AI‑assisted Medical Imaging & Diagnostics
Building AI pipelines that enhance diagnostic imaging quality and enable reliable automated analysis for clinical use.
Overview#
Work includes unpaired learning for handheld fundus photography enhancement and joint learning systems for lesion detection using optical imaging features. Emphasis is on improving downstream diagnostic performance and robustness in real clinical scenarios.
Key projects#
- Automated enhancement for handheld fundus photography via unpaired learning.
- Joint learning for automated diagnosis of non‑melanoma skin cancer using optical attenuation features.
Representative publications#
- “Automated Enhanced Handheld Fundus Photography via Unpaired Learning” (IEEE Trans. Instrum. Meas., 2025)
- “Automated Diagnosis of Non‑Melanoma Skin Cancer: A Joint Learning Approach Using Optical Attenuation Coefficients” (npj Digital Medicine, 2025)
If you want code, data, or preprints related to these projects, please reach out — I can share where available for collaboration and reproducibility.