Deepnoid, a first-generation medical AI company in South Korea, announced on Tuesday that it plans to revolutionize the medical AI landscape through the commercialization of M4CXR, a cutting-edge multimodal generative AI-based medical device.
M4CXR, an innovative chest X-ray interpretation solution powered by generative AI, can analyze for 41 types of lesions in mere seconds and automatically produce draft reading reports. The company has initiated multi-center clinical trials following approval from the Ministry of Food and Drug Safety in August, marking its entry into the early stages of commercialization.
Unlike existing unimodal solutions that merely identify lesions and suspicious areas, M4CXR generates comprehensive interpretation reports suitable for actual clinical use. This breakthrough offers healthcare professionals immediate, actionable insights for diagnosis, capturing significant industry attention.
A key feature of M4CXR is its multimodal learning capability, enabling simultaneous analysis of both images and text. By integrating Vision Transformer (ViT) and Large Language Model (LLM) technologies, it achieves a holistic understanding of X-ray images.
M4CXR goes beyond simple diagnoses, producing detailed, professional draft reading reports comparable to those written by experienced physicians. Trained on an extensive dataset of over 10 million chest X-rays and corresponding reports, it demonstrates high accuracy even in rare and complex cases, promising to enhance both the diagnostic efficiency of medical staff and the reliability of patient care.
Deepnoid’s CEO, Choi Woo-sik, emphasized, “Generative AI-based medical imaging technology is an innovation that enhances both healthcare accuracy and accessibility. The commercialization of M4CXR will solidify Deepnoid’s position as a technological leader in the field.”
Meanwhile, Deepnoid is expanding its diagnostic capabilities by developing a new solution that incorporates CT imaging, aiming to broaden its AI-based diagnostic technologies for diverse clinical environments.
The company is also set to launch a Total AI Interpretation Service Platform, integrating its proprietary teleradiology interpretation system with medical AI solutions. This comprehensive platform, designed to streamline the entire teleradiology interpretation process and boost healthcare professionals’ efficiency, is targeted for development completion in the first half of 2026, with commercialization plans to follow.
