Muhammad Abdul Kadir
@du.ac.bd
Associate Professor, Department of Biomedical Physics and Technology, Faculty of Science
University of Dhaka
RESEARCH INTERESTS
Biomedical instrumentation and medical applications of electrical bioimpedance techniques, Biomedical signal & image analysis and machine learning techniques for disease diagnosis.
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Md. Amimul Ihsan, Abrar Faiaz Eram, Lutfun Nahar, and Muhammad Abdul Kadir
Institute of Electrical and Electronics Engineers (IEEE)
Along with day-to-day communication, receiving medical care is quite challenging for the hearing impaired and mute population, especially in developing countries where medical facilities are not as modernized as in the West. A word-level sign language interpretation system that is aimed toward detecting medically relevant signs can allow smooth communication between doctors and hearing impaired patients, ensuring seamless medical care. To that end, a dataset from twenty distinct signers of diverse backgrounds performing 30 frequently used words in patient-doctor interaction was created. The proposed system has been built employing MobileNetV2 in conjunction with an attention-based Bidirectional LSTM network to achieve robust classification, where the validation accuracy and f1- scores were 95.83% and 93%, respectively. Notably, the accuracy of the proposed model surpasses the recent word-level sign language classification method in a medical context by 5%. Furthermore, the comparison of evaluation metrics with contemporary word-level sign language recognition models in American, Arabic, and German Sign Language further affirmed the capability of the proposed architecture.
Md. Mozzammel Haque, Muhammad Abdul Kadir, and Richard Bayford
AIP Publishing
Biological systems are often exposed to mechanical perturbations, which may modulate many biochemical processes. Ligand binding involves a wide range of structural changes in the receptor protein, from hinge movement of entire domains to minor sidechain rearrangements in the binding pocket residues. Hydrophobic ligand binding to protein alters the system’s vibrational free energy, allowing different conformational states of allosteric proteins. Excess hydrophobicity in protein–ligand binding generates mechanical force along the peptide backbone through the hydrophobic effect. We describe mechanically strained peptide structures involved in protein aggregation to determine the transition between the initial condensation of hydrophobic polypeptide chains into ordered fibrillar structures. This transition is due to the excess attractive hydrophobic force by ligand binding within proteins into fibrillar assemblies. The process of fibrillar formation has a mechanosensitive nature, which significantly influences the pathogenesis of several neurodegenerative diseases.
Kanchon Kanti Podder, Maymouna Ezeddin, Muhammad E. H. Chowdhury, Md. Shaheenur Islam Sumon, Anas M. Tahir, Mohamed Arselene Ayari, Proma Dutta, Amith Khandakar, Zaid Bin Mahbub, and Muhammad Abdul Kadir
MDPI AG
Every one of us has a unique manner of communicating to explore the world, and such communication helps to interpret life. Sign language is the popular language of communication for hearing and speech-disabled people. When a sign language user interacts with a non-sign language user, it becomes difficult for a signer to express themselves to another person. A sign language recognition system can help a signer to interpret the sign of a non-sign language user. This study presents a sign language recognition system that is capable of recognizing Arabic Sign Language from recorded RGB videos. To achieve this, two datasets were considered, such as (1) the raw dataset and (2) the face–hand region-based segmented dataset produced from the raw dataset. Moreover, operational layer-based multi-layer perceptron “SelfMLP” is proposed in this study to build CNN-LSTM-SelfMLP models for Arabic Sign Language recognition. MobileNetV2 and ResNet18-based CNN backbones and three SelfMLPs were used to construct six different models of CNN-LSTM-SelfMLP architecture for performance comparison of Arabic Sign Language recognition. This study examined the signer-independent mode to deal with real-time application circumstances. As a result, MobileNetV2-LSTM-SelfMLP on the segmented dataset achieved the best accuracy of 87.69% with 88.57% precision, 87.69% recall, 87.72% F1 score, and 99.75% specificity. Overall, face–hand region-based segmentation and SelfMLP-infused MobileNetV2-LSTM-SelfMLP surpassed the previous findings on Arabic Sign Language recognition by 10.970% accuracy.
Sadeque Reza Khan, Xiaohan Wang, Tiantao Jiang, Wei Ju, Norbert Radacsi, Muhammad Abdul Kadir, Khondkar Siddique-e Rabbani, Steve Cunningham, and Srinjoy Mitra
MDPI AG
Accurate assessment of Respiratory Rate (RR) is the most important mechanism in detecting pneumonia in low-resource settings. Pneumonia is a disease with one of the highest mortality rates among young children under five. However, the diagnosis of pneumonia for infants remains challenging, especially in low- and middle-income countries (LMIC). In such situations, RR is most often measured manually with visual inspection. Accurate RR measurement requires the child to remain calm without any stress for a few minutes. The difficulty in achieving this with a sick child in a clinical environment can result in errors and misdiagnosis, even more so when the child is crying and non-cooperating around unfamiliar adults. Therefore, we propose an automated novel RR monitoring device built with textile glove and dry electrodes which can make use of the relaxed posture when the child is resting on the carer’s lap. This portable system is non-invasive and made with affordable instrumentation integrated on customized textile glove. The glove has multi-modal automated RR detection mechanism that simultaneously uses bio-impedance and accelerometer data. This novel textile glove with dry electrodes can easily be worn by a parent/carer and is washable. The real-time display on a mobile app shows the raw data and the RR value, allowing a healthcare professional to monitor the results from afar. The prototype device has been tested on 10 volunteers with age variation of 3 years to 33 years, including male and female. The maximum variation of measured RR with the proposed system is ±2 compared to the traditional manual counting method. It does not create any discomfort for either the child or the carer and can be used up to 60 to 70 sessions/day before recharging.
Arafat Rahman, Muhammad E.H. Chowdhury, Amith Khandakar, Anas M. Tahir, Nabil Ibtehaz, Md Shafayet Hossain, Serkan Kiranyaz, Junaid Malik, Haya Monawwar, and Muhammad Abdul Kadir
Elsevier BV
Mahjabin Mobarak, Muhammad Abdul Kadir, and K Siddique-e Rabbani
Walter de Gruyter GmbH
Abstract Probing deep regions of the lung using electrical impedance is very important considering the need for a low cost and simple technique, particularly for the low and medium income countries. Because of complexity and cost, Electrical Impedance Tomography is not suitable for this envisaged application. The simple Tetrapolar Impedance Measurement (TPIM) technique employing four electrodes is the age old technique for bioelectrical measurements. However, it has its limitations in respect of organ localisation and depth sensitivity using skin surface electrodes. Recently, a new 6-electrode TPIM with two current electrodes but two pairs of appropriately connected potential electrodes positioned on the front and back of the thorax, proposed by one of the authors, came with a promise. However, this work gave a qualitative proposal based on concepts of physics and lacked a quantitative evaluation. In order to evaluate the method quantitatively, the present work employed finite element method based COMSOL Multiphysics software and carried out simulation studies using this new 6-electrode TPIM and compared the results with those from 4-electrode TPIM, with electrodes applied either on the front or at the back of the thorax for the latter. Initially, it carried out a sensitivity distribution study using a simple rectangular volume conductor which showed that the 6-electrode TPIM gives better depth sensitivity throughout the lung region. Next it used a near life like thorax model developed by another of the authors earlier. Using this model, extensive studies were carried out to quantify the overall sensitivity over a target lung region, the contribution of the target lung to the total measured impedance, and several other parameters. Through these studies, the 6-electrode TPIM was established on a stronger footing for probing deep regions of the lungs.
Kanchon Kanti Podder, Muhammad E. H. Chowdhury, Anas M. Tahir, Zaid Bin Mahbub, Amith Khandakar, Md Shafayet Hossain, and Muhammad Abdul Kadir
MDPI AG
A real-time Bangla Sign Language interpreter can enable more than 200 k hearing and speech-impaired people to the mainstream workforce in Bangladesh. Bangla Sign Language (BdSL) recognition and detection is a challenging topic in computer vision and deep learning research because sign language recognition accuracy may vary on the skin tone, hand orientation, and background. This research has used deep machine learning models for accurate and reliable BdSL Alphabets and Numerals using two well-suited and robust datasets. The dataset prepared in this study comprises of the largest image database for BdSL Alphabets and Numerals in order to reduce inter-class similarity while dealing with diverse image data, which comprises various backgrounds and skin tones. The papers compared classification with and without background images to determine the best working model for BdSL Alphabets and Numerals interpretation. The CNN model trained with the images that had a background was found to be more effective than without background. The hand detection portion in the segmentation approach must be more accurate in the hand detection process to boost the overall accuracy in the sign recognition. It was found that ResNet18 performed best with 99.99% accuracy, precision, F1 score, sensitivity, and 100% specificity, which outperforms the works in the literature for BdSL Alphabets and Numerals recognition. This dataset is made publicly available for researchers to support and encourage further research on Bangla Sign Language Interpretation so that the hearing and speech-impaired individuals can benefit from this research.
Ferdous Karim Lucy, Khadiza Tun Suha, Sumaiya Tabassum Dipty, Md Sharjis Ibne Wadud, and Muhammad Abdul Kadir
IOP Publishing
Abstract Objective. Pneumonia is the single largest cause of death in children worldwide due to infectious diseases. According to WHO guidelines, fast breathing and chest indrawing are the key indicators of pneumonia in children requiring antibiotic treatments. The aim of this study was to develop a video based novel method for simultaneous monitoring of respiratory rate and chest indrawing without upsetting babies. Approach. Respiratory signals, corresponding to periodic movements of chest-abdominal walls during breathing, were extracted by analyzing RGB (red, green, blue) components in video frames captured by a smartphone camera. Respiratory rate was then obtained by applying fast Fourier transform on the de-noised respiratory signal. Chest indrawing was detected by analysing relative phases of regional chest-abdominal wall mobility. The performance of the developed algorithm was evaluated on both healthy and pneumonia children. Main results. The proposed method can measure respiratory rate with an overall mean absolute error of 1.8 bpm in the range 18–105 bpm. Phase difference between regional chest wall movements in the chest indrawing (pneumonia) cases was found to be 143 ± 23.9 degrees, which was significantly higher than that in the healthy cases 52.3 ± 32.6 degrees (p < 0.001). Significance. Being non-intrusive and non-subjective, this computer-aided method can be useful in the monitoring for respiratory rate and chest indrawing for the diagnosis of pneumonia and its severity in children.
Bikash Kumar Bhawmick, Muhammad Abdul Kadir, and Khondkar Siddique-e Rabbani
IEEE
Bioelectrical impedance based imaging has emerged as a non-ionizing technique for physiological monitoring in the human body. Pigeon Hole Imaging (PHI) is an impedance based novel imaging modality that maps the spatial distribution of electrical impedance within a volume conductor. In PHI, impedance is measured at multiple locations placing an array of electrodes on the surface of a volume conductor and an image of the localized impedance distribution is reconstructed. Recent simulation studies demonstrated the use of PHI in the detection of subcutaneous vein and similar applications. However, the instrumentation of a PHI system is in need for experimental measurements to investigate its utility in practice. The aim of the present work is to design and develop the necessary hardware and software including switching algorithm and data acquisition peripherals for a PHI measurement system. A constant current source capable of driving alternating current of amplitude 0.5 mA at 10kHz in biological tissues was designed. An amplifier circuit minimizing common mode noise was also designed for measurement of the pickup voltage. An algorithm for switching current drive and voltage sensing electrodes for impedance measurement on a 5×5 PHI matrix is designed using multiplexers. Experimental measurements were performed on a cubic phantom containing 5×5 PHI electrode array. Software was developed in Matlab platform for acquiring data and visual presentation of the image reconstructed from PHI measurements. The implemented PHI system may be useful in the imaging of impedance distribution of human body for disease diagnosis and monitoring.
Arafat Rahman, Muhammad E. H. Chowdhury, Amith Khandakar, Serkan Kiranyaz, Kh Shahriya Zaman, Mamun Bin Ibne Reaz, Mohammad Tariqul Islam, Maymouna Ezeddin, and Muhammad Abdul Kadir
Institute of Electrical and Electronics Engineers (IEEE)
Electroencephalography (EEG) based biometric systems are gaining attention for their anti-spoofing capability but lack accuracy due to signal variability at different psychological and physiological conditions. On the other hand, keystroke dynamics-based systems achieve very high accuracy but have low anti-spoofing capability. To address these issues, a novel multimodal biometric system combining EEG and keystroke dynamics is proposed in this paper. A dataset was created by acquiring both keystroke dynamics and EEG signals simultaneously from 10 users. Each user participated in 500 trials at 10 different sessions (days) to replicate real-life signal variability. A machine learning classification pipeline is developed using multi-domain feature extraction (time, frequency, time-frequency), feature selection (Gini impurity), classifier design, and score level fusion. Different classifiers were trained, validated, and tested for two different classification experiments – personalized and generalized. For identification and authentication, 99.9% and 99.6% accuracies are achieved, respectively for the Random Forest classifier in 5 fold cross-validation. These results outperform the individual modalities with a significant margin (~5%). We also developed a binary template matching-based algorithm, which gives 93.64% accuracy 6X faster. The proposed method can be considered secure and reliable for any kind of biometric identification and authentication.
Tawsifur Rahman, Muhammad E. H. Chowdhury, Amith Khandakar, Khandaker R. Islam, Khandaker F. Islam, Zaid B. Mahbub, Muhammad A. Kadir, and Saad Kashem
MDPI AG
Pneumonia is a life-threatening disease, which occurs in the lungs caused by either bacterial or viral infection. It can be life-endangering if not acted upon at the right time and thus the early diagnosis of pneumonia is vital. The paper aims to automatically detect bacterial and viral pneumonia using digital x-ray images. It provides a detailed report on advances in accurate detection of pneumonia and then presents the methodology adopted by the authors. Four different pre-trained deep Convolutional Neural Network (CNN): AlexNet, ResNet18, DenseNet201, and SqueezeNet were used for transfer learning. A total of 5247 chest X-ray images consisting of bacterial, viral, and normal chest x-rays images were preprocessed and trained for the transfer learning-based classification task. In this study, the authors have reported three schemes of classifications: normal vs. pneumonia, bacterial vs. viral pneumonia, and normal, bacterial, and viral pneumonia. The classification accuracy of normal and pneumonia images, bacterial and viral pneumonia images, and normal, bacterial, and viral pneumonia were 98%, 95%, and 93.3%, respectively. This is the highest accuracy, in any scheme, of the accuracies reported in the literature. Therefore, the proposed study can be useful in more quickly diagnosing pneumonia by the radiologist and can help in the fast airport screening of pneumonia patients.
Tawsifur Rahman, Amith Khandakar, Muhammad Abdul Kadir, Khandaker Rejaul Islam, Khandakar F. Islam, Rashid Mazhar, Tahir Hamid, Mohammad Tariqul Islam, Saad Kashem, Zaid Bin Mahbub,et al.
Institute of Electrical and Electronics Engineers (IEEE)
Tuberculosis (TB) is a chronic lung disease that occurs due to bacterial infection and is one of the top 10 leading causes of death. Accurate and early detection of TB is very important, otherwise, it could be life-threatening. In this work, we have detected TB reliably from the chest X-ray images using image pre-processing, data augmentation, image segmentation, and deep-learning classification techniques. Several public databases were used to create a database of 3500 TB infected and 3500 normal chest X-ray images for this study. Nine different deep CNNs (ResNet18, ResNet50, ResNet101, ChexNet, InceptionV3, Vgg19, DenseNet201, SqueezeNet, and MobileNet) were used for transfer learning from their pre-trained initial weights and were trained, validated and tested for classifying TB and non-TB normal cases. Three different experiments were carried out in this work: segmentation of X-ray images using two different U-net models, classification using X-ray images and that using segmented lung images. The accuracy, precision, sensitivity, F1-score and specificity of best performing model, ChexNet in the detection of tuberculosis using X-ray images were 96.47%, 96.62%, 96.47%, 96.47%, and 96.51% respectively. However, classification using segmented lung images outperformed that with whole X-ray images; the accuracy, precision, sensitivity, F1-score and specificity of DenseNet201 were 98.6%, 98.57%, 98.56%, 98.56%, and 98.54% respectively for the segmented lung images. The paper also used a visualization technique to confirm that CNN learns dominantly from the segmented lung regions that resulted in higher detection accuracy. The proposed method with state-of-the-art performance can be useful in the computer-aided faster diagnosis of tuberculosis.
Muhammad E. H. Chowdhury, Tawsifur Rahman, Amith Khandakar, Rashid Mazhar, Muhammad Abdul Kadir, Zaid Bin Mahbub, Khandakar Reajul Islam, Muhammad Salman Khan, Atif Iqbal, Nasser Al Emadi,et al.
Institute of Electrical and Electronics Engineers (IEEE)
Coronavirus disease (COVID-19) is a pandemic disease, which has already caused thousands of causalities and infected several millions of people worldwide. Any technological tool enabling rapid screening of the COVID-19 infection with high accuracy can be crucially helpful to the healthcare professionals. The main clinical tool currently in use for the diagnosis of COVID-19 is the Reverse transcription polymerase chain reaction (RT-PCR), which is expensive, less-sensitive and requires specialized medical personnel. X-ray imaging is an easily accessible tool that can be an excellent alternative in the COVID-19 diagnosis. This research was taken to investigate the utility of artificial intelligence (AI) in the rapid and accurate detection of COVID-19 from chest X-ray images. The aim of this paper is to propose a robust technique for automatic detection of COVID-19 pneumonia from digital chest X-ray images applying pre-trained deep-learning algorithms while maximizing the detection accuracy. A public database was created by the authors combining several public databases and also by collecting images from recently published articles. The database contains a mixture of 423 COVID-19, 1485 viral pneumonia, and 1579 normal chest X-ray images. Transfer learning technique was used with the help of image augmentation to train and validate several pre-trained deep Convolutional Neural Networks (CNNs). The networks were trained to classify two different schemes: i) normal and COVID-19 pneumonia; ii) normal, viral and COVID-19 pneumonia with and without image augmentation. The classification accuracy, precision, sensitivity, and specificity for both the schemes were 99.7%, 99.7%, 99.7% and 99.55% and 97.9%, 97.95%, 97.9%, and 98.8%, respectively. The high accuracy of this computer-aided diagnostic tool can significantly improve the speed and accuracy of COVID-19 diagnosis. This would be extremely useful in this pandemic where disease burden and need for preventive measures are at odds with available resources.
Rashida Haque, Muhammad Abdul Kadir, and K Siddique-e Rabbani
Walter de Gruyter GmbH
Abstract For probing deep organs of the body using electrical impedance, the conventional method is to use Electrical Impedance Tomography (EIT). However, this would be a sophisticated machine and will be very expensive when a full 3D EIT is developed in the future. Furthermore, for most low income countries such expensive devices may not deliver the benefits to a large number of people. Therefore, this paper suggests the use of simpler techniques like Tetrapolar Impedance Measurement (TPIM) or Focused Impedance Method (FIM) in probing deeper organs. Following a method suggested earlier by one of the authors, this paper studies the possibility of using TPIM and FIM for the stomach. Using a simplified model of the human trunk with an embedded stomach, a finite element simulation package, COMSOL, was used to obtain transfer impedance values and percentage contribution of the stomach region in the total impedance. For this work, judicious placement of electrodes through qualitative visualizations based on point sensitivity equations and equipotential concepts were made, which showed that reasonable contribution of the stomach region is possible through the use of TPIM and FIM. The contributions were a little over 20% which is of similar order of the cross-sectional area percentage of the stomach with respect to that of the trunk. For the case where the conductivity of the stomach region was assumed about 4 times higher, the contributions increased to about 38%. Through further studies this proposed methods may contribute greatly in the study of deeper organs of the body.
Shamor Kanti Roy, Mohammad Abu Sayem Karal, Muhammad Abdul Kadir, and Khondkar Siddique-e Rabbani
Springer Science and Business Media LLC
Muhammad Abdul Kadir and K. Siddique-e Rabbani
Walter de Gruyter GmbH
Abstract Tetra-polar electrical impedance measurement (TPIM) with a square geometry of electrodes is useful in the characterization of epithelial tissues, especially in the detection of cervical cancer at precancerous stages. However, in TPIM, the peak planar sensitivity just below the electrode surface is almost zero and increases to a peak value at a depth of about one third to one half of the electrode separation. To get high sensitivity for the epithelial layer, having thicknesses of 200 μm to 300 μm, the electrode separation needed is less than 1 mm, which is difficult to achieve in practical probes. This work proposes a conical conducting layer in front of a pencil like probe with a square geometry of TPIM electrodes to create virtual electrodes with much smaller separation at the body surface, thus increasing the sensitivity of the epithelial tissues. To understand the improvements, if any, 3D sensitivity distribution and transfer impedance were simulated using COMSOL Multiphysics software for a simplified body tissue model containing a 300 μm epithelial layer. It has been shown that fractional contribution of an epithelial layer can be increased several times placing a cylindrical conducting layer in between the tissue surface and the electrodes, which can further be enhanced using a conical conducting layer. The results presented in this paper can be used to choose an appropriate electrode separation, conducting layer height and cone parameters for enhanced sensitivity in the epithelial layer.
Rushdi Zahid Rusho and M Abdul Kadir
IEEE
Detection and localization of subcutaneous vein is important in health care for intravenous drug administration and blood drawing procedures. This paper presents a simple and non-ionizing method based on Pigeon Hole Imaging (PHI) for mapping subcutaneous vein patterns. In PHI, localized impedance values on multiple adjacent positions are measured placing an electrode array on the surface of a volume conductor and the conductivity distribution beneath the electrode array is obtained. In this study, the impedance measurements were simulated in an FEM based software COMSOL Multiphysics. An FEM model of a human forearm with different domains including skin, muscle, subcutaneous layer and veins was created. The conductivity of blood is higher compared to that of its surroundings which is the basis of localization of subcutaneous vein in the current study. It is shown that the proposed method can detect and map the pathways of subcutaneous veins fairly accurately using 5×5 matrix of PHI electrode array. However, the resolution of the image needs to be increased and the method needs to be evaluated in human subjects to justify its usefulness in practice.
Rushdi Zahid Rusho and M Abdul Kadir
IEEE
Pigeon Hole Imaging (PHI) is a non-ionizing imaging modality that maps spatial distribution of electrical conductivity within a volume conductor. In PHI, electrical impedance is measured placing an electrode array on the surface of a volume conductor and the conductivity distribution in a plane parallel to the electrode array is obtained. This paper describes an image reconstruction algorithm for PHI based on 4-electrode Focused Impedance Measurements (FIM). The impedance measurements were simulated in a finite element based simulation software COMSOL Multiphysics for 5×5 and 6×6 PHI matrices. The measured impedance on the volume conductor surface just above an object of different conductivity compared to the volume conductor was found to be different than that in the neighbouring regions which is the basis of the image reconstruction algorithm. The developed algorithm was tested by modelling volume conductors containing multiple objects of different conductivities at different depths. It was found that PHI has 3D sensitivity and the developed algorithm can fairly detect objects with diameter at least of the order of half the electrode separation with depth equal to or less than electrode separation. Being a low cost and non-ionizing method, there is a potential of using PHI in the imaging of organs, tumors and blood vessels in human body although the image resolution is low.
Muhammad Towhidur Rahman, Muhammad Abdul Kadir, A.H.M Zadidul Karim, and Md Abdullah Al Mahmud
IEEE
Electrocardiogram (ECG or EKG) is an image of the electrical activity of heart which is formed as line tracings on paper. ECG helps the trained physicians to diagnose the status of the heart. The objective of this work is to utilize cardio generated voltages for respiratory study. This work analyses real-time ECG formed from standard lead systems to extract respiration rate of lungs. The job will be performed with the ECG dataset available at physionet.org, an internet database bank. An algorithm in (Matrix Laboratory) MATLAB environment is developed for extraction of the exact amplitudes of the R-waves and furthermore, these R-wave amplitudes are used to form pulsatile waves due to respiration. The respiration wave is then used to evaluate the respiration rate of the heart. This matches completely with the respiration information of the same subject/patient. Thus this analysis will help the person who has already gone through the ECG test to know about his respiration rate without going any further diagnosis.
M. Abdul Kadir, Tanvir Noor Baig, and K. Siddique-e Rabbani
National Taiwan University
Conventional spirometry gives information on the overall ventilation of a person's lung; it cannot detect localized disorders in ventilation as occurring in pulmonary edema, pneumonia, tumor, TB, etc. Here we propose a new technique involving the recently developed focused impedance method (FIM) in combination with conventional spirometry to detect localized lung ventilation disorders. Electrical impedance of lung tissue changes as a function of air content and FIM provides a measurement of localized electrical impedance with sensitivity down to reasonable depths inside the body using a few surface electrodes; here we used a six-electrode version. At least four quadrants of the lungs in the frontal plane can be separately measured using a hand-held probe with spring backed skin surface electrodes. Firstly, spatial sensitivity distribution of the six-electrode FIM was obtained using finite element simulation which verified the focusing effect and its depth sensitivity. Percent change in impedance between maximum inspiration and expiration were measured at four quadrants of the chest of a healthy male subject giving four different values; that at the lower right quadrant was found to be the maximum, as also expected based on anatomy. Changes in impedance at this quadrant of the same subject were found to vary proportionately with exhaled air volumes, measured using a bellows-type spirometer. Similar FIM measurements at lower right lung of seven healthy subjects were found to be almost proportional (R2= 0.7) to the total exhaled air volumes (vital capacity). This was the basis of the new technique. For a healthy individual, the ratio of the local impedance change to vital capacity (VC) will fall within a certain range for each of the four lung quadrants. A lower value at any quadrant would indicate disorder within that quadrant, while a larger value would indicate disorder in a region outside the particular quadrant. The FIM electrode probe can then be moved to take measurements at the other quadrants to locate the region of disorder. This preliminary study indicates that FIM in combination with conventional spirometry could be used to detect localized ventilation defects.
Abdullah Al Amin, Shahnaj Parvin, M A Kadir, Tasmia Tahmid, S Kaisar Alam, and K Siddique-e Rabbani
IOP Publishing
When a breast lump is detected through palpation, mammography or ultrasonography, the final test for characterization of the tumour, whether it is malignant or benign, is biopsy. This is invasive and carries hazards associated with any surgical procedures. The present work was undertaken to study the feasibility for such characterization using non-invasive electrical impedance measurements and machine learning techniques. Because of changes in cell morphology of malignant and benign tumours, changes are expected in impedance at a fixed frequency, and versus frequency of measurement. Tetrapolar impedance measurement (TPIM) using four electrodes at the corners of a square region of sides 4 cm was used for zone localization. Data of impedance in two orthogonal directions, measured at 5 and 200 kHz from 19 subjects, and their respective slopes with frequency were subjected to machine learning procedures through the use of feature plots. These patients had single or multiple tumours of various types in one or both breasts, and four of them had malignant tumours, as diagnosed by core biopsy. Although size and depth of the tumours are expected to affect the measurements, this preliminary work ignored these effects. Selecting 12 features from the above measurements, feature plots were drawn for the 19 patients, which displayed considerable overlap between malignant and benign cases. However, based on observed qualitative trend of the measured values, when all the feature values were divided by respective ages, the two types of tumours separated out reasonably well. Using K-NN classification method the results obtained are, positive prediction value: 60%, negative prediction value: 93%, sensitivity: 75%, specificity: 87% and efficacy: 84%, which are very good for such a test on a small sample size. Study on a larger sample is expected to give confidence in this technique, and further improvement of the technique may have the ability to replace biopsy.
M.A. Kadir, K.S. Rabbani, and A.J. Wilson
Institution of Engineering and Technology
Making diagnostic measurements available to the widely geographically spread populations in developing countries is a challenge which is unlikely to be met by the technology used in the west. Multi-frequency electrical impedance techniques are attractive in this context as the instrumentation is relatively simple and research has demonstrated application in cardiac and respiratory medicine and in the characterisation of epithelial tissues. The development of focused impedance measurement (FIM) techniques has potentially gone some way to overcome the complex spatial sensitivity distribution which has been one of the limiting factors in using electrical impedance techniques in diagnosis. This paper describes a simple low cost system based on FIM that can be maintained and repaired in the field by the researchers allowing the potential for electrical impedance based diagnostic technique in developing countries to be evaluated. The microcontroller based multi-frequency system for FIM uses synchronous demodulation which is implemented using IC components readily available in developing countries. The drive current was generated from a microcontroller at 16 different frequencies. A Howland V to I converter delivers current to the tissue. The peak value of the measured voltage signal is determined by a micro-processor controlled analogue synchronous peak detector. Measurements on resistive and reactive phantoms gave a resolution that would allow impedance changes reported in clinical studies (e.g. respiration, epithelial tissue characterization and abdominal fat thickness) to be measured with the system. (6 pages)
M Abdul Kadir, Humayra Ferdous, Tanvir Noor Baig, and K Siddique-e-Rabbani
IOP Publishing
Focused Impedance Method (FIM) provides an opportunity for localized impedance measurement down to reasonable depths within the body using surface electrodes, and has a potential application in localized lung ventilation study. This however needs assessment of normal values for healthy individuals. In this study, localized ventilation maps in terms of electrical impedance in a matrix formation around the thorax, both from the front and the back, were obtained from two normal male subjects using a modified configuration of FIM. For this the focused impedance values at full inspiration and full expiration were measured and the percentage difference with respect to the latter was used. Some of the measured values would have artefacts due to movements of the heart and the diaphragm in the relevant anatomical positions which needs to be considered with due care in any interpretation.