Medical AI
Deep learning for digital pathology, radiology, and clinical decision support.
- • Lupus Nephritis Classification
- • Glioma Diagnosis
- • Breast Cancer Subtyping
- • Multiple Instance Learning
Research
Research & Publications
I work at the intersection of machine learning and medicine, building deep-learning systems for medical image analysis, computer vision, and applied AI. My published work spans gigapixel histopathology, multiple-instance learning, and curated medical datasets. Selected publications and research interests are listed below.
Selected Publications
- A pipeline for automated classification of lupus nephritis subtypes from gigapixel whole slide imagesMIDL 2024 (Medical Imaging with Deep Learning)·2024
Amit Sharma et al.
A deep-learning pipeline for classifying lupus nephritis subtypes directly from gigapixel digital whole-slide images, designed to operate efficiently in resource-constrained clinical environments.
- A multi-cohort, multimodal dataset for diffuse glioma diagnosis with paired histopathology and MRINature Scientific Data·2024
Amit Sharma et al.
A curated brain-tumor dataset benchmarking multiple-instance learning methods for glioma diagnosis, achieving a 12% improvement over prior state-of-the-art on 3-way classification.
- Multiple Instance Learning for Glioma Diagnosis using Hematoxylin & Eosin Whole-Slide ImagesarXiv preprint·2024
Amit Sharma et al.
Benchmarks of MIL approaches on H&E whole-slide images for diffuse glioma classification, with extensive ablations on feature extractors and aggregation strategies.
Research Interests
Machine Learning
Core methods that power my applied work — from representation learning to multimodal fusion.
- • Computer Vision
- • Self-supervised Learning
- • Transfer Learning
- • Multimodal Models
Applied Research
Bridging research and product, with a focus on usable, deployable AI systems.
- • Emotion AI
- • ECG-based Question Answering
- • Human-AI Interaction
- • Generative Models
Interested in collaborating?
I'm open to research collaborations and applied ML projects. Take a look at the portfolio or hire me for engineering work, or read more on the blog.