor. These implantable devices employ amplitude, pulse width, and frequency to deliver low-intensity electrical pulses to brain nerve centers.	119
Noninvasive bioelectronic
ClearUPTM Sinus Pain Relief, by Tivic Health Systems, Inc. in Menlo Park, CA.	FDA approved the device to treat allergic rhinitis-related sinus pain and is classified as a U.S.-FDA Class II and EU Class IIa medical device. Further, its use extends to allergies, the flu, and the common cold on March 24, 2021. The device stimulator produces microcurrent through its conductive tip and emits biphasic current at low frequency. The stimulation electrode is moved in a limited, repeated “H” pattern above and below the bilateral orbits for 5 min, spanning the nasal dorsum and the bilateral supraorbital and infraorbital nerve areas. The apparatus detects treatment spots, locations with a density of subcutaneous nerve fibers, and the ophthalmic and maxillary branches of the trigeminal nerve employing a proprietary algorithm.	120–122
Combined technology
Monarch external trigeminal nerve stimulation (Monarch® eTNS®) technology by NeuroSigma	U.S. FDA granted pre-market approval. Areas of the brain linked to ADHD are stimulated with low-level electrical pulses. The device improves the ADHD-RS and Clinical Global Impression scores as EEG results show.	123–127
Innosphere, a wearable hat device (AF-RNS®)	The product is undergoing clinical trials in Europe for ADHD with a special machine-learning ability.	128
Some other patented bioelectronics
Device	Indication	References
Programmable automatic implantable cardioverter/defibrillator and pacemaker system (Invasive)	For bradycardia support and/or tachycardia support.	129
Merci Retriever (Invasive)	Treatment of Ischemic shocks	130,131
Fetal Shunt (Implantable/Invasive)	Treating the fetus with a lower urinary tract	132
Impella RP (Invasive)	Assist systemic circulation for the right ventricle	133
S-ICD pulse generator (Invasive)	For patients with a high risk of cardiac shock	134
Micra transcatheter (Leadless Invasive)	Cardiac pacemaker	135
Renerve (Invasive)	Nerve regenerator	136
Prestige LP Cervical disc (Invasive)	For invertebrate mobility	137
Jarvik 2000 (Invasive)	For patients who present with end-stage heart failure	138
Heartmate (Invasive)	Provide hemodynamic support in heart failure patients	139
Magnetic Nose Clip (Noninvasive)	Anti-snoring	140

SCS, spinal cord stimulation; SNS, sacral nerve stimulation; VNS, vagus nerve stimulator; DBS, deep brain stimulator; ADHD, attention deficit hyperactivity disorder.

For opioid withdrawal symptoms

When a person gets dependent on opioids, brusquely cuts back on, or quits using them, it is known as opioid withdrawal.^141–143 Based on encouraging findings from retrospective research, the FDA approved a bioelectronic device, a percutaneous electrical nerve field stimulator for opioid withdrawal symptoms in 2018.¹⁴⁴ Transcutaneous auricular neurostimulation is cleared by the FDA under 510(k), which targets the auricular branch of the vagus nerve and the auriculotemporal nerve to treat opioid withdrawal symptoms.^145–147

For attention deficit hyperactivity disorder

The neuropsychiatric condition known as attention deficit hyperactivity disorder (ADHD) is highly heritable and associated with several genetic variables that impact the neurological and molecular systems.^148,149 Advances in knowledge of the molecular causes of ADHD have opened up new avenues for neurostimulation-based technology therapies using invasive DBS microelectrodes, noninvasive transcranial electric stimulation, transcranial magnetic stimulation, transcranial direct current stimulation, and transcranial alternating current stimulation.¹⁵⁰

The trigeminal nerve affects the anterior cingulate cortex and inferior frontal gyrus, which are linked to ADHD, while it is unclear if trigeminal nerve stimulation activates specific brain regions known to be related to ADHD or not.^151–153 But vector stimulation that transports pulses to the trigeminal nerve by using either subcutaneously implanted electrodes with an implantable pulse generator, or external electrodes with an on-body pulse generator is found to be useful in ADHD.¹⁵⁴ The gadget is made up of an electrode array accessory to activate the trigeminal nerve.¹⁵⁵ It uses radiofrequency energy to function, is easy to use, and is rechargeable. When the patient is asleep, the gadget is worn on the forehead and treatment is administered.¹⁵⁶

Overcoming the Obstacles in the Development of Bioelectronics

The field of bioelectronic medicine is at a pivotal moment in its development. To have a large, widespread impact in the real world, the area will need to overcome the major obstacles it faces today (shown in Table 4).¹⁷

Table 4.

Challenges in Developing Bioelectronic Therapeutics

Problem	Comments	References
Design-related	Miniature size, biocompatibility. Improvements in sensing capabilities and precise neural targeting. Major pitfalls in physiological compatibility, adaptability, feedback system, long-term stability, and reliability.	157–160
Technological glitches	➣The scale and material of production, flexibility, stability, and robustness of the material. ➣Electrodes and sensors shelf life, efficiency, reliability, size and number, and recording quality. ➣Improvement in ML and adaptive algorithms. ➣Power source: reliability, wireless, multiscale signal processing. ➣Low power consumption. ➣Additional circuit to rectify the generated AC output voltage.
Scientific	Neural coding, neural, and non-neural biomarkers.
Regulatory	Regulatory processes and clinical trial design.	12
Clinical acceptance	Determining the exact dose in terms of the intensity of electric current to be given for modulating the neural impulse. Precision in the identification of the correct nerve fiber for their implantation.
Others	Invasiveness, translation into medical practice, cost, investment, packaging, and standardization.

ML, machine learning.

Design explanation of the key to design

Nanotechnology materials can be a great approach to achieve a less invasive technique. The introduction of MXenes with superior mechanical and electrical properties helped to reduce the invasive techniques.^161–163 The implanted bioelectronic devices requiring an invasive procedure can be resolved by using a wireless implantable bioelectronic device.¹⁶⁴ Though challenges like efficiency, biocompatibility,^165,166 energy transmission, power source availability, positioning of external communication devices, associated advanced electronic components, feedback loops, and software still need to be addressed with appropriate optimization.^{165,167–169} For providing noninvasive cerebral stimulation, dry electrode technology, developed at Battelle Memorial Institute, uses an electrode material made of mixed ionic electronic conducting materials and is light weight, elastic, and flexible. Carbon nanotubes as the electronic conductor and hyaluronic acid as the ionic conductor scattered throughout a polymer matrix can also be used for the same.¹⁴³ Besides this, 3D printing has aided the design, as discussed in the “Structural Featuring of Bioelectronics” section.

The challenges continue in the footings of material modernization, device fabrication, and circuit design architecture. A single, specialized sensor that can sense and track every biological function is difficult. Multifunctional materials such as graphene, carbon, phthalocyanine, and copper may create lab-on-a-chip devices that can sense biochemical and physical parameters in reaction to environmental changes. To meet energy requirements, effective nanomaterials such as thermoelectric, photoelectric, piezoelectric, and triboelectric materials need to be thoroughly investigated.^170,171 The device’s physics needs to be studied before application.^172,173 To enable possible benefits, the fundamental challenges of safety, reliability, and long-term use must be overcome.^172,174,175

Decoding neuronal functioning and neural biomarkers

The functioning of bioelectronics solely depends on neural circuits and neural signaling. The scientific community has done intense research to decode the neuronal systems, identifying different fibers in nerve functioning and signaling to promote the selective effect of the stimulation.¹⁷⁶ The main goal of the research is to interpret neuronal activity to derive meaningful information regarding motor, sensory, or even disease states in the brain. Finding the characteristics of the signals, such as multiunit activity, a correlation between pairs of signals being watched, and firing rate in action potentials, is the next level of the brain decoding process. Machine learning (ML) techniques can be applied to identify patterns in the data and convert them into a format that can be associated with motor and sensory events.¹⁷⁷ For this neural biomarkers came to light and further BIOS Health is working to resolve the neural decoding problems through neural boundaries using closed-loop approaches that permit the isolation of neural biomarkers from continuous neural recordings for discovering and decoding neural biomarkers and is also working on wearable biosensors.^178–180

Neural markers can be suggested as biomarkers related to neural tissue activities that provide specific neural signals generated under normal, stress, or disease states. Hence, a closed-loop neuromodulation system that can modify brain and neuronal activity in real time and restore the natural state using stimulation that can be achieved through these neural biomarkers^181,182:

• (1)ENG method that interprets the biological signals’ rhythmic pattern and exogenous stimulus.¹⁷⁶
• (2)ECG studies act as a neural biomarker to study the baroreceptor activity and selective vagal stimulation for reducing blood pressure without inducing bradycardia and bradypnea.¹⁸³ Studying ECGs can enhance early detection of atrial fibrillation, hence prevention of stroke as established by AliveCor.¹⁸⁴ Moreover, initiatives such as the Apple Heart Study and Verily’s Project Baseline extended the research and screening programs.^17,185,186
• (3)The signal study of the vagus nerve extracts the neural profile that resembles the temporal evolution of the respiratory cycle, blood pressure, or both during baseline activity. Hence, systolic peaks P₁ and P₂, the dicrotic wave of the blood pressure signal, inflation, and the deflation phases of the respiratory cycle act as neural biomarkers for the same.^187–189
• (4)Mechanoreceptive information that measures the increase in vesicular pressure acts as a neural biomarker in pelvic disorders as concluded using a rat model with cuff electrodes placed in the pelvic nerve of rats.¹⁹⁰
• (5)Spectra of CNS activity in response to metabolic stimuli were found to be neural biomarkers of type II diabetes and glycemia as concluded using hook electrodes in rats.¹⁹¹
• (6)Signals generated during the modulation of bladder volume and mechanical stimuli in the perineal region are suggested as neural markers for the urinary tract. The conclusion is based on a cat-animal model, bladder control, and Lower Urinary Tract control studies utilizing high-density penetrating electrode arrays implanted in the Diagnosis Related Group (S₁–S₂). The Non-Linear Auto Regressive Moving Average modality provided the best bladder pressure prediction.^192–194
• (7)Spike patterns of ANS murine vagus nerve that originate due to the activation of specific cytokines from different types of nerve fiber are found to be a neural biomarker for hypo- and hyperglycemia-specific neural activity.¹⁹⁵ The study developed a decoding framework for cytokine-specific vagus nerve action based on firing rate activity.^196,197

Moving ahead of neuromodulation, an essential attribute is that a single bundle of nerve fibers consists of different pathways and functions. Neuronal activity to external variables and stimulation of the right fiber to achieve the desired result can be achieved by discriminative knowledge about the fiber. As explained by the undesired activation of laryngeal muscular fiber, resulting in difficulty breathing while stimulating the cervical vagal nerve, this can be improved by selective stimulation using neural biomarkers.¹⁹⁸ Studies suggest that extraneural electrodes can be used to analyze neural population activity of compound neural activity potentials recorded in the nerve. As a result, fibers in nerve bundles can be distinguished from one another based on their unique velocity conductance. Several stimulation parameters as well as the relationship between a given stimulation dose, neural recruitment, and physiological effects can be investigated with the help of evoked activity in terms of compound neural activity potentials, though the technique fails in the differentiation between fibers with conduction velocities.¹⁹⁹

Optimizing stimulation parameters and electrode design during vagal stimulation in dogs and pigs found that using VNS-elicited compound neural activity potentials has a strong correlation between stimulus-evoked compound neural activity potentials and the physiological effects of stimulation. By assessing physiological responses directly, this correlation might be utilized to infer neural fiber recruitment in human and animal models.^200,201 A new framework consisting of anatomical data from the histological examination of vagus nerve sections with hybrid neural models and recording electrodes has discriminating characteristics. A new measure known as the discriminative field potential has been developed, within this framework, to distinguish between various spatial configurations of fascicular neuronal activity.²⁰² By using this method, effective discrimination between afferent and efferent nerve fibers can be performed. Based on the above procedure it is possible to prevent side effects by optimizing VNS in anesthetized rats.²⁰³

Safe packaging of bioelectronics against electromagnetic interference

Metals have shown promising possibilities for bioelectronics as shielding and packaging materials to prevent abnormal circumstances caused by interference. The primary advantage of metallic shielding is its capacity to deflect unwanted or disruptive electromagnetic forces. However, the biological interactions between the biological tissues surrounding the metallically insulated device and the device itself may harm the shielding.^204,205 Therefore, packaging materials must be both electromagnetically compatible and biocompatible. Despite metallic shielding being immune to electromagnetic interference, bioceramics are being explored by researchers and biomedical engineers due to their excellent hermeticity. However, their application is limited by challenges related to processing at room temperature and the functional characteristics that depend on specific frequencies.²⁰⁶ Hermetically compatible materials may prove to be good candidates for these implantable gadgets as they demonstrate corrosion resistance and biocompatibility.^207,208

Regulatory guidelines for bioelectronics

Bioelectronics are classified as medical devices, related to various regulatory bodies and their associated affairs.

U.S. FDA

The FDA regulates medical devices through the Center for device and radiological health to safeguard and promote public health to ensure their safety, effectiveness, and quality. There exist three fundamental stages to obtaining FDA marketing approval for medical devices that is the Premarket approval process, the Premarket notification process, and the humanitarian device exemption process.²⁰⁹ To give an overview of the approval layout for devices, the device classification, appropriate premarket submission pathway, and Premarket Approval requirements of the device should be provided.²¹⁰

European Medical Agency

Three EC directives have been subjected to periodic amendments to address the approval of medical devices. Implantable devices are governed by Directive 90/385/EC and in vitro diagnostic devices by Directive 98/79/EC. For marketing, medical devices must bear the Conformité Européenne label to be marketed in any EU member state. Nonimplantable medical equipment and low risk are “self-marked” while high-risk gadgets require more thorough external scrutiny. In 2010, new regulations increased standards for device clearance based on similarities to past “predicate” devices by adding “proactive postmarket surveillance” and reporting the events to Eudamed. The database is currently accessible to the EC only.²¹⁰ The European Medical Agency informs organizations within the EU that are responsible for assessing the conformity of products before their launch in the market. Manufacturers can select any legally recognized body to perform the conformity assessment process.²¹¹

Therapeutic Goods Administration

The Australian Therapeutic Goods Administration (TGA) Approval Process for Medical Devices involves manufacturers determining their product’s classification using Schedule 2 of the Australian Therapeutic Goods (Medical Devices) Regulations. Devices are classified into Class I, Class IIa, Class IIb, and Class III. A TGA sponsor facilitates device registration and approval in Australia through a current technical file, design dossier, and declaration of conformity for submission. The sponsor submits the Manufacturer’s Evidence and Global Medical Device Nomenclature codes for devices except Class I. The sponsor applies online, along with the fee. The TGA approves or rejects the application. If approved, a certificate is provided and a public summary is available on the Australian Register of Therapeutic Good database.²¹²

Central Drugs Standard Control Organization

The Central Drugs Standard Control Organization is the National Regulatory Authority (NRA) in India, responsible for approving manufacturing, import, clinical trials, standards, sale, and distribution of medical devices. The NRA ensures public health and safety by conducting the Materiovigilance Programme of India (MvPI) with Indian Pharmacopoeia Commission and Sree Chitra Tirunal Institute for Medical Sciences and Technology. The MvPI aims at systematic safety data collection, enabling regulatory decisions and recommendations on the safe use of medical devices in India. It monitors medical device-associated adverse events, creates awareness among health care professionals about MDAE reporting in India, and monitors the benefit–risk profile of medical devices. The Central licenses approving authority (CLAA), the Drugs Controller General of India, is responsible for enforcing these rules in matters relating to the import of all classes of medical devices.²¹³ The CLAA classifies the device, sets safety standards, appoints bodies for conformity assessment, conducts postmarket surveillance, and issues warnings and recalls. The CLAA uses the Bureau of Indian Statistics and the International Organization for Standardization specifications for quality management systems.²¹⁴

Current Research Areas

The growing field of bioelectronics is blooming a lot these days with various ongoing research as shown in Figure 6.

FIG. 6.

Current research scenario in the field of bioelectronics.

Artificial retina for blindness

The first artificial retina implant to repair injured retinal cells was authorized by the FDA. The implant helped those with retinitis pigmentosa to possibly identify things, detect movement, recognize enormous letters, and enhance their orientation and mobility.²¹⁵ The implant does not relate to a closed-loop system and uses external stimuli to correct vision disorders.²¹⁶ In June 2015, Argus II was released commercially to improve visual function.²¹⁷

Retinal implant

Boston group was the first to unveil a retinal implant of the first generation. Later, to get around issues with the coil design, hermetic metallic sealing, and an advanced superior electrode array design, a second-generation device was introduced by the Boston group.^218–220 The gadget is composed of a hermetic brain stimulator and an external transmitter. Artificial vision is created by stimulating retinal tissue with impulsive electrical impulses that are generated by the implant’s processing mechanism using the camera installed on the spectacles. An electronic chip is positioned behind the conjunctiva, the components are outside the eye and affixed to the sclera, and an electrode array is implanted in the subretinal space to provide stimulus current to the retinal nerve.^221–223 To keep the gadget under intelligent, intellectual, and biological atmospheric circumstances, the team employed metallic isolation.^221,222,224

An on-chip inductive coil-based device as optical retinal prostheses made of vertical silicon nanowire photodiodes of iridium oxide electrode uses light as an energy source to produce current, inserted in the subretinal space. The study suggests that the system can generate a variety of stimulating currents proportional to the incident light’s wavelengths.²²⁵ A similar system with epiretinal (electrodes contact with ganglion cells) and subretinal (electrodes touch with photoreceptors) implantation describes that they use light energy to stimulate the retinal neurons using an integrated silicon nanowire-based photodetection circuit.²²⁶ To deliver stimulative signals to the retinal neurons at constant field intensity and currents, a novel electrode system of silicon nanowire array for retinal prosthetic devices could be thought.²²⁷ A new photovoltaic array-based subretinal implant could be developed to stimulate retinal neurons locally.²²⁸ Ongoing research for subretinal stimulation and optic nerve stimulation is effective. Subretinal entails implanting an electrode array between bipolar cells and the retinal pigment epithelium.^229,230 Implanted devices near the optic nerve and retina focus on recovering vision through closed-loop stimulatory responses.^231–233 These systems do not fully restore vision to natural eyesight as it would require millions of electrodes. Studies say that stimulating sensory neurons with brain implants may significantly improve patients’ vision, hence combining these two concepts would be a revolutionary option.²¹⁶ An organic electrolytic photocapacitor that consists of p-n semiconducting organic nanocrystals charged by photoillumination energy and a thin tri-layer metal was presented to stimulate subretinal neurons. The gadget is nontoxic, wireless, biologically friendly, and devoid of external prejudice.²³⁴

The team at Bionic Vision Australia conducted a human trial for a novel suprachoroidal retinal prosthesis. The device was made of biocompatible titanium, platinum, a titanium-copper-nickel alloy, and a silicon elastomer eliminating the requirement for any implanted devices by using a percutaneous connector as a direct electrical link, enabling flexible neurostimulation and electrode monitoring.²³⁵ By using copper as a sacrificial layer and the modified poly(3,4-ethylenedioxythiophene) polystyrene sulfonate in an electrochemical gelation process, a hydrogel-electrode system was created on a platinum substrate that offers a great deal of promise for the creation of flexible, functional electrodes.²³⁶ Novel approaches by combining photon energy-based stimulating sources with energy storage devices modeled after photocapacitors can help blind patients.²³⁷

Baroreflex activation treatment for hypertension

Up to a certain point, hypotensive medication therapy is effective; however, a considerable proportion of patients across the country have blood pressure readings above recommended levels while receiving treatment.²³⁸ The sympathetic nervous system serves as a useful homeostatic mechanism to adjust hemodynamics. Regretfully, in certain individuals, this process eludes the carotid sinus’s physiological regulation and, via a variety of pathways, results in resistant hypertension.^239,240

The Barostim Neo system, CVRx Inc., which is made up of a pulse generator and a carotid sinus lead, is the system used to administer baroreflex activation therapy.²⁴¹ The lead ends in a platinum-iridium disk electrode with a circular backer and an iridium oxide coating centered on the backer.^242,243 By creating a subcutaneous infraclavicular chest wall pocket to house the pulse generator, the implant is done similarly to a pacemaker.^244,245 A transverse cervical incision is made above the carotid bifurcation to expose the carotid sinus in preparation for electrode insertion^246,247 and a subcutaneous tube is used to deliver the lead’s opposite end to the pulse generator pocket, where it is connected to the device.^248,249 Sensitivity is determined by monitoring hemodynamic alterations like drops in heart rate and blood pressure.^250,251 When there are no adverse effects, including sharp drops in blood pressure or heart rate, therapy is started at a modest dose.^189,252,253

On August 16, 2019, the U.S. FDA authorized the Barostim Neo system to treat advanced heart failure symptoms. The device sends information to the brain by stimulating baroreceptor cells in the neck with electrical impulses. Data from the clinical study proved that the device increases the life quality and the distance patients could walk in 6-minute walking tests. The device’s implantation may result in infection, low blood pressure, nerve damage, allergic reactions, arterial damage, worsening heart failure, stroke, and even death. Patients with certain nervous system disorders, atherosclerosis, ulcerative plaques, and those with known allergies to silicone or titanium should not use the device.^254,255

VNS for rheumatoid arthritis

Patients who had a VNS to trigger the inflammatory response experienced a notable improvement in clinical signs and symptoms of rheumatoid arthritis (RA).^256,257 Vagus nerve signals are transduced across the nerve, resulting in the reduction of inflammatory cell activation,^258,259 and the generation of fewer mediators of systemic inflammation.^260–263 Moreover, lower circulating immune cell activation resulted in decreased joint injury, inflammation, and discomfort.^264,265

The U.S.-FDA approved an Investigational Device Exemption on December 11, 2017, allowing SetPoint Medical, a clinical-stage biomedical technology, to start a pilot trial for patients with drug-refractory RA in the United States. Device is surgically positioned on the vagus nerve and turned on by a present dosage schedule to evaluate the safety and efficacy of VNS.^266,267

Future Prospective

Bioelectronic therapeutics has the potential to bring a significant change in the health care sector through their closed-loop working system that can precisely detect and modulate the electrical signal or nerve impulse patterns in the nervous system.²⁶⁸ The closed-loop device system hints at the future of personalized bioelectric therapeutics²⁶⁹ as systems that can decode the collected data into interpretable biomarkers can potentially estimate the stimulation threshold that varies from patient to patient and condition to condition as shown in Figure 7.²⁷⁰ Various blogs, research, and reviews suggest that bioelectronic medicine may be used in conjunction with therapeutic approaches to bring about advancements for diseases for which there are presently no treatments and for which the available treatments are either not very successful or have serious adverse effects. According to the market survey, the approximate value of the global market for active biomedical implants is predicted to be 178.9 billion USD in 2020, and the market is expected to increase at a compound annual growth rate of 8.5% between 2021 and 2027.²⁷¹

FIG. 7.

Future of personalized closed-loop bioelectronic therapeutics where brain, neural, organ, organ system, and full-body state information are analyzed both for longer term and in real time to adjust the stimulation parameters for the best possible therapy results by interpreting both brain and non-neural biomarkers of a patient’s condition. This method advances the use of bioelectronic therapies for the comprehensive, individualized treatment of chronic disorders.

Bioelectronics have the potential to target Gastroesophageal Reflux Disease, bleeding and hemophilia, cancer, chronic pain, Chronic Obstructive Pulmonary Disease, Congestive Heart Failure, Crohn’s disease, diabetes, epilepsy, hearing disorders, hypertension, irritable bowel syndrome, lupus, mental, neurodegenerative, and psychiatric disorders, RA, sepsis, stroke, and spinal cord injury.²⁷² The use of VNS for treatment-resistant anxiety disorders, Alzheimer’s disease, chronic refractory headaches, obesity, and rapid cycling bipolar disorder has been shown in small open-label studies and case series reports; however, the FDA has not approved any of these indications.²⁷³

The 5th BEM Summit, October 2022, discussed the future of bioelectronic therapeutics that launched the concept of intersubject biological variations to the growth of bioelectronics. They emphasized the significance of comprehending the brain or molecular systems that mediate certain physiological activities as well as the scientific foundations of bioelectronic treatments.²⁷⁴ The summit concluded that the several facets of brain disease treatment using bioelectronic treatments elucidate the process of bidirectional communication between a medical device and the neurological system. Knowledge and activation of the maternal immune system is the future therapeutic research that will focus on the molecular and neurological underpinnings of persistent itching associated with Janus kinase I pathway.²⁷⁴

Artificial intelligence in bioelectronics

Artificial intelligence (AI) is the simulation of human behavior about neural processes and activity. The AI stimulation includes markings like neural activity, impulse generation, origin, direction of propagation, intensity, and any kind of change in them.²⁷⁵ Bioelectronics, when integrated with AI, brings new openings for value-added diagnosis, prediction, and more operative, personalized treatments. These devices are picking up an increasing variety of biological signals, such as ECG and electroencephalogram, which can provide the foundation for AI to improve and educate health care. Here are a few instances of them:

• Tens of thousands of ECG pairs were utilized by the Mayo Clinic’s cardiovascular AI team to train a neural network to recognize heart failure patients. The network was able to identify the likelihood of future cardiac issues.²⁷⁶

• AI algorithms developed by Google Deep Mind and the Moorfields Eye Hospital in the United Kingdom have demonstrated the ability to diagnose more than 50 eye illnesses with high accuracy.²⁷⁷

• AI facilitates the creation of biosensors as part of the AI–biosensor integrated system because biosensors can evaluate large datasets at an incredibly fast rate, which helps with pattern recognition, illness trajectory prediction, and customized therapy recommendation. Biosensor precision and accuracy are enhanced by AI.^278,279

• Various single-molecule bioelectronic sensing has supported the understanding of biomolecules. Such biosensing revealed deep insights into fundamental biological processes with the help of nanogapped electrodes and nanopores. AI algorithms effectively support the biosensors to extract meaningful features, detect subtle changes, improve signal-to-noise ratios, and uncover hidden patterns in massive data. This potentially helped in treating the diseases that are under the control of biomolecules like interleukins and cytokines, that is, inflammatory disorders.²⁸⁰

• AI-enabled biosensors helped to improve interpretation by eliminating drift, noise, and the need for feedback or control signals. In addition to providing intelligent health monitoring and producing high-precision forecasts and judgments that enable the operation of closed-loop systems, the development of AI has produced strong tools and algorithms for data processing and analysis. A number of wearable biosensors are being developed in conjunction with AI to monitor the respiration rate, heart rate, pulse, perspiration, and tears; implanted sensors for cardiovascular care, nerve signal collection, and neurotransmitter monitoring; and soft wearable electronics for precise treatment. AI-enhanced multimode sensors, AI-enhanced self-sustaining systems, and human–machine interfaces are also supported in closed-loop bioelectronics.²⁸¹

ML—the future of “closing the loop” of bioelectronics

ML is a facet of AI that can automate the process to provide structure to the analytical models. Based on the data received by the devices, ML allows them to adapt independently to new scenarios, enabling the software to successfully predict and react to the scenarios based on the data they are receiving. Integration of bioelectronics with ML helps these devices predict the activity of the neural system, directing the device to act as per the requirement. Change to ML has deep learning as a subset that deals with neural networks through multiple deep layers allowing ML to learn. These neural networks consider algorithms inspired by the human brain and neural networks throughout the body known as the artificial neural networks (ANN). Further, the ANN learn from a large number of neural activities both under normal as well as altered, that is, diseased state. Deep learning techniques help the devices to analyze complex neural signals, impulse propagation, origin, and any kind of change in them. Thus, ML helps in the development of closed-loop bioelectronics.^282,283

Limbic function monitoring^284,285 and distributed multimodal data streams gathered from various sensors, such as molecular sensing^286,287 and peripheral nerve activity recording,¹⁹⁷ are two methods that can be used to close the loop. However, real-time closed-loop brain stimulation requires sophisticated statistical methods that can combine sufficiently complicated data streams, spot patterns in the data, and determine when the patient’s physiology is changing. Future closed-loop bioelectronic medications that employ ML can also parse sensor data input to identify a disease occurrence and initiate a neutral stimulation. Currently, brain–machine interfaces are one of the biological applications of ML. To decipher complex physiological data, several strong ML techniques are available, such as deep neural networks, support vector machines, and support vector regressions.^288–290 This will make the decoding method quick, accurate, selective, and easily retrainable over time, as a tailored decoding algorithm trained on a subject’s physiological data has enabled gains in therapeutic efficacy.^291,292

Conclusions

Technological advancements, along with advances in anatomical and neurological studies, are enabling the health care sector to transcend conventional treatments for the treatment of various chronic diseases through the use of bioelectronic devices. Bioelectronic devices help in monitoring and understanding the disease at the neural signaling level. Besides this, the devices treat the disease at the target through the modulation of neural signals without generating any secondary pharmacological response as occurs in the case of synthetic moieties.

The implantation and functioning of bioelectronics require a precise biological, anatomical, and neural understanding. For the precise functioning of bioelectronic devices, both neural and non-neural biomarkers are required. These biomarkers support the functioning of a closed-loop device, allowing them to modulate the nerve impulse at the right time and neural center. The FDA has approved numerous bioelectronic devices targeting different neural centers of the body that can modulate the impulse activity for treating the respective neural center-controlled/associated disorders. Bioelectronic devices require thorough research for structuring and designing them, as these devices are highly invasive and complicated to manage. Through the application of nanoarchitectures and 3D electrodes, noninvasive or less invasive devices for bioelectronics could be developed. An integrated system of AI and ML for neural decoding and biomarkers will support the development of precise and smart closed-loop devices.

Author Disclosure Statement

I Ishu Garg on behalf of all the authors declare no conflict of interest.

Funding Information

No funding was received for this article.