Corologic

CoroLogic develops clinical imaging software for coronary angiography. Its core product, Angio-FILM, is a generative frame-interpolation model that reconstructs the temporal information missing from cine acquired at a reduced frame rate, so that operators can lower radiation dose by acquiring fewer frames while retaining smooth, diagnostically usable motion. The company positions this as a software layer addressing what it describes as a plateau in hardware-based dose reduction (flat-panel detector and X-ray tube refinements from vendors such as GE, Philips and Siemens).

The product is delivered as an on-premises edge platform with a zero-click DICOM workflow intended for existing catheterization labs and requiring no hardware changes. The described pipeline has five steps: acquisition at a lower frame rate; automatic DICOM transfer of the low-frame cine to the edge server; a "Smart Gating" classifier that identifies contrast-filled segments and filters non-diagnostic frames; Angio-FILM restoration; and return of the full-rate sequence via DICOM within seconds. The current workflow covers 7.5→15 fps, with prototype paths for 3.5→7, 7.5→15 and 7.5→30 fps, plus development tracks for sub-second real-time processing and an ultra-low-dose 5→15 fps model. The company states compatibility with GE, Philips, Siemens and Canon systems.

Business model

Two parallel revenue tracks: licensing Angio-FILM as an embeddable software module to angiography equipment manufacturers and OEMs for integration into their imaging platforms, and direct hospital adoption (B2H) in which the software is deployed into existing cath lab infrastructure.

Software licensing to imaging vendors/OEMs, plus a per-procedure model in which coronary angiography performed with Angio-FILM's radiation-reduction technology is intended to be recognized as a reimbursable clinical service under Korea's innovative medical technology reimbursement pathway, with planned expansion into reimbursement systems in other major markets.

Traction

Validation is retrospective and single-institution: a development dataset of 28,907 patients and more than 377,000 angiography video clips from Seoul National University Bundang Hospital, with reported blinded visual Turing test expert preference above 92%, VTT accuracy of 54% (chance = 50%), minimal lumen diameter correlation of 0.879 and MS-SSIM above 0.95. Prototype performance figures, including a 50% target dose reduction and seconds-scale restoration latency, are stated as engineering targets not yet validated clinically. The site solicits pilot assessments and licensing discussions.

Latest developments

An MFDS (Korea) submission is in progress, while FDA 510(k), CE-MDR and a multi-center randomized controlled trial are planned. Active development tracks include reducing 7.5→15 fps reconstruction latency to under one second and extending the interpolation model to 5→15 fps for further dose reduction.