From Scan to Treatment Plan, AI Helps Close Breast Cancer’s Deadliest Gaps

Breast cancer is the most commonly diagnosed cancer among American women, yet the U.S. healthcare system faces significant gaps in care, including missed annual screenings, a projected shortfall of tens of thousands of radiologists over the next decade, and weeks-long waits for treatment-informing genomic tests. To address these challenges, startups participating in the NVIDIA Inception program are building artificial intelligence applications designed to support clinicians across imaging, risk assessment, and treatment planning. iSono Health has developed an FDA-cleared wearable automated 3D ultrasound system called ATUSA, which uses NVIDIA GPU acceleration to capture standardized breast scans in about two minutes with greater sensitivity than traditional handheld devices. Another startup, Whiterabbit.ai, offers FDA-cleared software that automatically assesses breast density and estimates long-term cancer risk, aiming to reduce the burden on radiologists. Once patients are diagnosed, Ataraxis AI utilizes digital pathology slides and standard clinical variables powered by NVIDIA GPUs to predict treatment responses and recurrence risks without waiting weeks for external lab assays. Similarly, SimBioSys creates 3D models of tumors and soft tissue using NVIDIA MONAI and CUDA-X libraries to help guide surgeries and personalize treatment plans. Together, these companies use NVIDIA AI infrastructure to process complex medical data quickly, aiming to improve speed, efficiency, and accuracy across the breast cancer care continuum.