During the course of a normal day, there are numerous physical tasks to beaccomplished that require control of one’s arms and legs. Forindividuals with a neurological disease, motor disorder, or physical injury,the ability to control their arms or legs may be impaired or completelyremoved. One rehabilitative method to restore some function of movementwould be to use a bioelectronic interface to provide a communication pathwaybetween the nervous system and either the control of an external device orimproved control of the existing limb. Successful bioelectronic interfacesrequire efficient and robust communication between external hardware and thenervous system in order to drive human movements. An ultimate solution wouldbe a device that could “speak” the same language as the neuralnetwork it is communicating with to produce biofidelic functionality. In thefollowing chapters (Chapters 24–26), a description will be given ofhow devices such as retinal prostheses perform the task of converting visualscenes into neural stimulation patterns to restore a sense of vision.Subsequent chapters (Chapters 27–34) will cover the differenttechnologies available for using neural recordings for brain–machineinterfaces (BMI). Again, this requires decoding neural signals from thecortex to properly control objects such as a computer mouse. This chapterwill describe how neural recordings can be decoded to provide controlsignals for producing human movements in either the arms or the legs.Depending on the application, this may involve reanimating the existinglimbs through methods of functional electrical stimulation (FES) orproviding control signals derived from neural recordings to activateprosthetic limbs [1–12]. This chapter will begin with a briefdiscussion of the general flow process involved in designing an effectivebioelectronic interface for artificial movements. The next two sections willdiscuss examples of bioelectronic interfaces for upper arm movements (reachand grasp) and lower leg movements (locomotion).

Introduction

Mark Weiser in 1991 set the vision for ubiquitous computing: “The mostprofound technologies are those that disappear. They weave themselves intothe fabric of everyday life until they are indistinguishable fromit.”[1]. Ubiquitous computing in his vision is embedded in the worldaround us or over our body via sensor networks accessed through intelligentinterfaces: “Its highest ideal is to make a computer so embedded, sofitting, so natural, that we use it without even thinking aboutit.”[2]. He meant to reposition the center of our world by movingtowards human-centered computing, where technology assists us adapting tohuman needs and preferences while remaining in the background, silent untilrequired. Weiser predicted three ages of computing technology: the mainframeage when many people shared a computer; the personal computer age when oneperson has one computer; and the ubiquitous computing age when each personshares many computers via a rich interconnect fabric. The consequence ofthis evolution of computing is a change in how people can interact withtechnology: a more natural way of using the power of networked sensing andcomputing systems. In the ubiquitous computing world, people are connectednot just to the Internet or other computers, but to places, other people,even to everyday objects and things. Indeed the frontier of research is verymuch related to this vision: Internet of Things, Systems of Systems, andSwarms are all different facets of this encompassing vision, albeit eachtheme emphasizes one particular aspect.

The increasing maturity, performance, and miniaturization of processors,networking technologies, memory, displays, and sensors is enabling a movetowards pervasive computing, ubiquitous connectivity, and more adaptableinterfaces that are sensitive and responsive.

Introduction

Analytical platforms are used in the life sciences for the observation,identification, and characterization of various biological systems. Theseplatforms serve applications such as sequencing of deoxyribonucleic acid(DNA), immunoassays, and gene expression analyses for environmental,medical, forensics, and biohazard detection [1]–[3]. Biosensors are asubset of such platforms that can convey biological parameters in terms ofelectrical signals. Biosensors are utilized to measure the quantity ofvarious biological analytes and are often required to be capable ofspecifically detecting multiple analytes simultaneously. A goal in biosensorresearch is to develop portable, hand-held devices for point-of-care (POC)use, for example in a physician’s office, an ambulance, or at ahospital bedside, that could provide time-critical information about apatient on the spot [4].

The current demand for high-throughput, point-of-care bio-recognition hasintroduced new technical challenges for biosensor design and implementation.Conventional biological tests are highly repetitive, labor-intensive, andrequire a large sample volume [2], [5]. The associated biochemical protocolsoften require hours or days to perform at a cost of hundreds of dollars pertest. Instrumentation for performing such testing today is bulky, expensive,and requires considerable power consumption. Problems remain in detectingand quantifying low levels of biological compounds reliably, conveniently,safely, and quickly. Solving these problems will require the development ofnew techniques and sensors.

Neural recording

Neural recordings have been utilized for controlling brain–machineinterfaces, such as artificial arms, since the late 1960s [1]. Neuralrecording devices play a central role in paralysis prosthetics, stroke,Parkinson’s disease, prosthetics for blindness, and experimentalneuroscience systems. Several attempts have been conducted in recent yearsto implement large-scale multi-electrode neural recording [2], [3].Experiments have been conducted on a variety of species ranging from rats[4] and monkeys [5] to humans [6].

Figure 33.1 shows how a brain–machine interface (BMI) can help severalpatients. Figure 33.1(a) shows conceptually how BMIs can help patients whosuffer from damaged arm muscles, preventing the signal from reaching thebrain, or those who have artificial limbs [7]. Vivid examples of recentefforts in using BMIs to help paralyzed patients include the thought controlof a wheelchair [8], as seen in Figure 33.1(b), and thought control of arobotic arm by a patient suffering from paralysis for 15 years [9], as shownin Figure 33.1(c). However, better quality of life could be imparted tothese patients if a fully implantable wireless solution was available.

Figure 33.2 shows an illustrative picture of a brain implant system. Theimplant’s power is wirelessly harvested to allow battery-free operation.Wireless transmission is necessary to allow for non-invasive operation.

Conceptually, extending the premise of bioelectronic interfaces down to thescale of single molecules is a straightforward goal. In practice, thechallenges are purely technological. Single-molecule bioelectronic deviceswould have to involve features much smaller than state-of-the-artsemiconductor electronics, and successful design would have uniquerequirements for sensitivity and stability.

These imposing specifications are balanced by the potential of enormousrewards, because single-molecule bioelectronics would be a breakthroughtechnology for biochemical research and applications. By peering past theensemble behaviors of traditional characterization, single-moleculetechniques aim to directly observe the stochastic fluctuations,instantaneous dynamics, and non-equilibrium behaviors that make up amolecule’s full functionality. Moreover, single-molecule measurementscan uncover the unusual reaction trajectories of a genetically mutatedprotein or a receptor interacting with pharmacological inhibitors. Buildinga better understanding of the precise roles of proteins in complexbiological processes is a grand challenge for biology, biochemistry, andbiophysics in the twenty-first century.

These potential benefits have spurred the development of a variety ofsingle-molecule techniques. Single-molecule fluorescence, specificallyFörster resonance energy transfer (FRET), has become a standard tool forsingle-molecule biochemistry [1]. Meanwhile, single-molecule bioelectronicshas remained elusive, despite the wide-ranging capabilities of modern solidstate electronics.

In the first demonstration of human brain-to-brain control, a scientistwearing an electrical brain-signal reading cap triggered motion in hiscolleague across campus [1]. While research on brain–machineinterfacing has been going on for years, this recent demonstrationrepresents an impressive signal of how advanced technology has become. Usingelectrical brain recordings and a form of magnetic stimulation, oneresearcher sent a brain signal to another on the other side of theUniversity of Washington campus, causing the recipient’s finger tomove on a keyboard. Similarly, other researchers have demonstratedbrain-to-brain communication between two rats, or between a human and arat.

The technologies used by the researchers for recording and stimulating thebrain are both well known. Electroencephalography (EEG) is routinely used byclinicians and researchers to record brain activity non-invasively from thescalp and is discussed in the first chapter of this part of the book.Transcranial magnetic stimulation, on the other hand, is a non-invasive wayof delivering stimulation to the brain to elicit a response. In theexperiment mentioned above, the stimulating magnetic coil was placeddirectly over the brain region that controls a person’s right hand.By activating these neurons, the stimulation “convinced” thebrain that it needed to move the right hand.

The rapid evolution of micro-nanoelectronics has led to unprecedentedopportunities to implement ultra-miniaturized electronic systems anddirectly interface them with biological systems. Thanks to the hugepossibilities in terms of design space, nanoscale microelectronic circuitscan allow the implementation of incredibly compact electronic systems withvery complex functionalities, including sensing, actuating, processing, andcommunicating, which could be potentially employed in countless newopportunities to realize “smart” (i.e. able to perform complexfunctions) human–machine interfaces in both directions, from human tomachine (including also humanoid robots) and vice versa, and so closing theinformation loop in order to allow bidirectional interactions. However, theencumbrance of the contact-based interfaces could greatly limit theexploitation of the emerging plethora of potential opportunities to buildsmart human–machine interfaces. To overcome these limitations, thereis growing interest in autonomous electronic systems that, in general, couldbe expected to be contactless, self- or remotely powered,ultra-miniaturized, non- or minimally invasive with negligible side effects,biocompatible, eco-friendly (i.e. green), low-cost, and so on, that we couldrefer to as zero-power [1] and more generally aszero-impact electronic systems, sometimes also referredto as “smart dust” [2]. These future and emerging technologiesare an extremely active macro-area of research [3] that, in spite ofenormous recent progress, is still at the early stage. In this generalcontext of grand challenges, the current silicon-based microelectronictechnologies can provide a huge range of opportunities for deliveringpotential solutions that could respond, at least in part, to the wishfulthinking for the most effective solutions of the future.

Introduction

Age-related macular degeneration (AMD) is one of the leading causes ofblindness in the developed world, with an incidence of 1:500 in patientsaged 55–64, and 1:8 in patients over 85 [1]. Retinitis pigmentosa(RP) is an inherited disease blinding about 1 in every 4000 individuals muchearlier in life [2]. In both of these conditions the photoreceptor layerdegenerates, while the inner retinal neurons survive to a large extent[3–5]. Electrically activating these neurons provides an alternativeroute for visual information and raises hope for the restoration of sight tothe blind.

In a normal retina, photoreceptors convert light into neural signals that areprocessed by inner retinal neurons, leading to generation of actionpotentials in the retinal ganglion cells (RGCs). These signals travel to thebrain through the optic nerve and serve as the basis for visual perception.Electrical stimulation of the retina with microelectrodes can also produceaction potentials in RGCs, creating spatially patterned percepts of lightcalled phosphenes. Indeed, recent clinical trials with retinal prostheticelectrode arrays have restored visual acuity to subjects blinded by retinaldegeneration up to 20/1200 using epiretinal placement (facing theganglion cell side) [6], and up to 20/550 with subretinalimplantation [7]. While this serves as an important proof of concept withclinically useful implications, existing retinal prosthesis designs have anumber of shortcomings.

Electrocardiography (ECG) is the acquisition of electrical activity of theheart captured over time by an external electrode attached to the skin. Eachof the cell membranes that form the outer covering of the heart cell has anassociated charge which is depolarized during every heartbeat. These appearas tiny electrical signals on the skin which can be detected and amplifiedby an electrical circuit.

History of ECG

Alexander Muirhead is reported to have recorded a patient’s heartbeatinitially in 1872. In the early twentieth century, Willem Einthoven used astring galvanometer to measure the electrical activity of the heart. Inthose days, patients had to immerse their limbs into salt solutions fromwhich the ECG was recorded. Nowadays, the ECG machine can be carried aroundin a patient’s pocket and has become truly mobile. Einthoven assignedthe letters P, Q, R, S, T to various deflections in the wave and identifiedfeatures of a number of disorders based on the deflections. He was awardedthe Nobel Prize in medicine for his discovery. Even though many advanceshave been made in the field of ECG, the basic principles are still valideven today.

Sample waveforms of ECG

The ECG signal is characterized by peaks and troughs as shown in Figure 28.1.For each beat of the heart, a similar sequence is generated owing tocontraction and relaxation of various muscles of the heart. The ECG signalis plotted on graph paper as shown in Figure 28.2.

Introduction

“How does the brain work?” This question, which has been askedthroughout the history of mankind, is addressed by all branches of science,in particular life sciences, from different perspectives. Although all seeka different answer, the common feature that triggers the research isobservation. This, together with curiosity, is whatmakes the beginning of a scientific study. From the electrical engineeringperspective, observations are performed by recording the electrical signalsgenerated by the neurons and interpreting the results. These interpretationsguide the research of scientists who are trying to map the brain, or tounderstand the mechanisms behind neurological disorders, or to implementbrain–machine interfaces [1]. Methods for recording the neuralsignals have evolved to the current state over decades, and the evolutionstill goes on. This chapter introduces the main concepts of thenew-generation neural recording systems: implantable wireless neuralrecording systems with a case study on in vivo epilepsymonitoring.

Current clinical practice in recording electrical activities of the brain isdominated by electroencephalography (EEG) which is a non-invasive procedureperformed along the scalp. Another type of EEG, intracranial EEG (iEEG; alsoknown as electrocorticography, ECoG), is an invasive procedure which isperformed by placing an electrode matrix (or array) onto the cortexfollowing the craniotomy as presented in Figure 32.1 [2]. Intracranial EEGis employed for epileptic focus localization prior to the resective surgery[3] which is performed to treat certain types of epilepsy. New-generationneural recording systems [4] aim to alter two main features of conventionaliEEG: (1) macro-sized iEEG electrodes will be replaced with microelectrodearrays (MEA) fabricated with microtechnology, and (2) transcutaneous wirescarrying neural information will be eliminated thanks to wireless datacommunication.

Although the term was first proposed in 1968, the birth ofbioelectronics as we know it today is much more recent. Infact, the first meaning of the word bioelectronics was the intermolecularelectron transfer found in biological systems [1], while its modern meaning isquite different, as discussed in this chapter.

Tony Turner, founder and Editor-in-Chief of the Elsevier journalBiosensors and Bioelectronics, wrote in 2005:“Bioelectronics is a recently coined term for a field of research thatworks to establish a synergy between electronics and biology” [2]. Overthe years, his journal became the main forum in the field of bioelectronics. Thejournal originally appeared in 1985 with the simple name ofBiosensors, and the title was expanded to include the termBioelectronics in 1992 [2]. The expressed scope of thejournal [3] explains that a key aspect of bioelectronics is the interfacebetween biological materials and electronics.

To better understand the modern state of the field, we can also consider themeaning of “advanced bioelectronics” as mentioned in April 2013 by the NationalInstitute of Standards and Technology (NIST – an agency of the US Department ofCommerce) and represented by the example of electronic DNA sequencing, assupported by singlemolecule mass spectrometry to develop innovative electronicdevices for healthcare [4]. The authors [4] refer to devices in which anelectric field drives individual molecules of single-stranded DNA through ananometer-scale pore (Figure 1.1). That approach is a step toward applicationsfor rapid sequencing of DNA [5] and DNA transcription complexes [6], and couldlead to further developments in protein sequencing [7] too. The application isclearly focused on devices for healthcare and, more generally, includes bothdiagnostic and therapeutic tools.

The concept of biomimetic systems was introduced in the early definition ofbioelectronics. As we have already seen in the first chapter, the originaldefinition of bioelectronics set by Wolfgang Göpel includes“structures [that] may consist… of chemicallysynthesized units such as molecules, supramolecules andbiologically active (biomimetic) recognition centers” [1]. Over theyears, the concept has been expanded in order to move from simplerecognition systems to biomimetic membranes for voltage shifts ingraphene-based transistors [2], systems for cell separation in the blood[3], electronic noses [4, 5], electronic tongues [6], smart info-chemicalcommunication systems [7], electronic design [8], pancreatic beta-cells [9],and neurons [10].

Artificial brain architectures, with all the neurons fully interconnected inparallel, show issues in terms of scalability, especially because the numberof interconnections scales exponentially with the number of neurons [11],while it would be desirable for it to scale like biologically plausiblearchitectures [11]. This brings us to the concept of bio-inspired orbiomimetic systems as possible solutions to solve problems emerging inextremely complex bioelectronics architectures. Over the years, severalbio-inspired and neuromorphic architectures have been proposed in theliterature for silicon neurons [10], synaptic and neural components made ofNiTi [12, 13], sensors [14], orientation tuning devices [15], and patternrecognition systems [16]. In the direction of more complexity andfunctionality, the present state-of-the-art in the field proposes artificialsystems for pancreas [9], skin [17, 18], cognitive architectures, and brains[19, 20].

Introduction: A survey of proteins in nanobioelectronics

The idea of using proteins to assemble hybrid electronic devices stems frommolecular electronics [1] and, as such, it is intimately connected with theadvent of nanosciences and nanotechnologies, dating back to the early 1990s.Since then, much technological effort but less scientific effort has beendeployed to try to implement devices that take advantage of the peculiarfeatures of proteins.

A technologist’s standpoint is that of regarding proteins asself-contained, nanometer-sized functional units, highly specialized andefficient in performing a certain functional task. Their efficiency istraced back to the fact that, being active parts of living beings, proteinsare taking advantage of billions of years of natural evolution inspecializing towards a given activity.

This way of thinking, which one can often encounter in ritten or spokenaccounts, appears to be questionable in light of a rather less naïveunderstanding of the theory of evolution, but is perhaps a good enoughstarting point to understand the historical motivations which led to theremarkable interest of a interdisciplinary part of the scientific communityin the use of proteins for assembling electronic devices.

The other aspect motivating the interest in proteins as elements inelectronic circuits is their size.

Introduction

The field of electrophysiology explores the mechanisms of electrical signalgeneration and propagation in living tissues. Contemporary electrophysiologyhas tended to focus on electrically excitable cells, for example observingaction potentials propagating along a neuronal membrane. However, the birthof electrophysiology was concerned with a more global approach, looking at“bioelectricity” in the whole organism. Galvani, the father of“bioelectricity”, hooked up lightning rods to cut nerves in afrog’s leg and observed twitching of the leg muscles during alightning storm. Matteucci, in 1831, was the first to measure the so-called“injury potentials” using a galvanometer in a cut nerve endingand demonstrated the existence of action potentials in nerves and muscles.His work was extended by Du Bois-Reymond, who in 1843 was able to directlymeasure the propagation of action potentials and also the injury potentialsfrom cuts in his own finger. While rapid changes in membrane conductance ofindividual cells may be viewed as a somewhat “obvious” targetthat has been intensely studied, other, slower bioelectric phenomena such asthose seen in wound healing have been somewhat neglected. The field of“bioelectricity” has seen a re-emergence in the past decade,thanks in part to new techniques in molecular physiology. In addition, theensemble of electric phenomena in biology is rarely considered. Instead ofsolely focusing on action potentials in individual cells, it can beinstructive to consider electrical characteristics in groups or arrays ofcells. In this chapter we focus on the electrical measurement of arrays ofcells or tissue layers. We attempt to widen the traditional definition ofelectrophysiology in a more general sense, as the electrical measurement ofion flow in biological systems. We review a subset of literature on methodsfor measuring ion flow in tissues in vitro in bothelectrically active and non-electrically active cells. We will particularlyhighlight dynamic methods for monitoring cell cultures, and the new trend ofusing transistors rather than simple electrodes with a special emphasis onthe use of conducting polymers to do so.

Introduction

Lack of physical activity (PA) and exercise is a widespread and prevalentproblem in the modern society. A study conducted by the US Department ofHealth in 2002 showed that the lack of PA is associated with a wide range ofconditions including obesity, diabetes, heart disease, stroke, andosteoporosis [1]. In 2008, the World Health Organization reported that 11%of the world’s population (at 25+ years of age) was estimated to besuffering from diabetes [2], and diabetes care alone may account for up to15% of national healthcare budgets [2]. These statistics reflects thenegative economic impact to healthcare systems worldwide.

Physical activity intensity (PAI) and energy expenditure (PAEE) can beestimated from measurements of oxygen consumption using portable gasanalyser systems. However, these indirect calorimeters are complex, bulky,heavy, obtrusive, and very expensive, and hence unsuitable for routine use.On the other hand, there is a large body of evidence (discussed later inthis chapter) suggesting that PA activity intensity and energy expenditurecan be estimated and continuously monitored using physiologicaland/or biomechanical information, owing to the relationship of theseparameters with oxygen consumption during aerobic exercise [3–5]. Inaddition, recent advances in microelectronics have enabled the developmentof miniaturized integrated circuits, medical sensors and micro-engineeredinertial sensors.