Ariful Islam
Assistant Professor
BE
www.bme.iu.ac.bd
About:
Assistant Professor in the Department of Biomedical Engineering at Islamic University, Bangladesh, with teaching and research interests in biomedical signal processing, biomedical instrumentation and diagnostic systems, medical device design and prototyping, biosensors, medical imaging, and artificial intelligence for healthcare. His research focuses on computational and engineering approaches for disease diagnosis, physiological signal analysis, biomarker discovery, and intelligent healthcare systems. He is actively involved in multidisciplinary research, student mentoring, and biomedical technology development.
Research interest:
*Biosignal Processing [ECG, EEG, EMG] *Biomedical Instrumentation *Biomedical Device Design & Fabrication *Biosensors *Medical Imaging, Computer Vision and XAI, *DL & ML in Healthcare; *Bioinformatics, Rehabilitation Techniques.
ResearchGate:
https://www.researchgate.net/profile/Ariful-Islam-59?ev=hdr_xprf
Google Scholar:
https://scholar.google.com/citations?user=_LfLHNEAAAAJ&hl=en
Publications
Integrated Bioinformatics and Machine Learning Analysis Identifies MT1F as a Potential Diagnostic Biomarker and Therapeutic Target in Breast Cancer
Atlantic Journal of Life Sciences
This study integrates bioinformatics and machine learning approaches to identify potential biomarkers for breast cancer. Gene expression datasets were analyzed using differential expression, protein–protein interaction networks, and machine learning models including Random Forest, Gradient Boosting, and XGBoost. MT1F emerged as a key hub gene with strong diagnostic potential, while additional analyses of methylation, immune infiltration, pan-cancer expression, and drug sensitivity suggested its possible therapeutic relevance in breast cancer.
2026-06-07
Click Here
Adaptive synthetic-based arrhythmia classification using machine learning techniques
IAES International Journal of Artificial Intelligence (IJ-AI)
This study presents an advanced arrhythmia classification framework that combines Adaptive Synthetic (ADASYN) sampling with multiple machine learning algorithms to address severe class imbalance in ECG-related data. Using the UCI Arrhythmia dataset, ten classifiers—including Random Forest, Gradient Boosting, LightGBM, and XGBoost—were evaluated, with the leading ensemble models achieving 97% accuracy. The findings demonstrate that integrating ADASYN with machine learning can improve automated arrhythmia classification and support more reliable cardiac diagnostic systems.
2026-06-03
Click Here
PSO-ML: Heartbeat Classifier Using Particle Swarm Optimization And Machine Learning Techniques
IOSR Journal of Dental and Medical Sciences (IOSR-JDMS)
This study presents a hybrid ECG heartbeat classification framework combining Particle Swarm Optimization (PSO) with machine learning algorithms to improve arrhythmia detection. Using the MIT-BIH Arrhythmia Database, four models—PSO-SVM, PSO-ETC, PSO-GBM, and PSO-XGBoost—were evaluated for classifying five heartbeat categories. The proposed models achieved 95–98% accuracy, demonstrating the potential of PSO-based machine learning for automated and reliable ECG analysis.
2023-10-01
Click Here
Alzheimer’s Disease Detection Using m-Random Forest Algorithm with Optimum Features Extraction
2021 1st International Conference on Artificial Intelligence and Data Analytics (CAIDA)
This study proposes a Modified Random Forest (m-RF) machine learning approach for early Alzheimer’s disease detection using data from the OASIS longitudinal MRI dataset. The model classifies subjects into three groups—non-demented, demented, and converted—using optimized feature extraction and data preprocessing techniques. The proposed m-RF model achieved 96.43% accuracy, outperforming several conventional machine learning methods and demonstrating strong potential for automated Alzheimer’s disease detection.
2021-04-20
Click Here
The Recording and Evaluation of an ECG Signal with Proper Electrode Placement and LEAD Configurations
Journal of Clinical Trials and Regulations
This study presents the recording and evaluation of electrocardiogram (ECG) signals with emphasis on proper electrode placement and lead configurations. ECG recordings were analyzed to evaluate important parameters including heart rate, rhythm, R–R interval, and QRS complex characteristics. The study highlights the importance of correct electrode placement and lead configuration for reliable ECG recording and interpretation.
2021-02-17
Click Here
Performance Analysis of F-OFDM, FBMC and UFMC Modulation Techniques in 5G Wireless Network Based on Deep Learning
Journal of Network Security Computer Networks
This study presents a comparative performance analysis of three multicarrier modulation techniques—F-OFDM, FBMC, and UFMC—for 5G wireless communication. Using MATLAB simulations, the techniques are evaluated mainly in terms of peak-to-average power ratio (PAPR) and bit error rate (BER) and compared with conventional CP-OFDM. The results demonstrate that each modulation technique offers different performance advantages, highlighting the trade-offs involved in selecting suitable waveform strategies for next-generation wireless networks.
2020-12-19
Click Here
Design and Development of Bluetooth Controlled Car Using Bluetooth Module HC-05
Global Scientific Journal (GSJ)
This study presents the design and development of a Bluetooth-controlled car using an Arduino UNO and HC-05 Bluetooth module. The system allows users to control the vehicle through an Android smartphone, providing forward, backward, left, right, and stop commands via Bluetooth communication. The prototype integrates an Arduino controller, motor driver, DC motors, and the HC-05 module, demonstrating a simple and practical approach to wireless vehicle control.
2020-09-10
Click Here
