OvaScope

OvaScope — Ovarian Ultrasound Analysis with Inspectable Evidence OvaScope is a women’s-health research prototype exploring how AI can extract useful information from existing ovarian ultrasound images, without requiring patients to purchase additional devices. This project brings [verified features: image upload, lesion segmentation, trained mass classification, interactive overlays, evaluation dashboard] into one workspace. Users can inspect the original image alongside model-generated results rather than receive an unexplained chatbot response. Our research objective is to investigate whether Meta-powered visual representations and lesion localization can improve the distinction between benign and malignant ovarian masses. We separate segmentation from classification: outlining a region does not establish that it contains cancer. OvaScope’s contribution is an inspectable workflow connecting image analysis, model provenance, and [held-out evaluation, if completed]. Its intended long-term impact is to support more accessible, evidence-based ovarian-mass assessment. This submission is an investigational prototype—not a clinically validated screening or diagnostic tool. Early-stage cancer detection and improved patient outcomes remain objectives requiring further research.