Deepnoid Receives MFDS Approval for Clinical Trial Plan of Generative AI 'M4CXR'

Full-Scale Clinical Trials Begin for Generative AI-Based Digital Medical Devices
Drafts Highly Reliable Reports by Interpreting 41 Types of Lesions
Delivers Rapid Preliminary Results in Emergency Situations

Deepnoid, a first-generation medical artificial intelligence (AI) specialist company in Korea, announced on August 19 that it has received approval from the Ministry of Food and Drug Safety for an Investigational New Device (IND) clinical trial plan for its generative AI medical device, 'M4CXR'.

This approval is in accordance with the Digital Medical Products Act Enforcement Rules and signifies the full-scale initiation of clinical trials for generative AI-based digital medical devices, which have been evaluated to outperform existing neural network models in medical image interpretation and reasoning.

Deepnoid will verify the clinical efficacy and safety of M4CXR through a multi-center, retrospective, confirmatory clinical trial involving Kangbuk Samsung Hospital and Boramae Medical Center. The company plans to begin clinical trials next month and aims to obtain medical device approval in the near future. By utilizing the new medical technology evaluation deferment system, Deepnoid intends to enter the clinical field as a non-reimbursed service and will pursue step-by-step inclusion in the national health insurance reimbursement list.

M4CXR interprets 41 types of lesions from chest X-ray images and drafts a consistent and highly reliable preliminary radiology report within seconds. The technology was developed by training on more than 10 million radiology report datasets matched with chest X-ray images, enabling medical staff to receive immediate preliminary results in emergency situations. This supports healthcare professionals in focusing more on critical clinical decisions and patient care, while also assisting radiologists in their interpretations.

Kim Sunghyun, Director of the Human Radiology Center, explained, "Chest X-rays are the most frequently performed examination in clinical settings, but there have been limitations in interpretation speed due to a shortage of radiologists." He added, "With rising demand for image interpretation, we expect that even in emergencies, preliminary results can be delivered to medical staff immediately upon request. This will enhance diagnostic speed and contribute to greater accuracy by supporting radiologists in their interpretations."

Director Kim further stated, "The significance lies in the fact that this is a radiology interpretation solution created with a large language model (LLM)-based generative AI, which is expected to overcome the limitations of existing neural network models."

Choi Woosik, CEO of Deepnoid, said, "The upcoming clinical trials will serve as a momentum not only for Deepnoid but for the entire medical AI industry, as they will verify the clinical efficacy of generative AI-based medical devices." He emphasized, "Following the commercialization roadmap for M4CXR, we will steadily pursue medical device approval and health insurance reimbursement registration, thereby solidifying Deepnoid's technological leadership."

Deepnoid has secured a stable exclusive market through collaboration with domestic medical image remote interpretation specialist hospitals and is currently providing remote interpretation services to tertiary hospitals and public health centers. The company plans to launch a 'Total AI Interpretation Service Platform' that integrates the entire remote interpretation process, evolving from its existing Tele-Radiology System (TRS). Leveraging its accumulated technological expertise and seamless infrastructure, Deepnoid is presenting a differentiated integrated solution that covers everything from on-site interpretation by medical staff to remote interpretation services, leading the medical AI market.

This content was produced with the assistance of AI translation services.