Projects

CXRGenie-Multimodal Chest X-ray Diagnostic Assistant

B.Tech Thesis, MiRL Lab, IITM

Developed a modular AI agent for chest X-ray interpretation using GPT-4o-mini, integrating segmentation, classification, grounding, VQA, and report generation, supported by a RAG-enabled vector database for retrieval-augmented reasoning.

Achieved transparent and safe diagnostic support with human-in-the-loop feedback and benchmarked performance on ChestAgentBench and CheXBench datasets.

Defect-Aware Denoising and Deblurring Network

EE5179-Deep Learning for Imaging (KLA Project), IIT Madras

Built an enhanced encoder–decoder architecture with Spatial-Channel Attention and Non-Local Blocks for defect-preserving image restoration.

Comparative Analysis of Z-axis 3D Super-Resolution Techniques

MiRL Lab, IIT Madras

Comparative study of z-axis 3D MRI super-resolution using GAN, Diffusion, and Implicit Fields.

Analyzed trade-offs between image quality and inference speed for clinical deployment.

FusionDecompNet: A Multimodal Multitask Learning Framework for Alzheimer’s Diagnosis

MICAIH Lab, NIT Trichy (Ram Thaila Scholarship Project)

2D-to-3D Coronary Artery Reconstruction and stenosis detection

Jun 2024 – May 2025 | MiRL Lab, IITM & Meril Life Sciences

Developed algorithms for 2D-to-3D coronary artery reconstruction from X-ray angiography, enhancing stenosis detection accuracy and reducing computation time via pseudo-transient vFFR.

CVS Classification and hepatocystic anatomy segmentation in laparoscopic cholecystectomy

Jun 2025 – Aug 2025 | MiRL Lab, IITM

Developing optimized deep learning models for laparoscopic cholecystectomy: video-based CVS classification and hepatocystic anatomy segmentation to assist surgical workflows.