This paper presents a multi-frequency time-variance-based Electrical Impedance Tomography (EIT) ASIC, designed to balance high-speed performance with low power consumption for biomedical applications. The ASIC leverages a dual-mode impedance readout architecture, incorporating time-to-digital demodulation at high frequencies and digital I-Q demodulation at low frequencies, which together address limitations in conventional EIT systems regarding data acquisition speed and power efficiency. The proposed system includes a current generator with an integrated auto-calibration scheme and a flexible clock generator with multiple PLLs, supporting a wide frequency range from 15.6 kHz to 1 MHz. Fabricated in a 65 nm CMOS process, the ASIC achieves a total power consumption of 2.43 mW and enables detailed multi-frequency impedance analysis while maintaining compactness and cost-effectiveness. In-vitro measurements with water tank validate its efficacy in different frequency-dependent impedance changes.

A heartbeat-powered harvester for cardiac monitoring will be demonstrated. The device harnesses mechanical vibrations from heartbeats and converts them into electrical energy to power pressure sensors, enabling real-time monitoring of heart’s health without an external powering source. The system integrates biocompatible piezoelectric flexible materials with sensors utilizing power electronics. Live demonstrations illustrate the system's feasibility for continuous intracardiac monitoring. Initial results suggest promising performance in terms of energy harvesting efficiency and biosignal acquisition accuracy for various medical and wearable devices applications.

This work presents a high-resolution plantar pressure measurement system designed to capture detailed foot information for gait analysis. After training, the customized regression model can estimate lower body joints positions in real-time using foot pressure data. The demonstration will feature a UK size 8 insole, equipped with 253 resistive pressure sensors wirelessly connected to the server. Visitors will have the opportunity to experience walking with the insole, observing the plantar force distribution displayed and lower body joints as predicted by the model on the computer.

Gait analysis is a crucial method for evaluating and monitoring an individual's health. A critical aspect of this analysis is understanding how forces are distributed across the foot while walking. Existing plantar pressure insole systems often lack the resolution needed for detailed foot analysis. To address this, a real-time insole system is presented with 253 high-density resistive pressure sensors (4 sensors per cm2) for each foot with a wireless transfer rate of 60 Hz. In addition, our work combines the insole hardware with a custom convolutional neural network (CNNs) and long short-term memory (LSTM) model to predict six lower body joint landmark positions. The prediction achieves a coefficient of determination (R2) of 0.83 and a mean squared error (MSE) ranging from 7.0e-4 to 9.6e-4. With an inference time of 0.6 ms, this system provided accurate, high-resolution plantar foot pressures and insights into 3D joint movements in the lower body. It is a promising tool for applications in rehabilitation and sports performance optimisation.

Capacitive sensing is commonly employed to measure a wide range of physical quantities, including displacement, humidity, and pressure. Readout accuracy, speed, and power consumption are among the key performance characteristics during the design process for meeting the specific demands of the sensor application. Minimum parasitic effects in capacitance-to-time converters offer high suitability in precise, high-resolution applications that demand low sensitivity to environmental effects, such as in biomedical sensing devices. However, the design tradeoffs in the measurement range, resolution, and speed often limit device performance and require manual adjustment of the sensing front-end. This paper describes a low-power readout circuit for high-resolution capacitance sensing across a wide input range of up to 200 pF. A pulse-width-modulation-based, capacitance-to-time (C-T) converter is presented in 65-nm CMOS technology for miniaturised applications. Real-time monitoring of the output enables automatic adjustment of the sensing parameters that determine range, resolution, and speed, providing a flexible solution to the application needs. The system provides rapid conversion of single- and differential-mode capacitors with a 50-μs readout time while consuming 25 μW from a 1-V supply.

Rehabilitation following total knee replacement (TKR) principally aims to reduce pain and restore knee muscle strength and function, which necessitates objective monitoring. Given the growing interest in wearable technologies for rehabilitation, this paper investigates the feasibility of using electrical impedance tomography (EIT) to monitor and estimate muscle activity, specifically focusing on resultant knee angle (KA) and muscle strength (MS). A data collection system was designed to simultaneously capture EIT data, kinematic and physiological reference data, namely KA and MS, followed by two experiments. Data processing methods, such as region of interest (ROI) selection and modelling of the correlation between EIT and reference data, were used to assess feasibility. The results indicate that the long short-term memory (LSTM) model achieved an R2 value of 0.97, suggesting a strong correlation between the EIT data and KA. While EIT demonstrated potential in detecting KA, it did not show effectiveness in capturing changes in muscle activity and MS during isometric contractions.

This work presents a novel interaction method using bioimpedance (Bio-Z) to enhance virtual reality (VR) experience. The platform consists of two electrode bands (with 4 and 5 electrodes), a high-performance bioimpedance measurement system, a recurrent neural network deployed on a laptop and a VR headset. The system captures the Bio-Z features from muscle contraction and bone movement in the wrist and upper arm. After training, the regression model can predict hand grasp angle and forearm motion in both horizontal and vertical directions. This system can be used as a VR human-computer interaction (HCI) peripheral. Visitors will experience VR games using this demonstration platform.

Central venous catheterisation (CVC) involves the precise and challenging task of needle insertion, which can be difficult and prone to complications. This paper presents a bioimpedance (BIOZ) system for precision needle navigation in CVC. The system integrates BIOZ with a haptic feedback mechanism to assist clinicians during the catheterisation procedure. Four types of embedded needle probes were designed for high-sensitivity BIOZ sensing. The probe integrates with a compact electronic handle system for real-time BIOZ measurement and data communication, enabling real-time tissue classification and haptic feedback for optimal needle positioning. In 24 experimental trials with user participation using a phantom model, the system demonstrated a 100% success rate in venous entry, with a root mean square error (RMSE) of 0.66 mm and an 87% probability of achieving a position within ±1 mm of the vein centre. The compact and portable design ensures compatibility with clinical settings, laying a foundation for future improvements in robotic-assisted CVC and procedures alike.

This work introduces a unique bioimpedance (BIOZ) guided haptic needle technology designed to enhance central venous catheterisation (CVC) procedures. The portable system combines real-time BIOZ analysis, needle posture tracking, and a haptic device specifically developed for CVC research. By providing real-time haptic feedback and BIOZ analysis, the system significantly improves precision, ensuring a 100% success rate for vein entry. During the demonstration, visitors will first experience the challenges of CVC without assistance. Then, using the platform, they will experience how the system improves accuracy by guiding the needle into a phantom of the vein, greatly increasing the insertion success rate.

This paper presents a scalable, differential resonant pulser for wearable/portable ultrasound applications, such as hand gesture recognition for prosthesis control and gaming. Presently, the mainstream pulser topologies include the class-D amplifier, multilevel pulse shaping, charge-recycling, and LC-resonant. A resonant pulser topology is adopted because theoretical analysis has shown that the resonant pulser has an inherent power advantage over other types of pulsers. The current state-of-the-art resonant pulser, while achieving excellent fCV2power dissipation reduction, lacks scalability. This work aims to tackle this shortcoming by using high-voltage DMOS transistors in a more area-efficient manner. On average, the proposed circuit uses three DMOS transistors per single-ended transducer or six DMOS transistors per differential transducer, making the proposed circuit attractive for multi-channel systems. Post-layout simulation results show that the proposed differential resonant pulser can achieve 69% fCV2dynamic power reduction, which is comparable to the state-of-the-art. To the best of the authors’ knowledge, the proposed differential resonant pulser for ultrasound applications is the first of its kind.

Ingestible capsules with position tracking capabilities bring a significant advancement in the diagnosis and treatment of gastrointestinal disorders. However, the wearability of these capsule positioning systems is restricted by the weight and shape of the transmitter (Tx). To enhance the wearability and flexibility of the system, a novel positioning system using particle filter (PF) is proposed. By employing PF, the system can calculate the capsule’s position in 3D space using only a belt-shaped Tx coil, paired with a single Rx coil and an inertial measurement unit (IMU). PF iteratively updates the estimated position based on the capsule movement and changes in mutual inductance (MI). The performance of PF with different noises is analysed. Simulation results show that the system has a 22.5 × 11.5 × 27 cm3field of view (FOV), a root mean square error (RMSE) of 13.89 mm under noise.

Multimodal machine learning with missing modalities is an increasingly relevant challenge arising in various applications such as healthcare. This paper extends the current research into missing modalities to the low-data regime, i.e., a downstream task has both missing modalities and limited sample size issues. This problem setting is particularly challenging and also practical as it is often expensive to get full-modality data and sufficient annotated training samples. We propose to use retrieval-augmented in-context learning to address these two crucial issues by unleashing the potential of a transformer’s in-context learning ability. Diverging from existing methods, which primarily belong to the parametric paradigm and often require sufficient training samples, our work exploits the value of the available full-modality data, offering a novel perspective on resolving the challenge. The proposed data-dependent framework exhibits a higher degree of sample efficiency and is empirically demonstrated to enhance the classification model’s performance on both full- and missing-modality data in the low-data regime across various multimodal learning tasks. When only 1% of the training data are available, our proposed ICL-CA method outperforms the best baseline by 5.9%, 5.9%, 5.3% and 10.8% on four datasets across various missing states. Notably, our method also reduces the performance gap between full-modality and missing-modality data compared with the baseline. Code is available 1 1 GitHub repository . . • We address data scarcity in missing-modality tasks, highlighting the limitations of existing parametric approaches, where adequate sample sizes are essential. Our study reveals that focusing adaptively on both missing- and full-modality data improves performance. • We introduce a novel, data-dependent in-context learning method to enhance sample efficiency and optimize learning across missing- and full-modality data. • Experiments on four datasets (medical and vision-language tasks) show that our method boosts performance by 5.9%, 5.9%, 5.3% and 10.8% over the best baseline in the low-data regime.

This paper presents an optimization methodology to enhance the linearity of direct digital synthesis (DDS)-based on-chip sinusoidal current generators (SCGs) for bioimpedance measurements. After a detailed analysis of the origin of harmonics and their impact on measurement accuracy in conventional DDS-based SCGs, three different harmonic cancellation schemes are proposed. These are designed to decrease the harmonics in various applications with differing hardware requirements, significantly improving the precision and efficiency of DDS-based SCGs. The approach was implemented in a DDS-based SCG in 65 nm CMOS technology. Both simulations and measurements demonstrate a marked improvement in linearity with a figure-of-merit at least 9.4 times better than prior work. Although tailored for bioimpedance applications, the approach is also suitable for other low power, high linearity sinusoidal signal generators.

The current generator is one of the most essential parts in electrical impedance tomography (EIT) systems. State-of-art current generators suffer from the trade-off between current efficiency, linearity, output impedance and well-controlled output current level. To address these challenges, a current DAC based current generator with a fourth-order current-mode filter is proposed. The current generator was designed and simulated in a 65 nm CMOS process. By utilizing the proposed current-mode filter and the master and slave current DAC, the current generator achieves a well-controlled output current leveling with 255 tunning steps, 83.9% current efficiency, 0.11% THD at 1 mAppoutput current and 2 MΩ output impedance at 500 kHz.

This paper presents a differential current driver based on a current feedback structure, designed to drive hydrogel sensors. It achieves low power consumption, low common-mode signal on the load, and high current efficiency. The use of a negative unit gain buffer reduces the common-mode signal on the load arising from process variations. The current driver was designed in a 65-nm CMOS technology with a 3.3 V supply. Simulation results demonstrate a THD of 0.4% at 125 kHz, for 40 μAp-poutput current. The common-mode voltage on the load is reduced by 98.96% compared to a conventional topology using two independent drivers. The total current consumption is 60.3 μA, resulting in a current efficiency of 82.9%. The simulated output impedance is 2.76 MΩ at 125 kHz and 1.47 MΩ at 300 kHz.

This paper presents a differential T/R SPDT switch, designed in a 0.18 µm HV BCD technology for PMUTbased biomedical ultrasound systems. It incorporates a bootstrapping technique to pass high-voltage pulses and a shunt branch for improved isolation. The differential T/R switch is designed to interface with bimorph electrodes although a singleended version can also interface with conventional PZT/CMUT transducers. Post-layout simulation results show that the switch circuit exhibits 64 Ω on-resistance and -62 dB off-isolation. A figure-of-merit is proposed to compare ultrasound T/R switches. To the best of the authors’ knowledge, this differential switch is the first of its kind being reported for PMUT biomedical ultrasound systems.

Wireless power transfer with inductively coupled coils is widely used for powering implantable devices. For a continuously moving receiver coil (Rx), real-time detection of its position is important for the magnetic field distribution control. This paper proposes an Rx positioning system that uses three varying-pitch coils and applies a trilateration algorithm for positioning. It operates at a frequency of 125 kHz. The detection area is approximately 224 cm2, when using a 2 cm radius Rx. Experimental results show the position detection system has a 1 cm2resolution with an average deviation of 6.8 mm.

This work presents a wearable, eight-channel A-mode ultrasound system for hand gesture recognition and interactive gaming. The wearable system consists of a custom-built forearm bracelet with eight piezoelectric transducers (1 MHz) evenly distributed, an electronic hub housed on a bicep strap, a laptop with a bespoke machine learning algorithm and a robotic hand for actuation. The electronic hub drives the transducers to produce ultrasound pulses that will travel into the forearm muscles. The reflected ultrasound signals will vary based on the forearm muscular morphology (movement-dependent). By decoding the reflected ultrasound signals using machine learning techniques, a robotic hand can be controlled based on the user’s hand motions. In contrast to the conventional surface electromyography methods, the visitor can experience seamless control of a robotic hand and a maze navigation game via ultrasound as a novel sensing modality.

Amyotrophic Lateral Sclerosis (ALS) has been a grossly misrepresented end user group when developing coadaptive algorithms for Brain Computer Interfaces (BCI). Researchers have credited this issue to the difficulty of progressing disease in patients with ALS. This non-stationarity reduces accuracy over time. This paper introduces an online model, usable for a BCI using ALS patients data. The automatic coadaptive model effectively decodes 3 class motor imagery (MI) of the left, right hand and rest while adapting to address non-stationarities of Electroencephalography (EEG) over time caused by various factors over the study duration. Adapting Filter bank Common Spatial Pattern (FBCSP) algorithm, where we show it could enable above 70% detection of hand MI in ALS end users longitudinally, previously lacking evidence. The evaluation results demonstrate that the model achieves average accuracies of 72.6% over a 1–2 month period of usage involving 8 ALS patients. This work shows the first auto-adaptive model with ALS patient EEG data providing a stronger incentive for further investigation by setting benchmark models on longitudinal datasets contributing to the solution of multiple challenges in this field.

An improved bootstrapped circuit topology is proposed for the design of a high-voltage, differential, single-pole double-throw transmit/receive switch for an ultrasound hand gesture recognition system. The differential transmit/receive switch is designed in a$0.18~\mu $m HV BCD technology. It is intended to interface with bimorph piezoelectric micromachined ultrasonic transducer electrodes, although a single-ended version can also interface with conventional ultrasound transducers. A comprehensive analysis has produced useful insights, namely the dependence of off-isolation on high-voltage input slew rate, and the dependence of common-mode rejection ratio on threshold mismatch and timing misalignment. In addition, an extended model has been developed to predict harmonic distortion through a DMOS transistor. Measured results verify the effectiveness of the design guidelines and switch operation. The switch circuit exhibits$67~\Omega $on-resistance and -63 dB off-isolation while occupying a modest area of$289~\mu $m$\times 295~\mu $m. The switch achieved a figure-of-merit with 74% improvement over the state-of-the-art. To the best of the authors’ knowledge, this differential switch is the first of its kind reported for piezoelectric micromachined ultrasonic transducer biomedical ultrasound systems.

This paper presents a fully-implantable closed-loop opto-electro stimulation interface for motor neuron disease studies, designed for experiments with freely moving rodents. A low power consumption Bluetooth data link is used to wirelessly control 64 opto-electro stimulation channels and receive neural recording data. The implant is powered by a wirelessly rechargeable lithium-ion battery, which can support 2.5 hours continuous operation with a stimulation output up to 10 mA. The battery is recharged using a QI standard wireless inductive power link, which can deliver >100mW power at a distance of 2 cm. The total size of the implant system is 29 mm × 20 mm × 13 mm. Its performance is compared with the state-of-the-art.

The purpose of lung recruitment is to improve and optimize the air exchange flow in the lungs by adjusting the respiratory settings during mechanical ventilation. Electrical impedance tomography (EIT) is a monitoring tool that permits measurement of regional pulmonary filling characteristics or filling index (FI) during ventilation. The conventional EIT system has limitations which compromise the accuracy of the FI. This paper proposes a novel and automated methodology for accurate FI estimation based on EIT images of recruitable regional collapse and hyperdistension during incremental positive end-expiratory pressure. It identifies details of the airway tree (AT) to generate a correction factor to the FIs providing an accurate measurement. Multi-scale image enhancement followed by identification of the AT skeleton with a robust and self-exploratory tracing algorithm is used to automatically estimate the FI. AT tracing was validated using phantom data on a ground-truth lung. Based on generated phantom EIT images, including an established reference, the proposed method results in more accurate FI estimation of 65% in all quadrants compared with the current state-of-the-art. Measured regional filling characteristics were also examined by comparing regional and global impedance variations in clinically recorded data from ten different subjects. Clinical tests on filling characteristics based on extraction of the AT from the resolution enhanced EIT images indicated a more accurate result compared with the standard EIT images.

This paper presents a stimulator for a cardiac neuroprosthesis aiming to restore the parasympathetic control after heart transplantation. The stimulator is based on time-to- current conversion, instead of the conventional current mode digital-to-analog converter (DAC) that drives the output current mirrors. It uses a DAC based on capacitor charging to drive a power efficient voltage-to-current converter for output. The stimulator uses 1.8 V for system operation and 10 V for stimulation. The total power consumption is Istlmx 10 V + 18.4 μW during the biphasic current output, with a maximum Istimof 512 μA The stimulator was designed in CMOS 0.18 μm technology and post-layout simulations are presented.

A real-time discrete wavelet transform-based adaptive voice activity detector and sub-band selection for feature extraction are proposed for noise classification, which can be used in a speech processing pipeline. The voice activity detection and sub-band selection rely on wavelet energy features and the feature extraction process involves the extraction of mel- frequency cepstral coefficients from selected wavelet sub-bands and mean absolute values of all sub-bands. The method combined with a feedforward neural network with two hidden layers could be added to speech enhancement systems and deployed in hearing devices such as cochlear implants. In comparison to the conventional short-time Fourier transform- based technique, it has higher Fi scores and classification accuracies (with a mean of 0.916 and 90.1%, respectively) across five different noise types (babble, factory, pink, Volvo (car) and white noise), a significantly smaller feature set with 21 features, reduced memory requirement, faster training convergence and about half the computational cost.

High-density neural electrodes in microchannel interfaces require in-situ amplification of the neural signals and rejection of high-voltage stimulus pulses leaking to the channel in order to adequately detect neural signals in the presence of concurrent stimulation. This paper presents the design of an active microchannel neural interface in 0.18 μm CMOS employing neural recording and stimulation. To reduce stimulus artifacts, a novel method is proposed that disconnects the recording module during concurrent channel stimulation and automatically applies detection and reduction of stimulus artifacts from adjacent channels using a tunable filter. Simulations show that the method provides at least 54 dB artifact attenuation.

This paper presents a prototype integrated bidirectional stimulator ASIC capable of mixed opto-electro stimulation and electrophysiological signal recording. The development is part of the research into a fully implantable device for treating motor neurone disease using optogenetics and stem cell technology. The ASIC consists of 4 stimulator units, each featuring 16-channel optical and electrical stimulation using arbitrary current waveforms with an amplitude up to 16 mA and a frequency from 1.5 Hz to 50 kHz, and a recording front-end with a programmable bandwidth of 1 Hz to 4 kHz, and a programmable amplifier gain up to 74 dB. The ASIC was implemented in a 0.18μm CMOS technology. Simulated performance in stimulation and recording is presented.

This paper presents an antenna-based RF wireless power and data transfer system for wireless neonate lung function monitoring in an intensive care unit using electrical impedance tomography (EIT). The proposed dual-band system comprises a directional slotted patch transmit antenna and a slotted flexible omnidirectional monopole antenna operating at 2.51 GHz and 5.1 GHz. A slotting technique unique to each antenna is proposed to achieve dual-band operation and size miniaturization for each antenna. Also, the arraying technique combined with parasitic elements is employed to increase the receive antenna gain and in turn the received power level. Both transmit and receive antennas were simulated and fabricated. Following FCC safety regulations, measurements show 20 mW received power when the receiver and the transmitter are spaced 19 cm apart.

This paper describes a simple and reproducible methodology towards a universal figure-of-merit (FoM) for evaluating the performance of electrical impedance tomography (EIT) systems using reconstructed images. Based on objective full-referencing and signal-to-noise ratio, the method provides a visually distinguishable hot-map and two new FoM factors, to address the issues where common electrical parameters are not directly related to the quality of EIT images. The paper describes the method with simulation results and develops a 16 electrode EIT system using an ASIC front-end for evaluation using the proposed method. The measured results show both visually and in terms of the proposed FoM factors, the impact on recorded EIT images with different current injection amplitudes.

A wearable multi-sensory platform for closed-loop optical stimulation control will be demonstrated. The system is developed for restoring hand grasp movement to patients suffering from muscle paralysis, for example, amyotrophic lateral sclerosis (ALS), by optical stimulating engrafted stem cell-derived motor neurons that reinnervate the extensor and flexor muscles. Movements induced by the stimulation will be monitored by using electrical impedance tomography (EIT) to detect the impedance variation caused by muscle contraction, and using inertial measurement units (IMUs) to measure the finger movements. Learning algorithms are employed for pattern recognition and the results are used for closed-loop stimulation control. During the live demonstration, visitors can control a multi-LED stimulator via a number of different hand movements with a wearable EIT system on their forearm and IMUs on their fingers.

The medial olivocochlear neuron feedback loop in the human auditory system plays a critical role in speech-in-noise perception. Recent studies have successfully simulated the mechanism of the neuron feedback and developed corresponding speech enhancement algorithms for hearing applications. These algorithms apply time-varying attenuation to hearing applications by reducing the gain in each frequency band over time. The speed of gain reduction is characterized by a time constant. In hearing applications, it is known that speech intelligibility is related to the gain regulating time constant. However, the effect of the time constant on enhancement performance is unclear. This paper studies its effect on speech intelligibility in hearing aids by incorporating a time constant variable speech enhancement algorithm modified from an existing neuron feedback model into a hearing aid model. The study demonstrates intelligibility improvement of over 30% at SNR between 5 and 10 dB. The longer time constants (≥ 1000 ms) show greater intelligibility improvement (about 20%) at SNR ≤ 10 dB, whilst the shorter time constants show more benefits at SNR > 10 dB.

Internet-of Things (IoT) is getting connected to an increasing number of mobile devices such as autonomous vehicles, drones and robots. Termed as Mobile Industrial Internet-of Things (MI2oT) devices in this paper, a key requirement of these devices is to accurately estimate range and Doppler in various applications, in addition to data communication. Research efforts, therefore, include incorporating MI2-oT devices with high-data rate communications together with Frequency Modulated Continuous Wave Radar (FMCW) sensing capabilities. Range and Doppler sensing, in FMCW radars is undertaken by a two-stage Fast Fourier Transform (FFT) which is computationally demanding. It is challenging to design baseband processing with FFTs that can be implemented as low computational hardware or application specific integrated circuits (ASIC) in MI2oT devices. This paper, presents a novel range and Doppler sensing technique based on Goertzel filtering, leading to considerable reduction in computations compared to the FFT. FMCW radar with Goertzel filtering and FFT are examined in three cases viz., sensing the range and velocity of a car, vibration and respiration monitoring. Simulation results show a computation reduction of the order of 6.3×, 7.7× and 8.1× in giga-operations per second (GOPS) for the three cases respectively. The reduced computations increase the feasibility of implementing range and Doppler sensing in MI2oT devices which have restricted computational resources.

A new approach for monitoring myocardium edema is introduced. It utilizes electrical impedance spectroscopy (EIS) to characterize edema tissue. The system has been tested on an isolated pig heart to distinguish the edema from non-edema tissue by a decrease of impedance for edema. The impedance reduction for left ventricular was 10 Ω at 100 kHz. The algorithm used for the demodulation in the developed EIS system is based on Goertzel filter which is utilized to replace the traditional coherent demodulation technique. Multisine excitation with 16 tones in the frequency range up to 1 MHz with more than 1000 samples was used for the measurements.

Cryogenic CMOS is a crucial subcomponent of quantum-technological applications, particularly as control electronics for quantum computers. Simulation is an important first step in designing any CMOS circuit. However, the standard BSIM4.5 model is only applicable for temperatures between 230 K and 420 K. In this work, N-type MOSFETs with different dimensions in a 65-nm CMOS technology were characterized at room temperature and liquid helium temperature (4.2 K). These measurements were compared with corresponding simulations from the BSIM4.5 model. A model of drain current in the triode region was constructed, where key parameters, such as threshold voltage and effective mobility, were modified. By adjusting these temperature-dependent parameters, the modified model predicted the triode region currents with an error reduced to 7.6%. Thus, the modified model can be utilized to simulate transistor behavior in the triode region at cryogenic temperatures.

Wireless data telemetry for implantable medical devices (IMDs) has, in general, been limited to a few Mbps, and used for applications such as transmitting recordings from an implanted monitoring device, or uploading commands to an implanted stimulator. However, modern neural interfaces need to record high resolution potentials from hundreds of neurons; this requires much higher data rates. While fast wireless communication is possible using existing standards such as WiFi, power consumption demands are far too high for IMDs. Short range inductive link based telemetry, in particular impulse-based systems such as pulse-harmonic modulation (PHM), have demonstrated transfer speeds of up to 20 Mbps with a small power budget. However, these systems require complex and precise circuits, making them potentially susceptible to inter-symbol-interference. This work presents a new method named Short-range Quality-factor Modulation (SQuirM), which retains the low power consumption and high data rate of PHM, while improving the resilience of the system and simplifying the circuit design. Transmitter and receiver circuits were fabricated using 0.35 micron CMOS. The circuits were capable of transceiving data at speeds of up to 50.4 Mbps, with a BER of <; 4.5E-10, and a transmitter energy consumption of 8.11pJ/b.

Wireless biosensors are playing a pivotal role in health monitoring, disease detection and management. The development of wireless biosensor nodes and networks strongly relies on the design of novel low-power, low-cost and flexible CMOS sensor readouts. This paper presents a CMOS potentiostat that integrates a control amplifier, a dual-slope ADC and a wireless unit on the same chip. It implements a novel time-based readout scheme, whereby the counter of the dual-slope ADC is moved to the receiver and the sensor current is encoded in the timing between two wireless pulses transmitted via pulse-harmonic modulation across an inductive link. Measured results show that the potentiostat chip can resolve a minimum input current of 10 pA at a sampling frequency of 125 Hz and a power consumption of 12 uW.

A color-coded full reference signal-to-noise (FR-SNR) method is proposed as a simple and reproducible performance evaluation method for EIT systems. It will be demonstrated with a 16-electrode high-speed EIT system and a resistive phantom. The performance of the EIT system can be altered by artificially introducing non-ideal operating conditions such as low drive current, low ADC resolution, dc offset in the readout front-end and large electrode impedance. The performance under these conditions will be measured and displayed using the FR-SNR method.

A 32 channel wearable torso shape detection belt will be demonstrated. The belt is designed to measure the torso shape of a neonate and provide real-time boundary information to assist the electrical impedance tomography (EIT) system to produce high quality lung respiration images. The system is fully integrated on a flexible printed circuit board which is encapsulated in a silicon wearable cover. During the live demonstration, while EIT images are reconstructed, the boundary shape can be changed to improve the image.

This paper describes the design of fully differential sine pulse-width-modulation (SPWM) wave current generator for bioimpedance measurement applications. The current generator has been designed in a 0.18-μm CMOS technology. Its analog front-end operates from ±1.65 V and has a current consumption of Iout+ 22 μA + (fclk× 1.7 pA) where Ioutis the output current and fclkis the operating frequency. It can provide outputs from 50 μAppto 1 mAppof SPWM current up to 98 kHz with a maximum voltage compliance of ±1.25 V. Using linear current feedback, the current generator has a designed transconductance of 1 mA/V. Feedback also enables cancellation of common mode signals and a high output impedance.

Background and Significance: Mechanical alternans (MA) is a biomarker associated with mortality and life-threatening arrhythmias in heart failure patients. Despite showing prognostic value, its use is limited by the requirement of measuring the continuous blood pressure (BP), which is costly and impractical. Objective: To develop and test, for the first time, non-invasive MA surrogates based on photoplethysmography (PPG). Methods: Continuous BP and PPG were recorded during clinical procedures and tests in 35 patients. MA was induced either by ventricular pacing (Group A, N = 19) or exercise (Group B, N = 16). MA was categorized as sustained or intermittent if MA episodes were observed in at least 20 or between 12 and 20 consecutive beats, respectively. Eight features characterizing the pulse morphology were derived from the PPG, and MA surrogates were evaluated. Results: Sustained alternans was observed in 9 patients (47%) from Group A, whereas intermittent alternans was observed in 13 patients (68%) from Group A and in 10 patients (63%) from Group B. The PPG-based MA surrogate showing the highest accuracy, V1M, was based on the maximum of the first derivative of the PPG pulse. It detected both sustained and intermittent MA with 100% sensitivity and 100% specificity in Group A and intermittent MA with 100% sensitivity and 83% specificity in Group B. Furthermore, the magnitudes of MA and its PPG-based surrogate were linearly correlated (R2= 0.83, p <; 0.001). Conclusion: MA can be accurately identified non-invasively through PPG analysis. This may have important clinical implications for risk stratification and remote monitoring.

This paper presents an application specific integrated circuit (ASIC) for use in a three-dimensional microchannel neural interface. The device is assembled with seven stacked ASICs with silicone microchannels instrumented in between. Each ASIC comprises tripolar electrodes for seven channels allowing recording or stimulation from any one of the 49 microchannels. The ASIC is implemented in 0.35μm high-voltage CMOS technology and occupies an active area of 4 mm2. The device has been tested demonstrating recording of 1 mV signals, and current controlled stimulation in the range of 5 μA to 500 μA (with 40 V compliance) and up to 50 kHz stimulation frequency. The device overcomes limits on numbers of connected microchannels in previous designs.

Point-of-care (PoC) diagnostics rely on the design of low-power and miniaturized readout units that can offer rapid and accurate test results, replacing the need for specialized equipment. CMOS technology can be exploited in order to design complex systems while achieving high energy efficiency for suitable operation in a mobile settings. This paper presents the design of a novel energy-efficient 4-channel wireless potentiostat chip, based on a dual-slope ADC architecture, that features a low-complexity wireless unit and a calibration approach that does not require additional circuitry. The chip was designed in a 0.35μm CMOS process. The simulated results suggest that each potentiostat channel can achieve an estimated energy efficiency of 2.5 pJ/bit from a 1.2 V supply.

There remains a need for satisfactory integrated sensors that can monitor humidity inside small active implanted devices in which delicate electronics are operating. This paper describes a simple capacitance-to-digits readout circuit for a capacitive humidity sensor for ultra-small medical implants. The proposed readout circuit provides a larger dynamic range than conventional capacitance readout circuits by removing the parasitic capacitance from the measured capacitive input. The circuit was designed in XFAB's 0.6-μm CMOS technology and provides 1 % resolution in measuring relative humidity.

A wearable electrical impedance tomography (EIT) system for hand prosthesis motion control is demonstrated. The system captures the user's hand motion by measuring the impedance alterations caused by muscle and bone movement inside the forearm. These impedance data are sent to an artificial neural network for motion classification which is then used to manipulate a hand prosthesis. During the live demonstration, a sensor band is put on a volunteers' forearm for data acquisition. After signal processing, hand gestures learnt by the neural network can be recognized and the same hand motion can be recreated through the hand prosthesis in real-time.

This paper presents a high accuracy hand gesture recognition system based on electrical impedance tomography (EIT). The system interfaces the forearm using a wrist wrap with embedded electrodes. It measures the inner conductivity distributions caused by bone and muscle movement of the forearm in real-time and passes the data to a deep learning neural network for gesture recognition. The system has an EIT bandwidth of 500 kHz and a measured sensitivity in excess of 6.4 Ω per frame. Nineteen hand gestures are designed for recognition, and with the proposed round robin sub-grouping method, an accuracy of over 98% is achieved.

This paper presents the design of a 1024-channel dual-modality CMOS biosensor suitable for both capacitive sensing and impedance spectroscopy. The chip serves as a platform for detection, localization and monitoring of bacteria and can be adopted for affinity-based assays. The chip features a 32×32 array of unpassivated metal electrodes formed on the top metal of a 0.18 μm CMOS process, with an overall sensing area of 2.06 mm2. The system design is based on a shared in-pixel integrator that can be used as a charge amplifier for capacitive sensing (CS) or as part of a transimpedance amplifier for electrical impedance spectroscopy (EIS). The CS mode is capable of a operation bandwidth of 50 MHz at a current consumption of 82 μA per pixel. The EIS channel operates over a bandwidth between 100 Hz and 1 MHz with a total input-referred current noise of 48 pArms and a current consumption of 210 μA per channel.

A wearable electrical impedance tomography system for lung respiratory monitoring will be demonstrated. The system features an active electrode integrated circuit (IC) fabricated in 0.35-μm CMOS technology. The IC is mounted on a flexible printed circuit board which is then embedded in a textile belt. During the live demonstration, the belt can be put on the volunteers' chest, and through a back-end signal processing hub, the system is able to provide real-time lung respiratory images on a remote computer.

This paper presents a fully implantable multi-channel neural prosthesis for epidural stimulation. The prosthesis features three telemetry-operated independent stimulators providing in total eighteen stimulation channels. The stimulator circuits were implemented in a 0.6-μm CMOS technology. The prosthesis is protected in a hermetically sealed ceramic enclosure and encapsulated in medical grade silicone rubber. In-vitro measured results with electrodes in saline are presented.

A fully implantable multi-channel neural prosthesis for epidural stimulation will be demonstrated. The prosthesis features three telemetry-operated independent stimulators providing in total eighteen stimulation channels. The stimulator circuits were implemented in a 0.6-μm CMOS technology. The prosthesis is protected in a hermetically sealed ceramic enclosure and encapsulated in medical grade silicone rubber for long-term implantation. During the live demonstration in-vitro tests with electrodes in saline will be performed with the prosthesis operated wirelessly from a remote computer.

A wideband, CMOS current driver for bioimpedance measurement applications has been designed employing nonlinear feedback. With the introduction of phase compensation, the circuit is able to operate at frequencies higher than the pole frequency of the output transconductor with minimum phase delay. Moreover, it isolates the poles required for stability from the high frequency characteristics of the output transconductor. The circuit has been simulated in a 0.35-μm CMOS technology and operates from ±2.5 V power supplies. Simulations show that for a 1 mAp-p output current, the phase delay is less than 1° for frequencies up to 3 MHz, rising to 1.5° at 5 MHz. Dual frequency currents to the load are demonstrated.

This paper describes the design of an active electrode integrated circuit (IC) for a wearable electrical impedance tomography (EIT) system required for real time monitoring of neonatal lung function. The IC comprises a wideband high power current driver (up to 6 mAp-poutput current), a low noise voltage amplifier and two shape sensor buffers. The IC has been designed in a 0.35-μm CMOS technology. It operates from ±9 V power supplies and occupies a total die area of 5 mm2. Post-layout simulations are presented.

An integrated circuit has been developed for a basic nonlinear negative feedback current driver which isolates the poles required for stability from the high frequency characteristics of the output transconductor. It provides a 1 MHz, 1 mAp-pcurrent with a phase delay of 16° into a 1kΩ load. While it can be designed as a stand-alone circuit its main use is for application in a proposed circuit using multiple copies of this basic circuit, which significantly extends its high frequency performance by substantially reducing phase delay at high frequencies. The circuit was fabricated in a 0.35-μm CMOS technology and measured results are presented. It is intended for bioimpedance applications.

This paper presents a design methodology for power optimized neural frontend interfaces (NFIs). The minimum power is analytically derived based on the constituent building blocks of the NFI, namely low noise amplifier, programmable gain amplifier and analog-to-digital converter (ADC), all of which are bound by noise. In the optimization process two types of ADC are considered: the successive approximation register (SAR) and the comparator-based switched-capacitor (CBSC) cyclic converter. It is shown that for a resolution of 7 bits (optimum for spike sorting) the use of the CBSC cyclic ADC results in ~30% power reduction compared to the conventional SAR ADC. The derived minimum power limits take into account the accuracy requirements for spike sorting and are compared with a number of NFIs in the literature.

A high voltage supply is necessary for implantable stimulators that need to deliver high stimulus current to load tissue and overcome large electrode impedance. A constantly high supply voltage results in unnecessary power wastage in low-voltage stimulation and is potentially dangerous at the neural tissue interface. This paper presents the authors' recent progress in developing a stimulation system with self-adapting supply based on the previous Active Books system. The improved design is able to measure the peak electrode voltage on any specified anode and feed back the information to a central hub unit. If the peak electrode voltage is significantly lower than the current supply voltage, the adjustable voltage regulator in the hub will issue a lower but high enough supply voltage to power up the stimulator chip. The circuit has been designed and simulated in a 0.6-μm HV CMOS process.

The comparator-based switched-capacitor (CBSC) technique has been used in low power analog to digital converters (ADCs). The objective of this paper is to derive the theoretical power bound for CBSC-based pipelined ADCs with digital error correction (1.5 bit/stage). To achieve this, the constituent building blocks whose performance is limited by noise, are examined to derive the power bound. The optimum values of design parameters influencing the power bound are also investigated including the optimal output ramp rates needed to achieve a given linearity constraint, comparator bias current and delay time. The accuracy of the derived equations is verified through behavioral simulation in MATLAB.

The aim of this paper is to compare three recording methods for interference rejection from tripolar nerve cuffs by testing in a saline-filled tank. The three methods were: the quasi-tripole (QT), the modified-quasi-tripole (mQT) and adjusted true tripole (aTT). When the interfering electric field that surrounds the cuff is solonoidal and coaxial with the cuff, the mQT allows the interference that happens to be present due to asymmetry to be neutralized at one frequency, and reduced across the ENG band. The aTT is better in this situation, allowing almost complete neutralization at all frequencies. However, when the electric field is altered after adjustment, the interference was greater for the aTT than the mQT. In an implant, the interference is likely to come from nearby muscles and the fields will be continuously changing as the action potentials propagate. Therefore, it is likely that mQT will be better than the aTT.

This paper describes the stimulation management unit for a multichannel vestibular neural prosthesis. The unit is designed as part of a stimulator ASIC in the implantable subsystem of the prosthesis. This digital unit provides the stimulator ASIC the ability to generate biphasic current pulses at specified amplitude, duration and pulse rate to drive electrodes in the semicircular canals. The circuit was implemented in 0.6-μm CMOS technology and post-layout simulations are presented to show its operation.

Impedance detection plays a crucial role in the development of Lab-on-a-Chip (LOAC) technologies. Various types of impedance biosensors have been developed over the years, with the main focus being centered on interdigital sensors. Optimization of impedance sensors has only recently attracted interest. In this paper we present a conformal mapping approach to the optimization of bipolar and tetrapolar impedance sensors. This technique enables an analytic equation for the sensitivity of the sensors to impedance changes to be obtained. Optimization of these sensors using this analytic equation was performed and a number of optimized sensor designs are presented in this paper.

We propose a novel dual-mode neural stimulator circuit. For stimulation requiring small or medium amount of charge, the stimulator supplies a constant current to the stimulation load in the same way as any standard current-mode stimulator. For high-charge stimulation applications, the stimulator supplies a variable stimulus current, depending on the voltage available across the current generator circuit. By adjusting the bias current accordingly, the stimulator maintains a high output impedance, independent of the stimulus current and the load voltage at the time. The total stimulus charge from the proposed stimulator is theoretically as high as voltage-mode stimulators. Simulated results using a 0.6-μm CMOS process are presented. At low stimulus current, the output impedance of the proposed stimulator circuit is at least six times higher than a conventional current-mode generator with constant bias current.

We present a current generator circuit which supplies two anodic currents for tripolar nerve stimulation. Depending on applications and the more convenient method of adjustment, the difference of the two anodic currents can be programmed using either of the two different modes. The reference current circuit in the current generator utilizes a gate-drain tied depletion transistor working in the triode region as the degenerating resistor in the Beta-multiplier topology. The negative temperature coefficient of the depletion transistor reduces the total sensitivity of the output current to temperature variation. The proposed current generator circuit is also stable over a wide range of supply voltages. Simulated results using a 0.6-μm CMOS process are presented. The temperature coefficient and power supply rejection ratio of the current generator are less than 0.06%/°C and 2%/V, respectively.

Recording neural information using implanted cuff electrodes in a tripolar arrangement is affected by myoelectric interference present at the implantation site. The performance of the recently proposed mQT was suboptimal in that tuning was required to compensate for such interference. In this work we present an improvement where interference in the mQT is removed by modeling the impedance profile of our electrodes. Simulation data show that by employing this “adaptive” approach, a 33% imbalance in the axial resistances when the interference is of 10 mV at 500 Hz and 10 kHz, results in an interference breakthrough that is about 7230 and 100 times less, respectively, at the amplifier's output.

A weakly inverted MOS transconductor with wide input linear range and very low transconductance is proposed. The transconductor is achieved by efficient input voltage attenuation formed by connecting identical simple source coupled pairs in a scheme that provides very high extensibility of voltage scaling. Post-layout simulated results in a 0.35-μm standard CMOS process show that Gmin the range of 0.05-1.3 nA/V is achievable along with 380 mVp-plinear range, at 1% Gmvariation under 1.5-V supply. Application in first-order low-pass sub-hertz filtering (with tunable cut-off frequency from 0.3-3 Hz) is demonstrated.

Deep brain stimulation (DBS) is a widely accepted clinical tool adopted for the treatment of a number of motor disorders. Despite its clinical efficacy, its underlying mechanisms have not been yet fully understood. One major issue that we identify as partly responsible for this lack of understanding is related to the poor control over the size, shape and location of the distribution of the depolarizing field around the electrode. With this and a parallel work the authors are proposing to develop and implement techniques that would allow for some degree of control over the distribution of the potential fields. This paper presents the application to DBS of a technique based on a tripolar current source configuration, with adjustable current flow through the lateral (anodic) branches of the tripole. The behavior of potential fields in the tissue were simulated by adopting FEM models of DBS electrode implanted in brain tissue. The profiles of simulated and measured fields were in agreement and they have shown how a dynamic tripolar current source can be adopted to obtain increase the focus and control the location of distribution of the fields around the electrode.

In this study, the authors investigate the performance-complexity tradeoff of convolutional codes for broadband fixed wireless access systems by considering the effects of quantisation and path metric memory in practical Viterbi decoding implementations. They show that in systems with limited antenna diversity, low-memory codes achieve a better error-rate performance compared to that of high-memory codes. Only in systems with considerable antenna diversity, can the performance of a convolutional code be improved by increasing its memory size. Nevertheless, the authors demonstrate that the coding advantage offered by the high-memory codes is not large enough to justify the significant increase in implementation complexity. In particular, memory-2 convolutional codes achieve a coding gain of up to 1.2 dB over their memory-8 counterparts in single-input single-output fixed wireless access systems. The situation is reversed when multiple antennas are used, but the decoder of memory-8 codes occupies at least 130 times more silicon area than that of memory-2 codes.

An integrated amplifier for electroneurogram (ENG) recordings from tripolar cuff electrodes is described. The amplifier is dedicated to urinary incontinence and other functional electrical stimulation (FES) applications. To remove myoelectric (EMG) interference a parallel RC network is used to balance the electrode impedances in the quasi-tripole amplifier configuration. The various ENG amplifier settings, such as resistance and capacitance trimming for the neutralization RC network, amplifier gain and filter cut-off frequencies, are controlled by an external microcontroller which communicates with the embedded SPI (Serial Port Interface) block in the amplifier. By this topology control of the ENG amplifier is executed in software allowing for the system parameters to be re-configured after implantation. The analog system blocks and details of the SPI logic are described. The amplifier was fabricated in a 3-V 0.35-mum BiCMOS process technology and preliminary measured results are reported. Input signals as low as 1muV can be reliably detected. The amplifier occupies an area of 1.5mm2and consumes about 1.4mW when configured to detect sub-microvolt neural signals.

In this paper we present a passive impedance network that offers, to conventional QT configuration, the ability to null EMG interference without requiring tuning. The network is based on the Randles circuit model which fits the electrode- electrolyte impedance profile of our platinum electrode book, throughout the ENG band. The calculated RMS noise input voltage of the system was found to be 0.73 uV, thus making the solution plausible in terms of noise performance.

We implemented an ASIC for readout and control of a contactless inductive position sensor. This type of sensor is used to measure linear and angular positions in industrial applications. An integrated front-end design is described that reduces phase sensitive offsets. The ASIC was designed in a 5 V 0.35-mum CMOS technology and preliminary measured results are presented. The integration of the sensor signal conditioning circuits in an ASIC offers considerable advantages in terms of size and performance compared to existing discrete solutions.

A signal acquisition system is presented which is suitable for small signals in the presence of low-frequency interference. In a conventional design filters precede the input stage to remove offset and interference. However, filters reduce the available recording bandwidth, are difficult to integrate and remove potentially valuable information about the interference signal and they yield long settling time after disturbance. Instead of removing signal content, the proposed approach extends the available input range. Floating batteries and a simple digital control are used to prevent the input amplifier from saturating. A layout of the system in a 0.35 mum CMOS process is presented and transistor level simulation results are discussed to confirm the operation of the system.

There is a great demand to reduce the volume of implantable multi-channel stimulators for functional electrical stimulation applications. Miniaturization is currently limited by the size of the large off-chip blocking capacitors that are required for safety. This paper describes a fail-safe stimulator output stage circuit utilizing the principle of high-frequency current-switching, which allows the blocking capacitors to be integrated on-chip. The proposed circuit is based on the bridge rectifier circuit, hence fully utilizing the bidirectional current through the blocking capacitor. The approach has been verified by post-layout simulations in a 1 μm CMOS SOI technology.

The prime concern in the design of biomedical implants is safety. This paper addresses several safety features and proposes a new scheme to transmit power and data from the central implant to the stimulator "pods" while maintaining failsafe operation. The proposed scheme demultiplexes the stimulation command close to the electrodes and uses three wires in total to provide power and data to all stimulator "pods" from the central implant. Thus, it greatly reduces the number of wires routed to the stimulation sites and is able to stimulate more stimulator "peas". The circuit used to recover the power is shown, utilizing the three wires as the only inputs. The power recovery circuit features a varactor as the storage capacitor which gives a faster pumping speed than a linear capacitor. The idea is verified with circuit simulations in a 0.6μm CMOS technology. The proposed scheme ensures that in the event of cable failure, the current will be charge-balanced, thereby avoiding the generation of harmful electrochemical products.

This paper describes a 10-bit analog-to-digital converter (ADC) for digital video broadcasting-handheld (DVB-H) systems. The ADC is based on a 2.5-2.5-2.5-4 bits-per-stage pipeline architecture and occupies an area of 1.3 mm2in a 0.35 mum CMOS process. At the target sampling rate of 20.48 MS/s, measured results show that the converter consumes 19.5 mW from a 1.5 V power supply and achieves 56 dB SNR and 60 dB SFDR. Effective resolution bandwidth is 100 MHz and energy consumption per conversion is 0.19 pJ

The possible benefits of moderate inversion on harmonic distortion for MOSFETS in linear and squarer transconductors are examined. The application of the simple equation on which the EKV model is based provides analytic information on the minima in the harmonic distortion at certain bias points. The effect on harmonic distortion of increasing the input signal amplitude is also discussed. Theoretical and simulated results are compared.

The use of blocking capacitors in the output stages of implantable stimulators for functional electrical stimulation (FES) applications, is common practice for safety reasons. However, these capacitors tend to dominate the implant volume. This paper describes a stimulator output stage circuit which overcomes this limitation. The circuit features a novel capacitor reduction technique for implant miniaturization, and a simple failure-checking mechanism for enhanced reliability. The approach was verified by simulations in a 0.35 mum CMOS technology

The output signal-to-interference (ENG/EMG) ratio of amplifier configurations connected to tripolar nerve cuff electrodes is affected by systematic interface errors which severely degrade neural recordings. These errors are contained in the term "cuff imbalance" and known causes include inhomogeneous tissue growth after cuff implantation and cuff manufacturing tolerances. In this paper we report on an additional contribution to cuff imbalance that stems from variations in orientation and distance (both factors contained here in the term proximity) of the tripolar cuff relative to the external interference source (represented here by a dipole). Interference amplitude is also shown to depend on proximity of the muscle to the cuff.

Electrical bioimpedance spectroscopy is a fast and relatively easily applicable method for tissue characterization. This paper describes a CMOS instrumentation amplifier based on the principle of current feedback for use in this application. The amplifier was designed in a 0.35 /spl mu/m CMOS technology and dissipates 420 /spl mu/W from /spl plusmn/1.5 V power supplies. Simulations showed that it has a bandwidth of 2 kHz-1 MHz, a mid-band gain of 20 dB, a CMRR of 101 dB at 1 MHz, and an input-referred noise voltage floor of 11.5 nV//spl radic/Hz.

We propose a new architecture which enables a motion estimation (ME) processor to be constructed using analogue components. At present, due to the arduous scale and nature of the computations required in digital hardware realisations, the ME processor is invariably the most complex and expensive component of a modern video codec. Our preliminary results indicate that very considerable reductions in size, power consumption and hence cost are available using our proposed method.