Corologic's AI Reconstructs Coronary Angiography to Cut X-Ray Dose

For a patient undergoing a coronary angiogram, the procedure is a necessary but stark trade-off. The live X-ray video that guides a cardiologist's catheter also delivers a cumulative dose of ionizing radiation, a known risk for both the patient and the clinical team performing the lengthy intervention. It is a fundamental tension in the cath lab, one that Seoul-based startup Corologic is trying to resolve not by reinventing the hardware, but by rewriting the video feed with AI.

Founded in May 2026 as a spinout from Seoul National University Bundang Hospital (SNUBH), Corologic has developed Angio-FILM, a software medical device that uses generative AI to reconstruct low-frame-rate coronary angiography footage in real time [corologic.io, retrieved 2026]. The clinical premise is direct: if the system can convincingly interpolate the intermediate frames between actual X-ray exposures, the lab could theoretically halve the frame acquisition rate while preserving a smooth, diagnostic-quality view of the coronary vessels and the interventional tools [VentureSquare, October 2026]. The company announced an undisclosed seed round in September 2026 from Kakao Ventures and Seoul National University Technology Holdings to fund its path to regulatory markets [Seoul Economic Daily, September 2026].

The clinical wedge in video frame rates

The product's wedge is technical but its intended outcome is humane. Coronary angiography, the gold standard for diagnosing blockages in heart arteries, requires continuous X-ray imaging to visualize contrast dye flowing through vessels and to safely navigate wires and stents. High frame rates (typically 15 frames per second or more) provide fluid motion but come with proportional radiation exposure. Corologic's software is designed to sit in the imaging pipeline, taking a lower-frame-rate input,say, 7.5 fps,and using a trained AI model to generate the in-between frames, outputting a restored video that appears to run at the higher, smoother rate [Asia Economy, September 2026]. The company claims this allows for a significant reduction in X-ray dose without compromising the procedural view [corologic.io, retrieved 2026]. Its first regulatory milestone was a Korean registration completed just three months after founding [Jayden Won LinkedIn, retrieved 2026].

A team built for the regulated device path

Corologic's founding team appears structured to navigate the long, evidence-based journey of a medical device. The clinical anchor is Professor Sihyuck Kang, an interventional cardiologist at SNUBH and the co-inventor of the core technology, who serves as Chief Medical Officer [KORIT, September 2026]. He is joined by CEO KyungRae Hwang, who brings 24 years of medical device industry experience in business and management, and CTO Youngrak Choi, who has a background in developing angiography equipment and medical imaging AI [KORIT, September 2026]. This blend of deep clinical domain expertise, regulated commercial experience, and technical product development is a deliberate architecture for a company whose immediate roadmap is dominated by clinical validation and regulatory filings, not just software iteration.

The company has outlined a clear, capital-intensive use of proceeds from its seed round, which points squarely at the barriers to entry in medical AI.

• Multicenter trials. Proceeds are earmarked for multicenter clinical validation studies, the essential evidence required for regulatory submissions [Asia Economy, September 2026].
• Dual regulatory path. Corologic is pursuing approval from South Korea's Ministry of Food and Drug Safety (MFDS) followed by a 510(k) clearance from the U.S. Food and Drug Administration [Asia Economy, September 2026].
• Quality certification. The company also plans to obtain ISO 13485 certification, the international quality management standard for medical devices [Asia Economy, September 2026].

This regulatory focus is a defining characteristic, separating it from many AI tools that operate in an adjunct, non-diagnostic capacity. Angio-FILM is intended to be part of the primary imaging workflow, a position that demands a higher standard of validation and regulatory scrutiny.

The risks of inference in a high-stakes field

The ambition is significant, and so are the technical and commercial hurdles. The most profound challenge is the risk profile of generative inference in a live, life-critical procedure. While frame interpolation is a known technique in computer vision, its application in medical imaging,where every pixel can inform a clinical decision,carries a unique burden of proof. The AI must not only make the video look smooth; it must never hallucinate or distort anatomic features or device positions in a way that could mislead the operator. Corologic will need to demonstrate through rigorous clinical trials that its interpolated frames are diagnostically equivalent to native high-frame-rate images across a wide range of patient anatomies and disease states.

Furthermore, the competitive landscape in medical imaging AI is crowded and global. While no direct competitors were named in the available sources, the field of computational angiography is active, with large incumbent device manufacturers and software specialists exploring dose-reduction techniques. Corologic's success will depend on proving not just efficacy, but also smooth integration into existing cath lab workflows from major vendors like Siemens Healthineers, Philips, and GE Healthcare. The company's on-premise focus for clinical imaging workflows is a practical nod to hospital IT security, but it also adds implementation complexity [corologic.io, retrieved 2026].

The next twelve months of evidence

For Corologic, the coming year is less about commercial rollout and almost entirely about evidence generation. The seed funding provides runway to execute the clinical studies that will form the backbone of its MFDS and FDA submissions. Key milestones to watch will be the initiation and results of those trials, the filing for Korean approval, and any strategic partnerships formed with research hospitals or imaging vendors to accelerate validation. The backing from Kakao Ventures, a firm with a growing healthtech portfolio, and the university's own holding company, provides not just capital but potentially valuable networks within the Korean healthcare system [Seoul Economic Daily, September 2026].

The patient population here is clear: individuals undergoing diagnostic or interventional coronary procedures, who number in the millions globally each year. For them, the standard of care today involves a direct, unavoidable trade-off between obtaining the necessary visual guidance and accumulating radiation dose. Cardiologists and cath lab staff, who perform these procedures daily, also face a lifelong occupational exposure risk. Corologic's bet is that an AI layer, quietly interpolating frames in real time, can meaningfully rewrite that equation. It is a technically elegant proposition that now faces the hard, iterative work of clinical science and regulatory review to prove it is both safe and effective.

Sources

1. [corologic.io, retrieved 2026] CoroLogic | Lower-dose coronary angiography, restored by AI | https://www.corologic.io/
2. [Seoul Economic Daily, September 2026] Cardiovascular AI Startup Corologic Raises Seed Funding From Kakao Ventures | https://en.sedaily.com/technology/2026/09/15/cardiovascular-ai-startup-corologic-raises-seed-funding
3. [Asia Economy, September 2026] Cardiovascular AI Startup Corologic Secures Seed Funding from Kakao Ventures and Seoul National University Holdings | https://www.asiae.co.kr/en/article/2026091508144099495
4. [KORIT, September 2026] CoroLogic、Kakao Ventures・ソウル大技術持株からシード資金調達 | https://www.korit.jp/funding/healthcare/platum-corologic-kakaoventures-seed-260917/
5. [VentureSquare, October 2026] 코로로직, X선 촬영 절반으로 줄이고 AI로 영상 복원한다 | https://www.venturesquare.net/1117395/
6. [Jayden Won LinkedIn, retrieved 2026] Jayden Won - LG Chem | https://www.linkedin.com/in/jayden-won/