A New Approach to Early Breast Cancer Detection
The ThermoMind Vision One breast cancer screening device uses multimodal sensing and advanced AI to track digital biomarkers for safe, radiation-free detection. The system relies on Gidel’s FantoVision edge computer to acquire high-speed, high-resolution image data in real time.
Breast cancer is a major global health concern, and early detection dramatically improves outcomes. Standard imaging tools such as mammography, ultrasound, and MRI involve challenges such as ionizing radiation, false positives, or high operational costs.
ThermoMind Breast Cancer Screening Device and Multimodal Imaging
ThermoMind developed a supplementary technology that combines multimodal sensors with advanced AI to identify metabolic abnormalities linked to cancer. The device captures over 300 data points using infrared and depth-sensing technology to create a digital skin model and monitor vascular and metabolic patterns.
The screening takes only four minutes and requires no radiation, no physical contact, and no invasive procedures. Unlike mammography or ultrasound, which focus on anatomical structures, ThermoMind detects physiological changes such as vascular flow and thermal anomalies.
Non-Invasive, No-Radiation Breast Cancer Screening Approach
ThermoMind’s Vision One is designed for comfort and privacy. The chest-height imaging ring houses infrared and depth sensors that capture a 180-degree view of the chest and axillae. Real-time AI algorithms analyze these thermal and depth images and provide insights such as vascular mapping and metabolic irregularities.
The system integrates LWIR, NIR, and 3D imaging technologies. It includes user-friendly software for technicians and supports remote consultations with oncologists and radiologists.
FantoVision Inside: High-Bandwidth Image Acquisition and Processing
ThermoMind partnered with Gidel to address the challenge of capturing multiple high-resolution data streams at once. “We needed to acquire many high-resolution sensor streams simultaneously and process them in real time,” says Najeeb Ayoub, CEO of ThermoMind.
Gidel’s FantoVision40 was the ideal choice. This compact embedded vision computer combines an NVIDIA Jetson module with Gidel’s FPGA frame grabbers supporting high-bandwidth interfaces such as 10GigE Vision and CoaXPress 12. Its footprint is only 134 × 90 × 60 mm, and the weight is 750 g, making it easy to integrate into medical systems with tight space constraints.
Because Vision One relies on CoaXPress cameras, the FantoVision40 delivers up to 40 Gb/s of imaging throughput. Fifteen sensors capture high-resolution images at 30–60 fps, and Gidel’s real-time compression prevents bottlenecks by processing data during acquisition.
“The real-time compression was essential. It let us analyze the data during acquisition rather than after it,” Ayoub explains.
Making Next-Generation Breast Cancer Screening Accessible
The breast cancer screening device is easier and more affordable to deploy than traditional imaging systems. It requires only a climate-controlled room and power supply, making it suitable for emerging regions with limited medical infrastructure.
ThermoMind is conducting a global multi-center clinical trial in collaboration with major institutions such as the University of Heidelberg, MD Anderson Cancer Center, and Assuta Medical Centers. The study evaluates thermal video streams combined with AI algorithms for early-stage breast cancer detection.
The goal is to validate a safe, cost-effective imaging modality suitable for women of all ages and breast densities, enabling earlier detection and improved patient monitoring.
Want to understand how ThermoMind’s workflow uses FantoVision for high-bandwidth imaging?
Watch the overview video below.
