Muhammad Awais Bin Altaf
@lums.edu.pk
Assistant Professor, Electrical Engineering Department
Lahore University of Management Sciences
Scopus Publications
Scopus Publications
Muhammad Sheeraz, Abdul Rehman Aslam, Emmanuel Mic Drakakis, Hadi Heidari, Muhammad Awais Bin Altaf, and Wala Saadeh
MDPI AG
Autism spectrum disorder (ASD) is a chronic neurological disorder with the severity directly linked to the diagnosis age. The severity can be reduced if diagnosis and intervention are early (age < 2 years). This work presents a novel ear-worn wearable EEG system designed to aid in the early detection of ASD. Conventional EEG systems often suffer from bulky, wired electrodes, high power consumption, and a lack of real-time electrode–skin interface (ESI) impedance monitoring. To address these limitations, our system incorporates continuous, long-term EEG recording, on-chip machine learning for real-time ASD prediction, and a passive ESI evaluation system. The passive ESI methodology evaluates impedance using the root mean square voltage of the output signal, considering factors like pressure, electrode surface area, material, gel thickness, and duration. The on-chip machine learning processor, implemented in 180 nm CMOS, occupies a minimal 2.52 mm² of active area while consuming only 0.87 µJ of energy per classification. The performance of this ML processor is validated using the Old Dominion University ASD dataset.
Gul Hameed Khan, Nadeem Ahmad Khan, Wala Saadeh, and Muahammad Awais Bin Altaf
Springer Nature Switzerland
Gul Hameed Khan, Nadeem Ahmad Khan, Muhammad Awais Bin Altaf, and Qammer Abbasi
MDPI AG
This paper presents a trainable hybrid approach involving a shallow autoencoder (AE) and a conventional classifier for epileptic seizure detection. The signal segments of a channel of electroencephalogram (EEG) (EEG epochs) are classified as epileptic and non-epileptic by employing its encoded AE representation as a feature vector. Analysis on a single channel-basis and the low computational complexity of the algorithm allow its use in body sensor networks and wearable devices using one or few EEG channels for wearing comfort. This enables the extended diagnosis and monitoring of epileptic patients at home. The encoded representation of EEG signal segments is obtained based on training the shallow AE to minimize the signal reconstruction error. Extensive experimentation with classifiers has led us to propose two versions of our hybrid method: (a) one yielding the best classification performance compared to the reported methods using the k-nearest neighbor (kNN) classifier and (b) the second with a hardware-friendly architecture and yet with the best classification performance compared to other reported methods in this category using a support-vector machine (SVM) classifier. The algorithm is evaluated on the Children’s Hospital Boston, Massachusetts Institute of Technology (CHB-MIT), and University of Bonn EEG datasets. The proposed method achieves 98.85% accuracy, 99.29% sensitivity, and 98.86% specificity on the CHB-MIT dataset using the kNN classifier. The best figures using the SVM classifier for accuracy, sensitivity, and specificity are 99.19%, 96.10%, and 99.19%, respectively. Our experiments establish the superiority of using an AE approach with a shallow architecture to generate a low-dimensionality yet effective EEG signal representation capable of high-performance abnormal seizure activity detection at a single-channel EEG level and with a fine granularity of 1 s EEG epochs.
Muhammad Sheeraz, Wala Saadeh, and Muhammad Awais Bin Altaf
IEEE
Current electroencephalogram (EEG) measuring systems are bulky, impose constraints on patients, and require pre and post-measuring procedures. Usually, the EEG systems use either wet or dry EEG sensors, with the former suffers from skin preparation, the issues of adhesive conductive gels, and one-time usability whereas the latter causes skin irritation, abrasion, and pain upon pressure. Hence, these sensors are not suitable for long-term measurements. This paper presents a novel, wireless, behind-the-ear wearable EEG acquisition device that incorporates flexible dry EEG sensors. Silver ink-printed flexible sensors are fabricated using screen printing to overcome the above-mentioned drawbacks and limitations of conventional EEG sensors. The flexible sensors form a capacitive link with the skin via an adhesive layer between the sensor and the person's skin and are capable of acquiring the EEG without any skin preparation or gel. The performance of the printed flexible EEG sensors is tested by comparing them with the standard Ag/AgCl pre-gelled sensors. The alpha wave test and evoked potential EEG test are also performed for verification. The proposed device has a small form factor similar to a hearing aid and an in-house configurable Analog Front End (AFE) and Digital Back End (DBE) Processor and is capable of acquiring continuous EEG for a longer duration in a user-friendly and socially discrete manner.
Abdul Rehman Aslam, Nauman Hafeez, Hadi Heidari, and Muhammad Awais Bin Altaf
Frontiers Media SA
Autism Spectrum Disorder (ASD) is characterized by impairments in social and cognitive skills, emotional disorders, anxiety, and depression. The prolonged conventional ASD diagnosis raises the sheer need for early meaningful intervention. Recently different works have proposed potential for ASD diagnosis and intervention through emotions prediction using deep neural networks (DNN) and machine learning algorithms. However, these systems lack an extensive large-scale feature extraction (LSFE) analysis through multiple benchmark data sets. LSFE analysis is required to identify and utilize the most relevant features and channels for emotion recognition and ASD prediction. Considering these challenges, for the first time, we have analyzed and evaluated an extensive feature set to select the optimal features using LSFE and feature selection algorithms (FSA). A set of up to eight most suitable channels was identified using different best-case FSA. The subject-wise importance of channels and features is also identified. The proposed method provides the best-case accuracies, precision, and recall of 95, 92, and 90%, respectively, for emotions prediction using a linear support vector machine (LSVM) classifier. It also provides the best-case accuracy, precision, and recall of 100% for ASD classification. This work utilized the largest number of benchmark data sets (5) and subjects (99) for validation reported till now in the literature. The LSVM classification algorithm proposed and utilized in this work has significantly lower complexity than the DNN, convolutional neural network (CNN), Naïve Bayes, and dynamic graph CNN used in recent ASD and emotion prediction systems.
Gul Hameed Khan, Nadeem Ahmad Khan, and Muhammad Awais Bin Altaf
IEEE
Epileptic patients ’ quality of life can be significantly improved by epileptic seizure prediction based on scalp electroencephalogram (EEG). With the advancement of brain e-health technologies, there is an essential need for a method that accurately predicts seizures while running on computing platforms with very low computing resources. Moreover, existing methods do not provide EEG analysis on an individual channel basis to identify the abnormalities in the data. In order to address this issue, we propose an efficient framework for patient-specific seizure prediction. A hybrid model comprising of a shallow autoencoder (AE) with only one hidden layer and a support vector machine (SVM) classifier has been developed. Both multi-channel and single channel EEG signal processing schemes have been developed. Generating a lower dimensional sparse signal with AE in the first stage and classifying the signal using SVM in the second stage are the two stages that the model separates into when processing EEG data. We initially train the AE to provide an optimum sparse signal and then use this sparse signal as input for an SVM classifier to categorize the EEG data. Using the 10-fold cross validation strategy, the proposed model tests 13 patients from the CHB-MIT dataset and achieves an average sensitivity of 98% and an average area under the curve (AUC) of 99%. We have compared our hybrid approach ’s performance with both deep learning models and traditional techniques. The proposed methodology outperforms state of the art seizure prediction methods, demonstrating its effectiveness.
Naureen Mujtaba, Ishtiaq Rasool Khan, Nadeem Ahmad Khan, and Muhammad Awais Bin Altaf
IEEE
Tone mapping is essential for displaying High Dynamic Range (HDR) images on Low Dynamic Range (LDR) screens. The quality performance of these algorithms has improved greatly over recent times, however their computational complexity has also considerably increased. Real-time applications are constrained because of this limitation in their deployment. Therefore, a fast tone-mapping algorithm is presented that maintains the same quality level as state-of-the-art techniques but has a considerably lower computational complexity. Based on an existing well-known Threshold vs. Intensity (TVI) model, a simplified human eye sensitivity model is derived that leads to a non-iterative method of TMO construction. The proposed model is used to construct a histogram of the HDR image such that the bins vary in size in the spatial domain but each of them covers the same amount of change in the luminance perceived by the human eye. Moreover, a simple yet effective solution is suggested to the over-enhancement and over-compression problems faced by histogram-based techniques. The proposed architecture is computationally light, hardware-friendly, produces superior visual results, and works significantly faster than state-of-the-art techniques.
Muhammad Sheeraz, Abdul Rehman Aslam, Nauman Hafeez, Hadi Heidari, and Muhammad Awais Bin Altaf
IEEE
High blood pressure (BP) is a major source of health problems related to mental stress, cardiac issues, kidney problems, vision, and brain. High BP bursts can damage and rupture blood vessels and cause strokes. Therefore, it is quite important to continuously monitor it for high BP patients. Conventional BP monitoring devices a) can cause discomfort and b) not suitable for intermittent monitoring. The photoplethysmographic (PPG) signals measure the volume changes in the human blood through human skin. This work presents a high BP classification processor using PPG signals through an artificial intelligence (AI) based boosted circuit. A data set of 25 participants was collected. Ten out of the 25 participants were high blood pressure patients with systolic BP (SBP) and, diastolic BP (DBP) values higher than 140mmHg and 90mmHg, respectively. The AI boosted circuit calculates the power spectral densities, power spectral densities difference, and the sum of the consecutive difference between PPG signals. The features are forwarded to a small 3-level Decision Tree (DT) classifier. The decision tree classifier classifies the high SBP and DBP as high or normal/low with 96.2% classification accuracy. The SBP values $\\geq$ 130mmHg and $<$ 130mmHg were classified as HIGH SBP or LOW/NORMAL SBP respectively. Similarly, the DBP values $\\geq$ 80mmHg and $<$ 80mmHg were classified as HIGH DBP or LOW/NORMAL DBP, respectively. The system was implemented on an Artix-7 FPGA which consumes power of $\\approx 18.23uW$ @ 50 MHz.
Muhammad Sheeraz, Abdul Rehman Aslam, and Muhammad Awais Bin Altaf
IEEE
Health problems related to stress are increasing globally and significantly affect the mental health and quality of life of human beings. Continuous suffering from stress may lead to serious psychological and physical health problems. But still, no effective and reliable stress detection methods are available. In this paper, a novel wearable device is presented to measure electroencephalogram (EEG) and electrocardiogram (ECG) simultaneously in a non-invasive approach. This system includes an analog front end (AFE) integrated with a machine learning-based digital backend (DBE) processor for mental stress prediction using only 3 electrodes. A PCB prototype is developed using the commercial off-the-shelf components. The developed prototype shows excellent noise performance of $0.1\\mu V_{rms}$ and predicts the mental stress with a classification accuracy of 92.7%. The proposed system is lightweight and easily wearable (behind the ear). The data is acquired from 25 participants for different stress scenarios including the Arithmetic Test and Stroop Color Word Test. Different EEG and ECG based features combinations are used for the classification of stress conditions using a shallow neural network (SNN) classifier.
Abdul Rehman, Muhammad Awais Bin Altaf, and Wala Saadeh
IEEE
Blood pressure (BP) is considered one of the key vital signs that provide valuable medical information about cardiovascular activity. Conventionally, cuff-based devices are used to measure BP which limits their usage for continuous monitoring. This paper presents a cuff-less BP estimation processor using photoplethysmography (PPG) signals with a Deep Neural Network (DNN). Spectral and temporal features are extracted from the PPG signals and then used to train and evaluate the machine learning (ML) algorithms. The proposed algorithm is evaluated using the MIMIC II database for systolic blood pressure (SBP) and diastolic blood pressure (SBP) estimation. The proposed BP estimation processor is implemented using a 180nm CMOS process with an area of 3.45mm2 and consumes $73 \\mu \\mathrm{W}$. It achieves a mean absolute error in systolic BP of $0.0657 \\pm 4.7$ mmHg and diastolic BP of $0.792 \\pm 4.61$ mmHg which outperforms the state-of-the-art BP estimation algorithms.
Muhammad Sheeraz, Abdul Rehman Aslam, Muhammad Awais Bin Altaf, and Hadi Heidari
IEEE
Electroencephalogram (EEG) play a vital role in the prediction of neurological disorders including epilepsy, narcolepsy, migraine, etc. The impedance between the electrodes and the skin interface should be within specific range in order to acquire good quality EEG signals. This work focuses on an intelligent, wearable device capable of measuring electrode-skin interface impedance (ESI) in parallel to the EEG acquisition in non-invasive manner. The proposed device includes an analog front end (AFE), back-end micro-controller (BEM), and ESI measuring unit. A novel technique is used for measurement of ESI by connecting known value resistances in parallel at the input terminals of the low noise amplifier (LNA) of AFE and measuring the change in the output signal. This unique technique is suitable in wearable medical devices as it is safe to use, results in high-quality EEG signals, and consumes negligible power. A PCB prototype is developed using the available commercial components. The designed prototype has a small form factor (52mm x 53mm), is lightweight, and easily behind the ear wearable.
Abdul Muneeb, Mubashir Ali, and Muhammad Awais Bin Altaf
IEEE
An electroencephalogram (EEG) based non-invasive 2-channel System on Chip (SoC) is presented to detect and report the seizure event of the epileptic patient. The SoC incorporates an area and power-efficient dual-channel analog front-end (AFE) and machine learning-based differential difference approximate computing seizure detection ($\\text{D}^{2}$ACSD) processor. The $\\text{D}^{2}$ACSD processor integrates approximate computing feature extraction and fixed-point linear support vector machine (LSVM) classifier to minimize the area-and-power utilization. The AFE comprises of two duty-cycled resistive MOSFET (DCRM) capacitively coupled instrumentation amplifier ($\\text{C}^{2}$IA), a programmable gain amplifier, and multiplexed SAR-ADC. The DCRM-C2IA utilizes proposed DCRM technique to boost the equivalent resistance of the integrator of the DC servo loop. The 5m$\\text{m}^{2}$SoC is implemented in 0.18$\\mu$m, CMOS process while achieving an average accuracy of 89.19%, sensitivity 92.18% and specificity 89.13% for the random and block-wise splitting of data in train/test sets. The implemented DCRM-C2IA achieves an integrated noise of 0.80$\\mu$Vrms over 0.5-100Hz frequency band. The realized system consumes $2.7\\mu \\text{J}/$classification to continuously detect seizure onset for timely suppression.
Abdul Rehman Aslam and Muhammad Awais Bin Altaf
Institute of Electrical and Electronics Engineers (IEEE)
An electroencephalogram (EEG)-based non-invasive 2-channel neuro-feedback SoC is presented to predict and report negative emotion outbursts (NEOB) of Autistic patients. The SoC incorporates area-and-power efficient dual-channel Analog Front-End (AFE), and a deep neural network (DNN) emotion classification processor. The classification processor utilizes only the two-feature vector per channel to minimize the area and overfitting problems. The 4-layers customized DNN classification processor is integrated on-sensor to predict the NEOB. The AFE comprises two entirely shared EEG channels using sampling capacitors to reduce the area by 30%. Moreover, it achieves an overall integrated input-referred noise, NEF, and crosstalk of 0.55 µVRMS, 2.71, and −79 dB, respectively. The 16 mm2 SoC is implemented in 0.18 um 1P6M, CMOS process and consumes 10.13 μJ/classification for 2 channel operation while achieving an average accuracy of >85% on multiple emotion databases and real-time testing.
Abdul Rehman Aslam, Nauman Hafeez, Hadi Heidari, and Muhammad Awais Bin Altaf
IEEE
Autism Spectrum Disorder (ASD) is the prevalent child neurological and developmental disorder causing cognitive and behavioral impairments. The early diagnosis is an urgent need for treatment and rehabilitation of ASD patients. This work presents an electroencephalogram (EEG) based ASD classification processor that targets a patch-form factor sensor that can be used for long time monitoring in a wearable environment. The selection of frontal and parietal lobe electrodes causes minimum uneasiness to the children. The proposed and implemented algorithm utilizes only four EEG electrodes. The processor is implemented and validated on Artix-7 FPGA which requires only 26K lookup tables and 15K flip flops. The hardware efficient implementation of the complex kurtosis value and Katz fractal dimension (KFD) features using kurtosis value indicator and KFD indicator with 54% and 38% efficient implementations, respectively, is provided. A hardware feasible shallow neural network architecture is used for the ASD classification. The system classifies the ASD with a high classification accuracy of 85.5% using the power and latency of $8.62 \\mu \\mathrm{W}$ and 2.25ms, respectively.
Zain Taufique, Bingzhao Zhu, Gianluca Coppola, Mahsa Shoaran, and Muhammad Awais Bin Altaf
Institute of Electrical and Electronics Engineers (IEEE)
Migraine is a disabling neurological disorder that can be recurrent and persist for long durations. The continuous monitoring of the brain activities can enable the patient to respond on time before the occurrence of the approaching migraine episode to minimize the severity. Therefore, there is a need for a wearable device that can ensure the early diagnosis of a migraine attack. This brief presents a low latency, and power-efficient feature extraction and classification processor for the early detection of a migraine attack. Somatosensory Evoked Potentials (SEP) are utilized to monitor the migraine patterns in an ambulatory environment aiming to have a processor integrated on-sensor for power-efficient and timely intervention. In this work, a complete digital design of the wearable environment is proposed. It allows the extraction of multiple features including multiple power spectral bands using 256-point fast Fourier transform (FFT), root mean square of late HFO bursts and latency of N20 peak. These features are then classified using a multi-classification artificial neural network (ANN)-based classifier which is also realized on the chip. The proposed processor is placed and routed in a 180nm CMOS with an active area of 0.5mm2. The total power consumption is $249~\\mu \\text{W}$ while operating at a 20MHz clock with full computations completed in 1.31ms.
Muhammad Rizwan Khan, Wala Saadeh, and Muhammad Awais Bin Altaf
IEEE
A wearable EMG based tremor detection and suppression system is presented. This work proposes a novel design enabling low-power consumption, wearability, lower computational cost and lower latency. An analog front end (AFE) is designed containing cascaded filters and a Driven-Right-Leg (DRL) feedback for high-level noise removal of up to 1V. A CC1352R microcontroller with an integrated BLE along with RTOS is utilized to achieve low-power processing. A user-friendly interface is provided using Android application (AP) that allows immediate sharing of data to caretakers or database. A 128-point FFT is employed with a simple implementation in terms of computation and a variable-voltage skin-impedance based muscle stimulation is being used. The system is operable on coin cell batteries for more than 3 weeks. The overall average power consumption of the system is 4.8mW with average current 1.35mA and a detection latency of <0.2s is achieved.
Shiza Shakeel, Niha Afzal, Gul Hameed Khan, Nadeem Ahmad Khan, Mujeeb ur Rehman Abid, and Muhammad Awais Bin Altaf
IEEE
Seizure type identification plays a pivotal part in the diagnosis and management of epileptic seizure disorder. Unfortunately, did not get much attention in past decades due to the unavailability of databases with seizure type marking. Seizure types not only assists the neurologist in deciding the correct drug and its dosage but precaution the epileptic patients about the seizure attack and its severity. In the recent past, a significant contribution has been made by applying machine and deep learning algorithms to the binary classification of generalized seizures. This work proposes and implements an early diagnostic and management (EDM) system to assist the neurologist in type identification (5-classes) of the seizure activity at run time and also features an interactive graphical user interface (GUI). In the GUI, temporal, spectral (along with source localization) and spatial plots can be viewed along with the seizure data classified based on its types. The system utilizes a discrete wavelet transform (DWT) and k-nearest neighbour, (KNN) based on feature extraction and classification, respectively. The system is validated using 31 patients' recordings from Temple University Hospital (TUH) EEG Database. Our system achieves a 5-class classification accuracy, sensitivity and specificity of 97.7%, 92.9%, and 98.7%, respectively, for patient-wise cross-validation.
Zain Taufique, Bingzhao Zhu, Gianluca Coppola, Mahsa Shoaran, Wala Saadeh, and Muhammad Awais Bin Altaf
IEEE
Chronic Neurological Disorders (CNDs) such as epilepsy [1], [2], migraine [3], and autism [4] can be persistent for extensive periods. Untreated CNDs may lead to perpetual debilities. Therefore, it is crucial to diagnose them at an early stage to perform a timely, meaningful intervention. A routine medical checkup often cannot provide the timely mediation required for CNDs. A chronic attack consists of pre-ictal, ictal, and post-ictal stages, while an effective intervention necessitates CND detection and remedial response during the pre-ictal stage. Therefore, monitoring CNDs 24/7 is crucial, irrespective of patient’s location and clinical state. The electroencephalogram (EEG) is utilized for monitoring and detection of most CNDs in a wearable environment [1]–[6].
Zain Taufique, Anil Kanduri, Muhammad Awais Bin Altaf, and Pasi Liljeberg
IEEE
Epilepsy is a pervasive disorder that causes abrupt seizure attacks. This paper presents an FPGA-based logic implementation that detects impending seizure attacks using the Electroencephalogram (EEG) data-set of epileptic patients. The feature extraction is done using a 2-dimensional Fast Fourier Transform hardware architecture, and the classification is done using a software-based Artificial Neural Network (ANN) classifier. This implementation is presented in two different models, i.e., an accurate model and an approximate model. The accurate model requires more operating power but provides highly accurate results. In comparison, the approximate model provides slightly lesser accurate results but consumes significantly lesser electrical power. The Application- and scenario-based trade-offs between these models are compared against the available energy resources in the device battery. The proposed solution achieved 80.83 % and 97.96 % sensitivity and specificity, respectively, against 218.95mW power using the accurate feature extraction. In contrast, 77.95 % and 95 % sensitivity and specificity were achieved at 173.32mW power requirements for the approximate model. There is a 21 % power saving in the approximate model with nearly 3% performance loss. The overall design was synthesized at 20MHz operating frequency and provided a complete 256-point FFT result in 650 µs.
Abdul Rehman Aslam and Muhammad Awais Bin Altaf
Elsevier
Abdul Rehman Aslam and Muhammad Awais Bin Altaf
Institute of Electrical and Electronics Engineers (IEEE)
Chronic neurological disorders (CND's) are lifelong diseases and cannot be eradicated, but their severe effects can be alleviated by early preemptive measures. CND's, such as Alzheimer's, Autism Spectrum Disorder (ASD), and Amyotrophic Lateral Sclerosis (ALS), are the chronic ailment of the central nervous system that causes the degradation of emotional and cognitive abilities. Long term continuous monitoring with neuro-feedback of human emotions for patients with CND's is crucial in mitigating its harmful effect. This paper presents hardware efficient and dedicated human emotion classification processor for CND's. Scalp EEG is used for the emotion's classification using the valence and arousal scales. A linear support vector machine classifier is used with power spectral density, logarithmic interhemispheric power spectral ratio, and the interhemispheric power spectral difference of eight EEG channel locations suitable for a wearable non-invasive classification system. A look-up-table based logarithmic division unit (LDU) is to represent the division features in machine learning (ML) applications. The implemented LDU minimizes the cost of integer division by 34% for ML applications. The implemented emotion's classification processor achieved an accuracy of 72.96% and 73.14%, respectively, for the valence and arousal classification on multiple publicly available datasets. The 2 x 3mm2 processor is fabricated using a 0.18 μm 1P6M CMOS process with power and energy utilization of 2.04 mW and 16 μJ/classification, respectively, for 8-channel operation.
Hafiz Talha Iqbal, Bilal Majeed, Uzair Khan, and Muhammad Awais Bin Altaf
IEEE
Breast cancer is the leading type of cancer among women in the 3rd world countries with < 50% survival rate. However, its early diagnosis can lead to cost-effective and successful treatment. Traditional breast cancer screening tools like mammography and MRI are not readily available to the population in low-income countries. Thermography (infrared imaging) is an FDA approved adjunct screening tool which can be an alternative solution. We present here the architecture of thermography based, application-specific Digital Back End (DBE) processor for a handheld off the shelf portable and intelligent screening device. A thermal image of the thorax taken by an infrared camera is pre-processed to get the regions of interest. To achieve efficient hardware implementation texture features are carefully selected, which are then fed to a dual classifier based on trained Linear Support Vector Machine (LSVM) and convolutional neural network (CNN) to decide the decision boundary. The proposed system achieves an overall sensitivity and specificity of 90.06% and 91.8%, respectively, with efficient hardware implementation by exploiting proposed classifier.
Arish Adil, Hassan Abid, Nadir Najib, Usama Jillani, Wala Saadeh, and Muhammad Awais Bin Altaf
IEEE
Continuous tracking of heart activity from Electrocardiogram (ECG) signal to assist physicians in treating Cardio Vascular Disease (CVD) is essential for patients to avoid complications and problems associated with Cardiac arrhythmia conditions. But it necessitates a patch form factor and energy-efficient design for long-term battery operation. This paper presents a novel non-invasive single- lead wearable ECG monitoring system for acquiring the ECG signals along with the proposed simplified processor for arrhythmia detection. To ensure correct functional verification, the proposed system is implemented in hardware and tested using the MIT-BIH ECG arrhythmia database and the Creighton University Ventricular Tachyarrhythmia Database. It achieves a classification accuracy of more than 90%. The proposed processor is also synthesized using 0.18um CMOS technology.
Wala Saadeh, Fatima Hameed Khan, and Muhammad Awais Bin Altaf
Institute of Electrical and Electronics Engineers (IEEE)
Accurate monitoring of the depth of anesthesia (DoA) is essential for intraoperative and postoperative patient's health. Commercially available electroencephalograph (EEG)-based DoA monitors are recommended only for certain anesthetic drugs and specific age-group patients. This paper presents a machine learning classification processor for accurate DoA estimation irrespective of the patient's age and anesthetic drug. The classification is solely based on six features extracted from EEG signal, i.e., spectral edge frequency (SEF), beta ratio, and four bands of spectral energy (FBSE). A machine learning fine decision tree classifier is adopted to achieve a four-class DoA classification (deep, moderate, and light DoA versus awake state). The feature selection and the classification processor are optimized to achieve the highest classification accuracy for the state of moderate anesthesia required for the surgical operations. The proposed 256-point fast Fourier transform accelerator is implemented to realize SEF, beta ratio, and FBSE that enables minimal latency and high accuracy feature extraction. The proposed DoA processor is implemented using a 65 nm CMOS technology and experimentally verified using field programming gate array (FPGA) based on the EEG recordings of 75 patients undergoing elective surgery with different types of anesthetic agents. The processor achieves an average accuracy of 92.2% for all DoA states, with a latency of 1s The 0.09 mm2 DoA processor consumes 140nJ/classification.