Publications

Study summaries and my contributions across six research areas.

Publication list

The selected papers include recent work and widely cited studies.

61 publications

Best subject quartile and Impact Factor · 2025 unless noted

Wearables & biomechanics

Smart insoles, gait modelling, and estimation of forces and motion.

2026Online 2025

Journal articleFirst author

3D Foot Kinetics Estimation From Distributed VGRF From Smart Insoles via 1D Domain Transformation

Sakib Mahmud, Muhammad E. H. Chowdhury, Faycal Bensaali

Reconstructed three-dimensional ground reaction forces, moments, and centre of pressure from distributed insole pressures. The study used instrumented-treadmill references and participant-held-out evaluation.

My contribution

I led preprocessing, development and evaluation of Ke2KeNet, and manuscript preparation. I also analysed which plantar regions carried the most useful information.

IEEE Journal of Biomedical and Health InformaticsQ1Impact Factor 7.7 · 30(3) · 2656-2667

2024

Journal article

Laser-Induced Graphene-Based Smart Insole to Measure Plantar Temperature

Amith Khandakar, Md. Ahasan Atick Faisal, Muhammad E. H. Chowdhury, Mamun Bin Ibne Reaz, Sawal Hamid Md Ali, Mohd Ibrahim bin Shapiai Abd. Razak, Ahmad Ashrif A. Bakar, Sakib Mahmud, Rayaz A. Malik

Used laser-induced graphene sensors in a smart insole to map plantar temperature in real time. Experiments characterised sensitivity, operating range, and response and recovery times.

My contribution

I helped assess the sensing system experimentally, interpreted the results, and reviewed the manuscript.

IEEE Sensors JournalQ1Impact Factor 4.5 · 24(2) · 1190-1199

2024Online 2023

Journal article

Robust and novel attention guided MultiResUnet model for 3D ground reaction force and moment prediction from foot kinematics

Md. Ahasan Atick Faisal, Sakib Mahmud, Muhammad E. H. Chowdhury, Amith Khandakar, Mosabber Uddin Ahmed, Abdulrahman Alqahtani, Mohammed Alhatou

Estimated three-dimensional ground reaction forces and moments from five shoe-mounted motion markers. The study evaluated the attention-guided model across two public gait datasets.

My contribution

I contributed to methodology, model development, cross-dataset evaluation, and manuscript writing.

Neural Computing and ApplicationsQ1 · 36(3) · 1105-1121

2023

Journal article

NDDNet: a deep learning model for predicting neurodegenerative diseases from gait pattern

Md. Ahasan Atick Faisal, Muhammad E. H. Chowdhury, Zaid Bin Mahbub, Shona Pedersen, Mosabber Uddin Ahmed, Amith Khandakar, Mohammed Alhatou, Mohammad Nabil, Iffat Ara, Enamul Haque Bhuiyan, Sakib Mahmud, Mohammed AbdulMoniem

Combined ground reaction force waveforms and gait-cycle features to distinguish healthy gait from gait patterns associated with Parkinson’s disease, Huntington’s disease, and ALS in a 64-participant dataset.

My contribution

I contributed to gait data analysis and model evaluation, interpreted findings, and reviewed and revised the manuscript.

Applied IntelligenceQ2Impact Factor 3.5 · 53(17) · 20034-20046

2023

Journal articleFirst author

Fiber Bragg Gratings based smart insole to measure plantar pressure and temperature

Sakib Mahmud, Amith Khandakar, Muhammad E.H. Chowdhury, Mohammed AbdulMoniem, Mamun Bin Ibne Reaz, Zaid Bin Mahbub, Kishor Kumar Sadasivuni, M. Murugappan, Mohammed Alhatou

Built a fibre Bragg grating insole to measure plantar pressure and temperature simultaneously. Static and gait experiments evaluated the prototype and its real-time pressure and temperature maps.

My contribution

I led system development and evaluation, data collection and analysis, validation, and manuscript preparation.

Sensors and Actuators A: PhysicalQ1Impact Factor 5.1 · 350 · Article 114092

2022

Journal article

Design and Implementation of a Smart Insole System to Measure Plantar Pressure and Temperature

Amith Khandakar, Sakib Mahmud, Muhammad E. H. Chowdhury, Mamun Bin Ibne Reaz, Serkan Kiranyaz, Zaid Bin Mahbub, Sawal Hamid Ali, Ahmad Ashrif A. Bakar, Mohamed Arselene Ayari, Mohammed Alhatou, Mohammed Abdul-Moniem, Md Ahasan Atick Faisal

Built a wireless insole prototype with 16 pressure sensors and eight temperature sensors per insole. The system displayed plantar pressure and temperature maps in real time.

My contribution

I contributed to methodology, experimental investigation, and manuscript review.

SensorsQ1Impact Factor 4.0 · 22(19) · 7599

2022

Journal article

An investigation to study the effects of Tai Chi on human gait dynamics using classical machine learning

Md. Ahasan Atick Faisal, Muhammad E.H. Chowdhury, Amith Khandakar, Md Shafayet Hossain, Mohammed Alhatou, Sakib Mahmud, Iffat Ara, Shah Imran Sheikh, Mosabber Uddin Ahmed

Analysed gait data from a longitudinal Tai Chi study of 57 participants. The analysis used scaled gait features and subject-wise validation to assess group differences during walking tasks.

My contribution

I contributed to gait-data analysis, classical machine-learning evaluation, interpretation, and manuscript review.

Computers in Biology and MedicineQ1 · 142 · Article 105184

Medical imaging

Renal CT and ultrasound, organ segmentation, and chest imaging.

2026

Journal articleFirst author

Enhancing kidney stone segmentation in computed tomography images through successive scope reduction and intra-framework feature transfer

Sakib Mahmud, Muhammad E.H. Chowdhury, Tariq O. Abbas, Faycal Bensaali

Developed a CT pipeline that narrows the search to kidney and stone slices before segmentation. On the study cohort, Dice was 97.44% with scope reduction and 90.13% when that step was omitted.

My contribution

I designed the scope-reduction and feature-transfer strategy and led model development, evaluation, analysis, and writing.

Biomedical Signal Processing and ControlQ1Impact Factor 5.7 · 119(Part B) · Article 109894

2024

Journal articleFirst author

Automated grading of prenatal hydronephrosis severity from segmented kidney ultrasounds using deep learning

Sakib Mahmud, Tariq O. Abbas, Muhammad E.H. Chowdhury, Adam Mushtak, Saidul Kabir, Sreekumar Muthiyal, Alaa Koko, Ahmed Balla Abdalla Altyeb, Abdulrahman Alqahtani, Amith Khandakar, Sheikh Mohammed Shariful Islam

Combined kidney localisation, segmentation, and multi-view ultrasound classification to estimate prenatal hydronephrosis severity. The study evaluated grading at the subject level for a curated cohort of 163 subjects.

My contribution

I led ultrasound data curation, developed the segmentation and grading models, and drafted the manuscript.

Expert Systems with ApplicationsQ1Impact Factor 9.4 · 255 · Article 124594

2023

Journal article

Deep Learning Framework for Liver Segmentation from T1-Weighted MRI Images

Md. Sakib Abrar Hossain, Sidra Gul, Muhammad E. H. Chowdhury, Muhammad Salman Khan, Md. Shaheenur Islam Sumon, Enamul Haque Bhuiyan, Amith Khandakar, Maqsud Hossain, Abdus Sadique, Israa Al-Hashimi, Mohamed Arselene Ayari, Sakib Mahmud, Abdulrahman Alqahtani

Benchmarked 24 architectures for liver segmentation in T1-weighted MRI. The study also evaluated a cascaded segmentation approach and described a workflow for cloud-based model updates.

My contribution

I contributed to software implementation, data curation, original drafting, and manuscript revision.

SensorsQ1Impact Factor 4.0 · 23(21) · 8890

2023

Journal articleFirst author

Kidney Cancer Diagnosis and Surgery Selection by Machine Learning from CT Scans Combined with Clinical Metadata

Sakib Mahmud, Tariq O. Abbas, Adam Mushtak, Johayra Prithula, Muhammad E. H. Chowdhury

Combined CT-derived features with clinical records to classify renal tumours and estimate the surgical approach in the study cohort. Feature analysis identified tumour volume and cancer stage as informative.

My contribution

I contributed to study design, methodology, software, data curation, validation, visualisation, analysis, and manuscript drafting as part of my MSc research.

CancersQ1Impact Factor 4.8 · 15(12) · 3189

2023

Journal article

Automated measurement of penile curvature using deep learning-based novel quantification method

Sriman Bidhan Baray, Mohamed Abdelmoniem, Sakib Mahmud, Saidul Kabir, Md. Ahasan Atick Faisal, Muhammad E. H. Chowdhury, Tariq O. Abbas

Estimated penile curvature from images of nine 3D models. The pipeline localised and segmented the models and derived the curvature angle from their masks.

My contribution

I assisted with the experiments and helped prepare and revise the manuscript.

Frontiers in PediatricsQ1Impact Factor 2.2 · 11 · Article 1149318

2022

Journal article

HipXNet: Deep Learning Approaches to Detect Aseptic Loosening of Hip Implants Using X-Ray Images

Tawsifur Rahman, Amith Khandakar, Khandaker Reajul Islam, Md Mohiuddin Soliman, Mohammad Tariqul Islam, Ahmed Elsayed, Yazan Qiblawey, Sakib Mahmud, Ashiqur Rahman, Farayi Musharavati, Erfan Zalnezhad, Muhammad E. H. Chowdhury

Localised hip implants on X-rays and assessed aseptic loosening with a classification pipeline. The authors released the annotated radiography dataset used in the study.

My contribution

I contributed to pipeline development and evaluation, data analysis, interpretation, and manuscript writing and review.

IEEE AccessQ1Impact Factor 4.2 · 10 · 53359-53373

2022

Conference paper

Classification of Viral, Bacterial, And Covid-19 Pneumonia using Deep Learning Framework from Chest X-Ray Images

Muhammad E. H. Chowdhury, Tawsifur Rahman, Amith Khandakar, Sakib Mahmud

Benchmarked 21 pretrained convolutional encoders on chest radiographs to distinguish COVID-19, viral pneumonia, bacterial pneumonia, and normal images.

My contribution

I contributed to model evaluation and analysis, and to manuscript preparation and review.

Signal, Image Processing and Embedded Systems Trends · 12(20) · 1-21 · Zurich, Switzerland

2021

Journal article

COVID-19 infection localization and severity grading from chest X-ray images

Anas M. Tahir, Muhammad E.H. Chowdhury, Amith Khandakar, Tawsifur Rahman, Yazan Qiblawey, Uzair Khurshid, Serkan Kiranyaz, Nabil Ibtehaz, M. Sohel Rahman, Somaya Al-Maadeed, Sakib Mahmud, Maymouna Ezeddin, Khaled Hameed, Tahir Hamid

Segmented lungs and infection regions in chest radiographs to localise COVID-19 pneumonia and estimate the extent of involvement for severity grading.

My contribution

I helped curate the benchmark dataset and contributed to dataset and model evaluation, interpretation, and manuscript review.

Computers in Biology and MedicineQ1 · 139 · Article 105002

2021

Journal article

Detection and Severity Classification of COVID-19 in CT Images Using Deep Learning

Yazan Qiblawey, Anas Tahir, Muhammad E. H. Chowdhury, Amith Khandakar, Serkan Kiranyaz, Tawsifur Rahman, Nabil Ibtehaz, Sakib Mahmud, Somaya Al Maadeed, Farayi Musharavati, Mohamed Arselene Ayari

Segmented lungs and infection regions in chest CT scans, then used the extent of involvement to classify COVID-19 severity.

My contribution

I contributed to investigation, data curation, validation, and manuscript review and editing.

DiagnosticsQ2Impact Factor 3.8 · 11(5) · 893

Biosignal AI & health monitoring

Machine learning for cardiovascular, neural, and metabolic signals from clinical, wearable, and contactless sensors.

2025

Journal article

Deep learning-based beat-to-beat arterial blood pressure estimation using distant radar signals

Farhana Ahmed Chowdhury, Md Kamal Hosain, Md Shafayet Hossain, Muhammad E. H. Chowdhury, Sakib Mahmud, Muhammad Ashad Kabir, Abdulrahman Alqahtani, Anwarul Hasan

Evaluated whether distant radar could recover continuous beat-to-beat arterial blood pressure waveforms from raw in-phase and quadrature signals across recording scenarios.

My contribution

I proposed investigating one-dimensional signal reconstruction for this task and contributed to model implementation, evaluation, interpretation, and manuscript review.

Neural Computing and ApplicationsQ1 · 37(21) · 16677-16702

2025

Journal article

A novel 1D generative adversarial network-based framework for atrial fibrillation detection using restored wrist photoplethysmography signals

Faizul Rakib Sayem, Mosabber Uddin Ahmed, Saadia Binte Alam, Sakib Mahmud, Md. Mamun Sheikh, Abdulrahman Alqahtani, Md Ahasan Atick Faisal, Muhammad E.H. Chowdhury

Restored wrist photoplethysmography (PPG) with a one-dimensional CycleGAN and tested whether restoration improved atrial fibrillation detection across two test splits.

My contribution

I adapted CycleGAN to one-dimensional physiological signals and contributed to study design, formal analysis, and writing.

Biomedical Signal Processing and ControlQ1Impact Factor 5.7 · 101 · Article 107233

2024

Journal article

QU-GM: An IoT Based Glucose Monitoring System From Photoplethysmography, Blood Pressure, and Demographic Data Using Machine Learning

Md Nazmul Islam Shuzan, Moajjem Hossain Chowdhury, Muhammad E. H. Chowdhury, Khalid Abualsaud, Elias Yaacoub, Md Ahasan Atick Faisal, Mazun Alshahwani, Noora Al Bordeni, Fatima Al-Kaabi, Sara Al-Mohannadi, Sakib Mahmud, Nizar Zorba

Combined PPG, blood pressure, and demographic data to estimate glucose and classify diabetes severity in a 139-participant study. The regression evaluation included Clarke Error Grid analysis.

My contribution

I contributed to machine-learning and physiological-signal analysis, technical evaluation, and manuscript review.

IEEE AccessQ1Impact Factor 4.2 · 12 · 77774-77790

2024

Journal article

ECG waveform generation from radar signals: A deep learning perspective

Farhana Ahmed Chowdhury, Md Kamal Hosain, Md Sakib Bin Islam, Md Shafayet Hossain, Promit Basak, Sakib Mahmud, M. Murugappan, Muhammad E.H. Chowdhury

Reconstructed ECG waveforms from contactless radar with a multiresolution network, evaluating performance during rest, the Valsalva manoeuvre, and apnoea.

My contribution

I contributed to methodology, model development, quantitative evaluation, and original drafting.

Computers in Biology and MedicineQ1 · 176 · Article 108555

2024

Journal articleFirst author

Restoration of magnetohydrodynamic-corrupted 12-lead electrocardiogram to enhance cardiac monitoring during magnetic resonance imaging

Sakib Mahmud, Muhammad E.H. Chowdhury, Moajjem Hossain Chowdhury, Abdulrahman Alqahtani, Zaid Bin Mahbub, Faycal Bensaali, Serkan Kiranyaz

Evaluated two methods for recovering 12-lead ECG and R-peak timing from magnetohydrodynamic distortion in MRI at 1, 3, and 7 tesla.

My contribution

I led study design, data curation, model and software development, experiments, analysis, and manuscript preparation.

Engineering Applications of Artificial IntelligenceQ1Impact Factor 9.0 · 133 · Article 108483

2024

Journal articleFirst author

Wearable wrist to finger photoplethysmogram translation through restoration using super operational neural networks based 1D-CycleGAN for enhancing cardiovascular monitoring

Sakib Mahmud, Muhammad E.H. Chowdhury, Serkan Kiranyaz, Malisha Islam Tapotee, Purnata Saha, Anas M. Tahir, Amith Khandakar, Abdulrahman Alqahtani

Restored noisy wrist PPG and translated it towards finger PPG morphology. The study assessed whether translation improved pulse rate and pulse rate variability estimates against ECG-derived references.

My contribution

I led development of the two-stage Super-ONN CycleGAN workflow, ran the experiments, and prepared the analysis and manuscript.

Expert Systems with ApplicationsQ1Impact Factor 9.4 · 246 · Article 123167

2024

Journal article

M2ECG: Wearable Mechanocardiograms to Electrocardiogram Estimation Using Deep Learning

Malisha Islam Tapotee, Purnata Saha, Sakib Mahmud, Abdulrahman Alqahtani, Muhammad E. H. Chowdhury

Reconstructed ECG from wearable measurements of cardiac vibration and rotation. The study evaluated waveform similarity and whether heart rate could be recovered on held-out participants.

My contribution

I contributed to methodology, data analysis, model evaluation, interpretation, and manuscript review.

IEEE AccessQ1Impact Factor 4.2 · 12 · 12963-12975

2024

Journal articleFirst author

Restoration of motion-corrupted EEG signals using attention-guided operational CycleGAN

Sakib Mahmud, Muhammad E.H. Chowdhury, Serkan Kiranyaz, Nasser Al Emadi, Anas M. Tahir, Md Shafayet Hossain, Amith Khandakar, Somaya Al-Maadeed

Used an attention-guided operational CycleGAN to restore motion-corrupted EEG. Evaluation on held-out participants examined signal recovery in the time and frequency domains.

My contribution

I led study design, development of AGO-CycleGAN, experiments, analysis, visualisation, and writing.

Engineering Applications of Artificial IntelligenceQ1Impact Factor 9.0 · 128 · Article 107514

2024

Journal article

LGI-rPPG-Net: A shallow encoder-decoder model for rPPG signal estimation from facial video streams

Moajjem Hossain Chowdhury, Muhammad E.H. Chowdhury, Mamun Bin Ibne Reaz, Sawal Hamid Md Ali, Seyed Mehdi Rakhtala, M. Murugappan, Sakib Mahmud, Nazmul Islam Shuzan, Ahmad Ashrif A. Bakar, Mohd Ibrahim Bin Shapiai, Muhammad Salman Khan, Amith Khandakar

Estimated contactless PPG waveforms from facial video using a shallow encoder-decoder. The evaluation compared both waveform fidelity and agreement of derived heart rates.

My contribution

I prepared the deep-learning pipeline, helped process the video and physiological data, and contributed to evaluation and manuscript review.

Biomedical Signal Processing and ControlQ1Impact Factor 5.7 · 89 · Article 105687

2024

Journal article

A novel deep learning technique for morphology preserved fetal ECG extraction from mother ECG using 1D-CycleGAN

Promit Basak, A.H.M Nazmus Sakib, Muhammad E.H. Chowdhury, Nasser Al-Emadi, Huseyin Cagatay Yalcin, Shona Pedersen, Sakib Mahmud, Serkan Kiranyaz, Somaya Al-Maadeed

Used a one-dimensional CycleGAN to recover full fetal ECG morphology from maternal recordings. Two PhysioNet datasets were used to evaluate waveform recovery, fetal QRS detection, and fetal heart rate.

My contribution

I contributed to the deep-learning methodology, model assessment, interpretation, and manuscript review.

Expert Systems with ApplicationsQ1Impact Factor 9.4 · 235 · Article 121196

2023

Journal article

A Deep Learning Framework for the Detection of Abnormality in Cerebral Blood Flow Velocity Using Transcranial Doppler Ultrasound

Naima Nasrin Nisha, Kanchon Kanti Podder, Muhammad E. H. Chowdhury, Mamun Rabbani, Md. Sharjis Ibne Wadud, Somaya Al-Maadeed, Sakib Mahmud, Amith Khandakar, Susu M. Zughaier

Classified abnormal middle cerebral artery blood-flow velocity from transcranial Doppler ultrasound waveforms using an attention-based residual network.

My contribution

I helped develop the MATLAB software, preprocess the data, analyse and visualise the results, and write and revise the paper.

DiagnosticsQ2Impact Factor 3.8 · 13(12) · 2000

2023

Journal article

Fetal ECG extraction from maternal ECG using deeply supervised LinkNet++ model

Arafat Rahman, Sakib Mahmud, Muhammad E.H. Chowdhury, Huseyin Cagatay Yalcin, Amith Khandakar, Onur Mutlu, Zaid Bin Mahbub, Reema Yousef Kamal, Shona Pedersen

Used a deeply supervised LinkNet++ to recover fetal ECG from maternal recordings. Leave-one-subject-out evaluation on two datasets assessed waveform fidelity and fetal QRS timing.

My contribution

I contributed to the LinkNet++ methodology, model development, experiments, validation, analysis, and manuscript writing.

Engineering Applications of Artificial IntelligenceQ1Impact Factor 9.0 · 123 · Article 106414

2023

Journal article

MultiResUNet3+: A Full-Scale Connected Multi-Residual UNet Model to Denoise Electrooculogram and Electromyogram Artifacts from Corrupted Electroencephalogram Signals

Md Shafayet Hossain, Sakib Mahmud, Amith Khandakar, Nasser Al-Emadi, Farhana Ahmed Chowdhury, Zaid Bin Mahbub, Mamun Bin Ibne Reaz, Muhammad E. H. Chowdhury

Used a multiscale residual U-Net to remove eye-movement and muscle artefacts from EEG. Comparisons assessed artefact reduction in both the time and frequency domains.

My contribution

I contributed to study design, methodology, data curation, formal analysis, and manuscript drafting and revision.

BioengineeringQ2Impact Factor 4.4 · 10(5) · 579

2023

Journal article

PCovNet+: A CNN-VAE anomaly detection framework with LSTM embeddings for smartwatch-based COVID-19 detection

Farhan Fuad Abir, Muhammad E.H. Chowdhury, Malisha Islam Tapotee, Adam Mushtak, Amith Khandakar, Sakib Mahmud, Anwarul Hasan

Used a convolutional variational autoencoder with LSTM features to identify unusual resting heart rate patterns in smartwatch records. The study examined anomalies before and after symptom onset in a cohort of 68 people with COVID-19.

My contribution

I contributed to the software, method and experiment analysis, interpretation, and manuscript review and editing.

Engineering Applications of Artificial IntelligenceQ1Impact Factor 9.0 · 122 · Article 106130

2023

Journal article

Self-Attention MHDNet: A Novel Deep Learning Model for the Detection of R-Peaks in the Electrocardiogram Signals Corrupted with Magnetohydrodynamic Effect

Moajjem Hossain Chowdhury, Muhammad E. H. Chowdhury, Muhammad Salman Khan, Md Asad Ullah, Sakib Mahmud, Amith Khandakar, Alvee Hassan, Anas M. Tahir, Anwarul Hasan

Detected ECG R peaks under magnetohydrodynamic distortion from MRI. The study evaluated an attention-based detector on 3- and 7-tesla recordings.

My contribution

I contributed to methodology, software development, validation, and manuscript drafting and revision.

BioengineeringQ2Impact Factor 4.4 · 10(5) · 542

2023

Journal article

Deep learning-based middle cerebral artery blood flow abnormality detection using flow velocity waveform derived from transcranial Doppler ultrasound

Kanchon Kanti Podder, Muhammad E.H. Chowdhury, Somaya Al-Maadeed, Naima Nasrin Nisha, Sakib Mahmud, Fatema Hamadelneil, Taif Almkhlef, Hind Aljofairi, Adam Mushtak, Amith Khandakar, Susu Zughaier

Combined generative waveform enhancement with a two-stream classifier to distinguish abnormal middle cerebral artery flow from transcranial Doppler recordings.

My contribution

I contributed to data curation, analysis of flow velocity signals, and original drafting.

Biomedical Signal Processing and ControlQ1Impact Factor 5.7 · 85 · Article 104882

2023

Journal articleFirst author

NABNet: A Nested Attention-guided BiConvLSTM network for a robust prediction of Blood Pressure components from reconstructed Arterial Blood Pressure waveforms using PPG and ECG signals

Sakib Mahmud, Nabil Ibtehaz, Amith Khandakar, M. Sohel Rahman, Antonio JR. Gonzales, Tawsifur Rahman, Md Shafayet Hossain, Md. Sakib Abrar Hossain, Md. Ahasan Atick Faisal, Farhan Fuad Abir, Farayi Musharavati, Muhammad E. H. Chowdhury

Reconstructed arterial blood pressure from PPG and ECG, then combined waveform and pressure-component estimates to recover a waveform in physical units. The evaluation examined systolic and diastolic pressure errors.

My contribution

I led development of NABNet, data preparation, experiments, analysis, visualisation, and writing.

Biomedical Signal Processing and ControlQ1Impact Factor 5.7 · 79 · Article 104247

2023Online 2022

Journal articleFirst author

MLMRS-Net: Electroencephalography (EEG) motion artifacts removal using a multi-layer multi-resolution spatially pooled 1D signal reconstruction network

Sakib Mahmud, Md Shafayet Hossain, Muhammad E. H. Chowdhury, Mamun Bin Ibne Reaz

Removed motion artefacts from single-channel EEG with a multiresolution reconstruction network. Leave-one-subject-out evaluation assessed signal recovery on held-out participants.

My contribution

I led development of MLMRS-Net, preprocessing, implementation, experiments, quantitative analysis, and manuscript preparation.

Neural Computing and ApplicationsQ1 · 35(11) · 8371-8388

2022

Journal article

PPG2ABP: Translating Photoplethysmogram (PPG) Signals to Arterial Blood Pressure (ABP) Waveforms

Nabil Ibtehaz, Sakib Mahmud, Muhammad E. H. Chowdhury, Amith Khandakar, Muhammad Salman Khan, Mohamed Arselene Ayari, Anas M. Tahir, M. Sohel Rahman

Translated PPG into arterial blood pressure waveforms with two cascaded networks. The evaluation measured waveform error and errors in derived systolic, diastolic, and mean arterial pressure.

My contribution

I contributed to data curation, methodology, validation, analysis, and drafting. I also worked with the first author to revise the study for publication.

BioengineeringQ2Impact Factor 4.4 · 9(11) · 692

2022

Journal article

Lightweight End-to-End Deep Learning Solution for Estimating the Respiration Rate from Photoplethysmogram Signal

Moajjem Hossain Chowdhury, Md Nazmul Islam Shuzan, Muhammad E. H. Chowdhury, Mamun Bin Ibne Reaz, Sakib Mahmud, Nasser Al Emadi, Mohamed Arselene Ayari, Sawal Hamid Md Ali, Ahmad Ashrif A. Bakar, Syed Mahfuzur Rahman, Amith Khandakar

Estimated respiration rate directly from PPG with a compact ConvMixer network. The study evaluated two public datasets and tested limited fine-tuning across datasets.

My contribution

I contributed to data curation, formal analysis and methodology, and to drafting the manuscript.

BioengineeringQ2Impact Factor 4.4 · 9(10) · 558

2022

Journal article

PCovNet: A presymptomatic COVID-19 detection framework using deep learning model using wearables data

Farhan Fuad Abir, Khalid Alyafei, Muhammad E.H. Chowdhury, Amith Khandakar, Rashid Ahmed, Muhammad Maqsud Hossain, Sakib Mahmud, Ashiqur Rahman, Tareq O. Abbas, Susu M. Zughaier, Khalid Kamal Naji

Modelled resting heart rate with an LSTM variational autoencoder to identify anomalies in wearable data around COVID-19 symptom onset. The study examined how its evaluation setup affected the balance between precision and recall.

My contribution

I contributed to wearable-data analysis, model evaluation, interpretation, and manuscript review.

Computers in Biology and MedicineQ1 · 147 · Article 105682

2022

Journal article

QUCoughScope: An Intelligent Application to Detect COVID-19 Patients Using Cough and Breath Sounds

Tawsifur Rahman, Nabil Ibtehaz, Amith Khandakar, Md Sakib Abrar Hossain, Yosra Magdi Salih Mekki, Maymouna Ezeddin, Enamul Haque Bhuiyan, Mohamed Arselene Ayari, Anas Tahir, Yazan Qiblawey, Sakib Mahmud, Susu M. Zughaier, Tariq Abbas, Somaya Al-Maadeed, Muhammad E. H. Chowdhury

Compared cough and breath audio for classifying COVID-19 in symptomatic and asymptomatic groups and integrated the classification workflow into an application prototype.

My contribution

I contributed to the investigation and methodology, including technical analysis and validation of the audio pipeline, and to drafting the manuscript.

DiagnosticsQ2Impact Factor 3.8 · 12(4) · 920

2022

Journal articleFirst author

A Shallow U-Net Architecture for Reliably Predicting Blood Pressure (BP) from Photoplethysmogram (PPG) and Electrocardiogram (ECG) Signals

Sakib Mahmud, Nabil Ibtehaz, Amith Khandakar, Anas M. Tahir, Tawsifur Rahman, Khandaker Reajul Islam, Md Shafayet Hossain, M. Sohel Rahman, Farayi Musharavati, Mohamed Arselene Ayari, Mohammad Tariqul Islam, Muhammad E. H. Chowdhury

Estimated blood pressure from PPG and ECG with a shallow one-dimensional U-Net. Evaluation used records from 942 MIMIC-II patients and an external dataset of 40 participants.

My contribution

I led study design, data curation, methodology, implementation, validation, analysis, visualisation, and writing.

SensorsQ1Impact Factor 4.0 · 22(3) · 919

2022

Conference paper

Deep Learning Technique to Denoise Electromyogram Artifacts from Single-Channel Electroencephalogram Signals

Muhammad E. H. Chowdhury, Md Shafayet Hossain, Sakib Mahmud, Amith Khandakar

Compared four one-dimensional convolutional models for removing muscle artefacts from single-channel EEG using EEGdenoiseNet. The evaluation examined recovery in the time and frequency domains and preservation of relative EEG band power.

My contribution

I contributed to the methodology, experiments, analysis, and manuscript preparation.

Signal, Image Processing and Embedded Systems Trends · 12(20) · 85-98 · Zurich, Switzerland

Raman spectroscopy & biosensing

Microbial sensing, cross-instrument alignment, and validation of Raman models.

2027Online 2026

ReviewFirst author

Machine learning-enabled label-free SERS for microbial sensing: Toward robust, generalizable, and deployable workflows

Sakib Mahmud, Md. Sakib Bin Islam, Mansura Naznine, Md. Shaheenur Islam Sumon, Susu M. Zughaier, Yiping Zhao, Muhammad E.H. Chowdhury

Reviewed 129 studies on label-free surface-enhanced Raman spectroscopy (SERS) for microbial sensing. The review traced six stages from sample preparation to validation and found external testing in only 6.2% of studies.

My contribution

I designed the review, led the evidence synthesis and validation audit, and prepared the figures, tables, and manuscript.

TalantaQ1Impact Factor 6.7 · 311 · Article 130241

2026

Journal article

SERS on analyte-enriched blood for rapid, culture-free sepsis recognition and causative pathogen identification with super operational neural networks

Manal Hassan, Md Sakib Bin Islam, Sakib Mahmud, Mahmoud Elgamal, Md. Shaheenur Islam Sumon, Ali Ait Hssain, Emad Ibrahim, Amit Kumar, Yiping Zhao, Saad Bin Abul Kashem, Mehmet Burcin Unlu, Gozde Durmus, Muhammad E.H. Chowdhury, Susu M. Zughaier

Used analyte-enriched blood spectra and super-operational neural networks to recognise sepsis and identify pathogens. The study included an independent external cohort of 70 samples.

My contribution

I developed the workflow from raw spectra through preprocessing, split design, model training, evaluation, and interpretation. I also helped document the experimental protocol.

TalantaQ1Impact Factor 6.7 · 301 · Article 129332

2026

Journal articleFirst author

Deep learning-based cross-device standardization of surface-enhanced Raman spectroscopy for enhanced bacterial recognition

Sakib Mahmud, Faizul Rakib Sayem, Manal Hassan, Yanjun Yang, Muhammad E.H. Chowdhury, Susu M. Zughaier, Faycal Bensaali, Yiping Zhao

Mapped SERS spectra from four portable instruments to a laboratory reference and evaluated bacterial classification after alignment. The study also examined how many reference spectra were needed for adaptation.

My contribution

I led development and evaluation of the spectral alignment and classification models and prepared the analysis, figures, and manuscript.

Spectrochimica Acta Part A: Molecular and Biomolecular SpectroscopyQ1Impact Factor 4.8 · 347 · Article 126931

Biomedical devices & simulation

Implants, simulation, ventilation, and computational haemodynamics.

2025

Journal article

Rapid wall shear stress prediction for aortic aneurysms using deep learning: a fast alternative to CFD

Md. Ahasan Atick Faisal, Onur Mutlu, Sakib Mahmud, Anas Tahir, Muhammad E. H. Chowdhury, Faycal Bensaali, Abdulrahman Alnabti, Mehmet Metin Yavuz, Ayman El-Menyar, Hassan Al-Thani, Huseyin Cagatay Yalcin

Predicted wall shear stress in abdominal aortic aneurysms by representing three-dimensional geometries from multiple views. The study compared model outputs with computational fluid dynamics estimates on real and synthetic geometries.

My contribution

I contributed to model implementation and integration, iterative evaluation, and troubleshooting with the modelling team.

Medical & Biological Engineering & ComputingQ2Impact Factor 3.1 · 63(7) · 2173-2190

2024

Journal article

Stress transfer, stress shielding, and micro-movement analysis in the implanted femoral bone region with various hip implant stem materials: A finite element approach

Md Mohiuddin Soliman, Sakib Mahmud, Muhammad E.H. Chowdhury, Touhidul Alam, Abdulmajeed M. Alenezi, Norbahiah Misran, Mohamed S. Soliman, Mohammad Tariqul Islam

Compared six hip-stem materials under walking, standing, and sitting loads using finite element simulations. The study examined stress transfer, stress shielding, and implant micromovement.

My contribution

I reconstructed the Summit stem in CAD, helped prepare the biomechanical simulations, and contributed to interpretation and writing.

Materials Today CommunicationsQ1Impact Factor 4.9 · 41 · Article 110932

2024

Journal article

ExoventQ: A Novel Low-Cost Portable Negative Pressure Ventilator Design and Implementation

Shahd Sameer Mohammed Gaben, Nada Abughanam, Nabil Ibtehaz, Sakib Mahmud, Ghada Al Noaimi, Abdulrahman Alqahtani, Amith Khandakar, Muhammad E. H. Chowdhury

Built and tested a portable negative-pressure ventilation prototype with continuous and cyclic operation, closed-loop pressure control, and emergency safety mechanisms.

My contribution

I supported engineering analysis and prototype evaluation, interpreted test results, and helped write and revise the paper.

IEEE AccessQ1Impact Factor 4.2 · 12 · 61828-61847

2023

Review

Advancement in total hip implant: a comprehensive review of mechanics and performance parameters across diverse novelties

Md Mohiuddin Soliman, Mohammad Tariqul Islam, Muhammad E. H. Chowdhury, Abdulrahman Alqahtani, Farayi Musharavati, Touhidul Alam, Ahmed S. Alshammari, Norbahiah Misran, Mohamed S. Soliman, Sakib Mahmud, Amith Khandakar

Reviewed how total hip implant materials and geometries affect stress shielding, micromotion, wear, and loosening. The synthesis examined trade-offs in design and gaps in biomechanical evaluation.

My contribution

I helped synthesise the literature on implant mechanics, finite element modelling, and performance measures and contributed to interpretation and manuscript revision.

Journal of Materials Chemistry BQ1Impact Factor 6.2 · 11(44) · 10507-10537

2023

Journal article

Design and Performance Evaluation of a Novel Spiral Head-Stem Trunnion for Hip Implants Using Finite Element Analysis

Md Mohiuddin Soliman, Muhammad E. H. Chowdhury, Mohammad Tariqul Islam, Farayi Musharavati, Sakib Mahmud, Muhammad Hafizh, Mohamed Arselene Ayari, Amith Khandakar, Mohammad Kaosar Alam, Erfan Zal Nezhad

Simulated a spiral trunnion at the head-stem interface of a hip implant. The study compared predicted wear, sliding distance, and fixation stability with a conventional design.

My contribution

I designed the spiral geometry in Fusion 360 and contributed to the finite element methodology, implementation, and analysis.

MaterialsQ2Impact Factor 3.7 · 16(4) · 1466

2023

Journal article

Surface Acoustic Wave (SAW) Sensors for Hip Implant: A Numerical and Computational Feasibility Investigation Using Finite Element Methods

Muhammad Hafizh, Md Soliman, Yazan Qiblawey, Muhammad Chowdhury, Mohammad Islam, Farayi Musharavati, Sakib Mahmud, Amith Khandakar, Mohammad Nabil, Erfan Nezhad

Modelled a surface acoustic wave sensor for detecting hip implant loosening. Finite element simulations compared voltage response and stress at two implant locations under three loading conditions.

My contribution

I contributed to simulation software, investigation, and formal analysis.

BiosensorsQ1Impact Factor 6.2 · 13(1) · 79

2022

Review

A Review of Biomaterials and Associated Performance Metrics Analysis in Pre-Clinical Finite Element Model and in Implementation Stages for Total Hip Implant System

Md Mohiuddin Soliman, Muhammad E. H. Chowdhury, Mohammad Tariqul Islam, Farayi Musharavati, Mohammad Nabil, Muhammad Hafizh, Amith Khandakar, Sakib Mahmud, Erfan Zal Nezhad, Md Nazmul Islam Shuzan, Farhan Fuad Abir

Reviewed biomaterial selection and preclinical finite element evaluation for total hip implants, covering the effects of stiffness mismatch, stress shielding, wear, fatigue, and micromotion.

My contribution

I contributed to the computational and technical analysis and helped design and prepare figures for the review.

PolymersQ1Impact Factor 5.8 · 14(20) · 4308

2019

Conference abstract

Towards next generation cannulation simulators

Abdulrahman Mahmoud, Uzair Khurshid, Aiman Abducarim, Sakib Mahmud, Omrane Abdallah, Elshaikh Mohamed, Abdullah Alsalemi, Faycal Bensaali, Abbes Amira, Ali Ait Hssain, Guillaume Alinier, Ibrahim Hassan

Developed an extracorporeal membrane oxygenation (ECMO) cannulation trainer with ultrasound-compatible pads and a closed-loop circulation model. The design incorporated sensing for simulated procedural errors.

My contribution

I contributed to prototype design and implementation, experimental and technical work, and preparation of the documentation and manuscript.

Qatar Medical JournalQ3 · 2019(2) · Article 61

2019

Conference paper

Preliminary Implementation of the Next Generation Cannulation Simulator

Abdulrahman Mahmoud, Uzair Khurshid, Aiman Abducarim, Sakib Mahmud, Omrane Abdallah, Elshaikh Mohamed, Abdullah Alsalemi, Faycal Bensaali, Abbes Amira, Ali Ait Hssain, Guillaume Alinier, Ibrahim Hassan

Presented a preliminary ECMO cannulation trainer that could reproduce complications such as low-pressure flow and guidewire or cannula mispositioning.

My contribution

I contributed to prototype design, implementation, and testing, as well as technical documentation and manuscript preparation.

2019 IEEE Student Conference on Research and Development (SCOReD) · 243-247 · Bandar Seri Iskandar, Malaysia

2018

Journal article

Towards the design and implementation of a human circulatory system for Extracorporeal Membrane Oxygenation simulation

Uzair Khurshid, Abdulrahman Mahmoud, Aiman Abducarim, Sakib Mahmud, Omrane Abdallah, Elshaikh Mohamed, Faycal Bensaali, Abbes Amira, Abdullah Alsalemi, Ali Ait Hssain, Guillaume Alinier, Ibrahim Hassan

Developed and tested a simulated human circulation system for ECMO training with pulsatile flow to support cannulation practice.

My contribution

I helped design, build, and test the prototype. I developed the Arduino control code for the pump motor and contributed to technical documentation and writing.

The Egyptian Journal of Critical Care MedicineQ4 (2019)Impact Factor 0.2 · 6(3) · 87-89

Energy & applied machine learning

Smart buildings, energy systems, and machine learning for other applied problems.

2026

ReviewFirst author

A comprehensive review of deep reinforcement learning applications from centralized power generation to modern energy internet frameworks

Sakib Mahmud, Aya Nabil Sayed, Yassine Himeur, Armstrong Nhlabatsi, Faycal Bensaali

Reviewed deep reinforcement learning across generation, storage, grid operations, and energy markets. The review mapped algorithms to operating problems and examined barriers to deployment, including safety and transfer from simulation.

My contribution

I led the review design, literature synthesis, taxonomy, comparative analysis, visualisations, and writing.

Renewable and Sustainable Energy ReviewsQ1Impact Factor 18.0 · 226 · Article 116481

2025

Journal articleFirst author

A cascaded deep learning framework for simultaneous non-intrusive load and occupancy monitoring using multi-channel aggregated smart meter data

Sakib Mahmud, Mahdi Houchati, Muhammad E.H. Chowdhury, Faycal Bensaali

Paired appliance-level load disaggregation with occupancy detection from multichannel smart-meter data. A cascaded model evaluated how individual device events could help infer occupancy on the ECO dataset.

My contribution

I led study design, data curation, software development, validation, analysis, visualisation, and writing.

Journal of Building EngineeringQ1Impact Factor 8.1 · 112 · Article 113731

2025

Journal article

Optimizing energy efficiency through precise occupancy detection: A tailored CNN architecture for smart buildings and beyond

Aya Nabil Sayed, Sakib Mahmud, Faycal Bensaali, Muhammad E. H. Chowdhury, Yassine Himeur

Encoded environmental sensor time series as images to estimate indoor occupancy. The study evaluated a tailored convolutional network across three benchmark datasets.

My contribution

I redesigned the occupancy classifier and contributed to the experimental setup, analysis, and manuscript.

Neural Computing and ApplicationsQ1 · 37(21) · 16487-16503

2025

Journal articleFirst author

Multimodal feature fusion and ensemble learning for non-intrusive occupancy monitoring using smart meters

Sakib Mahmud, Faycal Bensaali, Muhammad E. H. Chowdhury, Mahdi Houchati

Combined appliance-event, statistical, and time-of-day features to infer occupancy from smart meters. The study evaluated feature fusion, ensemble classifiers, and device selection on the ECO dataset.

My contribution

I led conceptualisation, data curation, methodology, software development, analysis, visualisation, and original drafting.

Building and EnvironmentQ1Impact Factor 8.4 · 271 · Article 112635

2024

Journal article

GearFaultNet: Novel Network for Automatic and Early Detection of Gearbox Faults

Proma Dutta, Kanchon Kanti Podder, Md. Shaheenur Islam Sumon, Muhammad E. H. Chowdhury, Amith Khandakar, Nasser Al-Emadi, Moajjem Hossain Chowdhury, M. Murugappan, Mohamed Arselene Ayari, Sakib Mahmud, S. M. Muyeen

Used four-channel vibration data to identify gearbox faults with a lightweight one-dimensional network, evaluated against ResNet baselines.

My contribution

I contributed to model evaluation and analysis, interpreted the results, and reviewed the manuscript.

IEEE AccessQ1Impact Factor 4.2 · 12 · 188755-188765

2024

Journal article

GCDN-Net: Garbage classifier deep neural network for recyclable urban waste management

Md Mosarrof Hossen, Azad Ashraf, Mazhar Hasan, Molla E. Majid, Mohammad Nashbat, Saad Bin Abul Kashem, Ali K. Ansaruddin Kunju, Amith Khandakar, Sakib Mahmud, Muhammad E.H. Chowdhury

Classified recyclable waste from images for both single-label and multi-label tasks. The study evaluated the method on the 25,000-image GIGO dataset and used Score-CAM to inspect which regions influenced predictions.

My contribution

I helped develop the deep-learning framework and contributed to evaluation, experimental analysis, interpretation, and manuscript review.

Waste ManagementQ1Impact Factor 7.3 · 174 · 439-450

2024

Book chapter

Rotational Vibration-Based Piezoelectric Energy Harvesting with Bandwidth Enhancement Through Magnetic Oscillators

Muhammad Hafizh, Abdelrahman Ali, Sakib Mahmud, Mohamed A. Ayari, Asan G. A. Muthalif

Used magnetic oscillators to broaden the operating bandwidth of a rotational piezoelectric energy harvester. The study paired analytical beam modelling with finite element analysis.

My contribution

I contributed to technical analysis, interpretation of the results, and manuscript review.

Proceedings of Ninth International Congress on Information and Communication Technology · 1003 · 27-36 · London, United Kingdom

2022Online 2021

Journal article

COVID-19 vaccine willingness and hesitancy among residents in Qatar: a quantitative analysis based on machine learning

Muhammad Hafizh, Yousif Badri, Sakib Mahmud, Amir Hafez, Pilsung Choe

Used survey responses from Qatar residents to model vaccine attitudes and examine factors associated with willingness and hesitancy.

My contribution

I contributed to quantitative and machine-learning analysis, interpretation, and manuscript preparation and review.

Journal of Human Behavior in the Social EnvironmentQ1Impact Factor 1.8 · 32(7) · 899-922