AIDOT, an AI-based women’s health diagnostics company, announced that it has successfully developed an artificial intelligence (AI) diagnostic support software for detecting HPV (human papillomavirus) using cervical colposcopy images. This achievement was realized through the “G-Valley Medical Device Development Support Center – Medical Data Enterprise Support Project.”
The AI model was developed in collaboration with Professor Donghoon Seo’s team from Seoul National University Bundang Hospital, Department of Obstetrics and Gynecology. Internal performance evaluations showed a sensitivity of 80.92% and specificity of 82.75%, confirming its potential for clinical application.
This technology introduces a new diagnostic paradigm by allowing immediate HPV detection from cervical images, compared to conventional DNA tests that take several days to deliver results. By instantly analyzing the captured cervical images, the AI can determine HPV infection status (Positive/Negative), significantly improving both speed and cost-efficiency.
The new HPV detection algorithm can be integrated with AIDOT’s Cerviray AI, which already classifies cervical lesions into CIN1, CIN2/3, and Cancer. This integration enables comprehensive assessment of disease severity, risk, and HPV infection status, while providing precise lesion location information to assist clinicians in making informed decisions.
AIDOT has begun MFDS clinical trials and domestic regulatory approval processes, with plans to publish related research in international academic journals. The company aims to commercialize this technology as a novel HPV screening tool globally, including in Korea.
Choi Hansol, Head of Research at AIDOT, stated: “The results of this study are expected to significantly enhance cancer prevention through early detection of high-risk patients. The technology can streamline patient triage, reduce unnecessary biopsies or cytology tests, improve timeliness across the care cascade, standardize interpretation, and ultimately increase cervical cancer survival rates by detecting CIN2+ lesions early.”
He added: “Image-based HPV detection expands the early cervical cancer detection system to a new level. Together with the on-device Cerviray AI, which recently received domestic MFDS certification, we will strengthen our technological leadership in the global women’s health market.”
Korea Economic TV / November 28, 2025 / Senior Reporter Jaejun Yang
