Saianiruth M — Writing https://saianiruth.me/writing Notes on medical-imaging AI, production ML systems, and what I've learned shipping research-grade work into clinical workflows. en-us Thu, 04 Jun 2026 03:29:30 GMT Grad-CAM for Medical Imaging: Reading a Heatmap on a Pediatric Chest X-ray https://saianiruth.me/notes/grad-cam-medical-imaging https://saianiruth.me/notes/grad-cam-medical-imaging Thu, 04 Jun 2026 00:00:00 GMT How Grad-CAM works in thirty lines of PyTorch, what it tells you on a chest X-ray, and the failure mode every medical-imaging engineer should know. sai2804aniruth@gmail.com (Saianiruth M) notes grad-cam cxr primer explainer interpretability Three Detectors, One Ensemble: Inside Our Shoulder-Fracture Detection System https://saianiruth.me/blog/shoulder-fracture-ensemble-faster-rcnn-efficientdet-rf-detr https://saianiruth.me/blog/shoulder-fracture-ensemble-faster-rcnn-efficientdet-rf-detr Wed, 03 Jun 2026 00:00:00 GMT An engineer-readable walkthrough of our shoulder-fracture detection ensemble — Faster R-CNN, EfficientDet, and RF-DETR — and what ensembles actually buy you when individual models are already strong. sai2804aniruth@gmail.com (Saianiruth M) blog object-detection ensemble shoulder-fracture faster-rcnn efficientdet rf-detr medical-imaging arxiv-paper Gemini vs CNN for Clinical Quality Control: Where Each Wins on 3,566 Chest X-Rays https://saianiruth.me/research/gemini-vs-cnn-clinical-qc-chest-xray https://saianiruth.me/research/gemini-vs-cnn-clinical-qc-chest-xray Mon, 01 Jun 2026 00:00:00 GMT A head-to-head benchmark of Gemini 2.5 Flash, Gemini 3 Flash Preview, and 14 CNN classifiers on the same chest X-ray QC task. Per-task numbers, cost, and latency. sai2804aniruth@gmail.com (Saianiruth M) research gemini llm cnn benchmarking quality-control chest-xray vlm CNN Components for Medical Imaging: What Kernels, Pooling, and Receptive Fields Actually Do https://saianiruth.me/notes/cnn-components-medical-imaging https://saianiruth.me/notes/cnn-components-medical-imaging Mon, 01 Jun 2026 00:00:00 GMT A primer on the four CNN primitives — kernels, conv layers, pooling, receptive fields — grounded in real activations and Grad-CAMs from a chest X-ray classifier. sai2804aniruth@gmail.com (Saianiruth M) notes cnn cxr primer explainer classification Shipping a PyTorch Model as a Windows Installer: Electron + PyInstaller + AES-256 Model Decryption https://saianiruth.me/blog/pytorch-windows-installer-electron-pyinstaller https://saianiruth.me/blog/pytorch-windows-installer-electron-pyinstaller Thu, 28 May 2026 00:00:00 GMT Shipping a PyTorch model as a Windows installer: the stack, the pitfalls, and the war stories nobody else writes about. sai2804aniruth@gmail.com (Saianiruth M) blog mlops electron pyinstaller windows aes deployment production Year One in Production Medical AI: The Honest Version https://saianiruth.me/blog/year-one-production-medical-ai https://saianiruth.me/blog/year-one-production-medical-ai Sat, 23 May 2026 00:00:00 GMT My first year as an AI Scientist — 24+ models, one arXiv paper, and a few lessons I wish I'd had on day one. sai2804aniruth@gmail.com (Saianiruth M) blog medical-imaging production-ml mlops ensembles vlm reflection year-in-review