Dr. Joddat Fatima

PhD Theme/Topic: Artificial Intelligence in Biomedical Imaging and Clinical Decision Support Systems

Supervisor: Dr. Joddat Fatima, Senior Assistant Professor, Bahria University
Contact #: +92-345-5170004
Email: joddat.fatima@gmail.com
Campus/School/Dept: BSEAS / CoE-AI / Department of Software Engineering
RAC Approved Supervisor for Research Areas: Artificial Intelligence, Biomedical Imaging, Machine Learning, Computer Vision

Supervisory Record:
PhD Produced: 0
PhD Enrolled: 0
MS/MPhil Produced: 06
MS/MPhil Enrolled: 05

Topic Brief Description:

The research focuses on advancing AI-driven biomedical imaging techniques, including medical data analysis, diagnostic estimation, and automated radiological report generation. It integrates machine learning and deep learning methods to enhance diagnostic accuracy and strengthen clinical decision-making in real-world healthcare environments. Future work will expand these capabilities by incorporating multimodal clinical data, developing more explainable and trustworthy AI models, and creating scalable, deployment-ready systems to support clinicians across diverse medical settings.

1. AI-Driven Multimodal Predictive Modeling for Cardiovascular Disease
   This research will develop advanced AI models that integrate ECG, imaging, clinical history, and lifestyle factors to predict cardiovascular disease progression. It will enable personalized early-intervention strategies through dynamic risk profiling and real-time clinical decision support

2. Spatiotemporal Deep Learning for Radon Concentration Forecasting
   This study will build deep spatiotemporal models using historical, synthetic, and GIS-based environmental data to predict radon levels in soil and water. The work will support environmental hazard assessment and provide early-warning insights for public health and safety planning.

3. Multiscale AI Modeling of Lung Cancer Risk from Radon Exposure
   This research will integrate ecological datasets, radon exposure metrics, and clinical cancer registries to create AI-based lung cancer risk estimation models. The multiscale framework will support population-level surveillance and early identification of high-risk communities.
   

Research Objectives/Deliverables:

1. Develop advanced AI models for biomedical image analysis, including automated diagnostic estimation and radiological report generation
2. Integrate multimodal clinical data (imaging, text, metadata) to enhance model accuracy and robustness
3. Design explainable and trustworthy AI frameworks to improve transparency and clinical acceptance
4. Build prototype systems that are scalable, deployment-ready, and suitable for integration into hospital workflows
5. Improve diagnostic support tools to assist clinicians in decision-making across diverse medical settings
6. Evaluate model performance using real-world clinical scenarios to ensure reliability, safety, and generalizability

Research Questions:

1. How can AI-driven biomedical imaging models improve accuracy and reliability in automated diagnosis and report generation?
2. What combination of multimodal clinical data (images + clinical text + metadata) yields the highest diagnostic performance?
3. How can explainability techniques be integrated to make AI decisions more interpretable and trustworthy for clinicians?
4. What architectural or training strategies ensure scalability and real-world deployment of AI systems in clinical environments?
5. How can AI tools be validated effectively to support clinical decision-making across different medical centers and patient populations?
6. What challenges arise when adapting AI systems to heterogeneous imaging modalities and varying clinical workflows?

Candidate’s Eligibility Profile:

Dr. Joddat Fatima holds a PhD in Computer Science specializing in Biomedical Image Analysis with over 12 years of academic and research experience. She has strong expertise in machine learning, deep neural networks, pattern recognition, and medical imaging. She has supervised numerous MS and undergraduate projects and authored 18+ publications, including IF journals.

She possesses strong programming experience (Python, C++, MATLAB) and is an active member of national and international research communities including PEC, CVPR, BIOMISA, and CoE-AI.