Dr. Asad Ullah Senior

PhD Theme/Topic: AI-Driven Early Detection and Risk Stratification of Cardiovascular Diseases Using Multi-Modal Medical Imaging and Clinical Data

Supervisor: Dr. Asad Ullah Senior Associate Professor
Contact #: +92312113166
Email: asadullah.buic@bahria.edu.pk
Campus/School/Dept: BUIC/CS
RAC Approved Supervisor for Research Areas:

• Artificial Intelligence
• Deep Learning
• Medical Imaging
• Intelligent Vision Systems
• Healthcare Analytics

Supervisory Record:
PhD Produced:
PhD Enrolled: 2
MS/MPhil Produced: 7
MS/MPhil Enrolled:6

Topic Brief Description:

Cardiovascular diseases (CVDs) remain a leading global cause of mortality, primarily due to delayed diagnosis and limited access to expert clinical evaluation. The proposed PhD project aims to develop an AI-powered, multi-modal cardiovascular analysis framework capable of early detection, classification, and risk prediction of major CVDs such as coronary artery disease, cardiomyopathies, arrhythmias, and heart-failure-related abnormalities.

Using deep learning on multiple data types—echocardiography videos, ECG signals, CT/MRI scans, and electronic health records (EHR)—the system will integrate image-based biomarkers, functional cardiac parameters, and clinical risk factors. The research will focus on explainable AI (XAI) to support clinical decision-making and increase trust among cardiologists. The final solution will be designed as a deployable CAD (Computer-Aided Diagnosis) tool for hospitals to support timely and accurate disease detection.

Research Objectives/Deliverables:

• To design a multi-modal deep learning architecture that fuses imaging (Echo, CT/MRI) and non-imaging clinical data for comprehensive CVD detection.
• To develop an early-risk stratification model that predicts severity and likelihood of future cardiovascular events using AI-driven biomarkers.
• To integrate explainable AI (XAI) mechanisms that highlight clinically interpretable features (e.g., myocardial wall motion, ejection fraction deviations).
• To build a scalable clinical decision support system (CDSS) for cardiologists, enabling triage, screening, and prioritization of high-risk patients.
• To validate the system with local hospitals (e.g., PIMS, Armed Forces Institute of Cardiology, Shifa) and generate publishable benchmarks.

Research Questions: 

• How can multi-modal medical data be effectively fused to improve early detection accuracy of cardiovascular diseases?
• Which AI-derived biomarkers are most reliable for predicting high-risk cardiac events, and how can their interpretability be ensured using XAI methods?
• How can the framework generalize across patient populations, imaging devices, and clinical settings while maintaining reliability and robustness?
• What clinical parameters significantly influence risk prediction, and how can the model integrate them without introducing bias?
• Can the developed system assist cardiologists in real-time decision-making, and how effective is it compared to standard diagnostic workflows?

Candidate’s Eligibility Profile:

1. The applicant must hold an MS/MPhil (Computer Science / AI / Data Science / Software Engineering / Biomedical Engineering) with CGPA ≥ 3.0
2. Strong background in machine learning, deep learning, medical imaging, or signal processing.
3. Experience in Python, PyTorch/TensorFlow, and handling medical datasets (ECG, MRI, CT, Echo) will be preferred.
4. Ability to conduct independent research, manage clinical collaborations, and contribute to high-impact publications.
5. Excellent verbal and written communication skills.