🩺 Biomedical Acoustic Signal Processing
My research in this area focuses on developing deep learning frameworks for analyzing physiological acoustic signals such as heart and lung sounds. The goal is to achieve robust real-time denoising, segmentation, and disease classification, enabling reliable computer-aided auscultation systems in noisy environments.

Journal Publications

• A Lightweight CNN Model for Detecting Respiratory Diseases from Lung Auscultation Sounds Using EMD-CWT-based Hybrid Scalogram
  IEEE Journal of Biomedical and Health Informatics, 25(7), 2595–2603, 2021.
  Developed a hybrid time–frequency approach for respiratory sound classification using CNNs.

• CardioXNet: A Novel Lightweight Deep Learning Framework for Cardiovascular Disease Classification Using Heart Sound Recordings
  IEEE Access, 9, 36955–36967, 2021.
  Proposed an efficient model for heart sound–based cardiovascular disease detection.

• An End-to-End Deep Learning Framework for Real-Time Denoising of Heart Sounds for Cardiac Disease Detection in Unseen Noise
  IEEE Access, 2023.
  Introduced an end-to-end model for robust heart sound denoising under unknown noise.

• NRC-Net: Automated Noise Robust Cardio Net for Detecting Valvular Cardiac Diseases Using Optimum Transformation Method with Heart Sound Signals
  Biomedical Signal Processing and Control, 86 (Part C), 2023.
  Enhanced robustness of cardiac disease detection using adaptive transformation methods.

• A Self-Attention-Driven Deep Denoiser Model for Real-Time Lung Sound Denoising in Noisy Environments
  Manuscript under review.
  Applies attention-based architectures for adaptive noise suppression in lung auscultation.

• A Multi-Stage Hybrid CNN-Transformer Network for Automated Pediatric Lung Sound Classification
  Manuscript under review.
  Combines CNN and Transformer modules to classify pediatric lung disorders.

• BUET Multi-disease Heart Sound Dataset: A Comprehensive Auscultation Dataset for Developing Computer-Aided Diagnostic Systems
  Manuscript under review.
  Curated a large-scale annotated dataset to advance heart sound AI research.

Theses

• Undergraduate Thesis: A Robust Deep Learning Framework for Real-Time Denoising of Heart Sound
  Supervisor: Dr. Taufiq Hasan

• Postgraduate Thesis: Automated Denoising and Classification of Lung Sounds Using End-to-End Deep Learning Models
  Supervisor: Dr. Taufiq Hasan

Projects

• AI-based software for heart sound denoising

• Low-cost ECG acquisition system for real-time arrhythmia detection

• Generative model-based framework for EEG denoising

🧠 2. Medical Imaging and Computer Vision
This research area explores deep learning–based disease diagnosis, segmentation, and detection from medical images, combining CNNs, Vision Transformers, and 3D architectures for improved interpretability and accuracy in ocular and pulmonary imaging.

Journal Publications

• Benchmarking Deep Learning Frameworks for Automated Diagnosis of Ocular Toxoplasmosis: A Comprehensive Approach to Classification and Segmentation
  IEEE Access, 2023.
  Benchmarked multiple CNNs for ocular lesion detection and segmentation.

• AutoLungDx: A Hybrid Deep Learning Approach for Early Lung Cancer Diagnosis Using 3D Res-U-Net, YOLOv5, and Vision Transformers
  Manuscript under review.
  Proposed a 3D multi-stage hybrid model for early detection of lung nodules.

Projects

• End-to-end deep learning framework for lung nodule detection and classification

• Low-cost mechanical ventilator prototype

🧬 3. Bioinformatics and Computational Genomics
This domain integrates AI and molecular bioinformatics, focusing on protein function prediction, DNA-binding protein classification, and biomedical data mining using wavelet analysis and transformer-based embeddings.

Journal Publications

• Application of Wavelet Transformation and Artificial Intelligence Techniques in Healthcare: A Systemic Review
  Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 15(2): e70007, 2025.
  Comprehensive review of AI–wavelet integration in medical data analysis.

• ResCap-DBP: A Lightweight Residual-Capsule Network for Accurate DNA-Binding Protein Prediction Using Global ProteinBERT Embeddings
  Manuscript under review.
  Introduced a capsule-based hybrid network leveraging transformer embeddings.

Projects

• Deep learning–based DNA-binding protein prediction using optimized feature extension techniques

⚙️ 4. Biomedical Device and System Design
This research domain focuses on hardware–software co-design for affordable healthcare technology. It includes embedded systems for physiological monitoring, medical IoT devices, and intelligent signal-driven diagnostic systems.

Conference Papers

• A Low-Cost, Low-Energy Wearable ECG System with Cloud-Based Arrhythmia Detection
  2020 IEEE Region 10 Symposium (TENSYMP), pp. 1840–1843.
  Designed a wearable ECG device with cloud-based arrhythmia classification.

• DengueGuard: An Innovative Non-Invasive Dengue Vital Monitoring Device
  ICECE 2024.
  Developed an integrated dengue monitoring prototype with clinical decision support.

Projects

• Wearable ECG acquisition device for arrhythmia detection

• Low-cost mechanical ventilator system

• Microneedle patch–based transdermal insulin delivery simulation