Cross-population generalizability audit of lung-cancer radiomics & dosiomics between Western (TCIA / Maastro) and South Asian cohorts. Forensic DICOM curation, 107 IBSI-compliant features, domain-shift statistics, and SHAP explainability.
Medical Physics & Biomedical Engineering researcher focused on radiomics, dosiomics, and explainable AI for equitable cancer care in South Asia, auditing whether models trained on Western data remain reliable for underrepresented populations.
I am a medical physics and biomedical engineering graduate focused on AI in oncology, radiomics and dosiomics, medical imaging, radiation therapy, and explainable AI. My research asks a practical question before deployment: do machine-learning models developed on one population remain reliable when applied to another, especially where South Asian patients are underrepresented in validation cohorts?
I combine hands-on clinical workflow experience (CT simulation, treatment-plan review, IMRT / VMAT / DIBH quality assurance) with reproducible computational research and open-source scientific software. I am seeking graduate research at the intersection of clinical medical physics, quantitative imaging, trustworthy machine learning, and equitable cancer care.
IELTS Academic: 7.5 (L 8.5 · R 8.0 · W 6.5 · S 7.0) · CGPA: 3.63 / 4.00
Square Hospitals Ltd.
·
Ahsania Mission Cancer and General Hospital
Collaborative research role across two oncology centres in Dhaka (one project, multi-site data and clinical exposure).
Center for Bio-Health Research Bangladesh (CBHR)
·
Remote · Competitively selected (FRONTIER-02)
Peer-reviewed Springer book chapterOpen-source research software and domain-shift tooling, designed for reproducibility and clinical relevance.
Cross-population generalizability audit of lung-cancer radiomics & dosiomics between Western (TCIA / Maastro) and South Asian cohorts. Forensic DICOM curation, 107 IBSI-compliant features, domain-shift statistics, and SHAP explainability.
Web service that audits domain shift between two hospital cohorts: upload imaging-feature tables, receive per-feature drift statistics (Mann-Whitney U, FDR correction) and a clear transportability verdict.
Research project using Python, PyRadiomics, and Explainable AI to predict treatment response to Transarterial Chemoembolization (TACE) in Hepatocellular Carcinoma (HCC).
Focused programs spanning trustworthy machine learning, medical imaging physics, and clinical radiation therapy.
Peer-reviewed presentations, awards, and accredited scientific events: national and international.
BSc in Medical Physics & Biomedical Engineering
CGPA 3.63 / 4.00 · University Merit Scholarship (multiple semesters)
Coursework: radiation oncology physics, treatment planning, dosimetry, radiation biology, digital signal & image processing, biostatistics, brachytherapy, nuclear medicine, radiological health and safety.
Notes on medical physics, research, and ideas I am learning and sharing. Also on Medium.
A plain-language look at optogenetics and why the 2026 Nobel Prize in Physiology or Medicine matters for how we understand and control living systems.
Read on Medium →What it means to study medical physics: the quiet work behind radiation therapy, precision, and a personal philosophy of making the invisible matter.
Read on Medium →Current programs and recent acceptances.