Co-Localized Dermoscopy and LC-OCT for AI-Assisted Margin Assessment of Basal Cell Carcinoma: Development of a...

Background/Objectives: The surgical treatment of basal cell carcinoma (BCC) remains challenging due to the time-consuming, expensive and invasive nature of Mohs micrographic surgery. The objective is to develop a standardized protocol for managing diagnosis, surgery, and margin control within a single patient visit. Methods: Several protocols were tested to establish a "BCC-One-Stop-Shop", combining in vivo and ex vivo margin mapping of BCC, pre- and postoperatively using Line-field confocal optical coherence tomography (LC-OCT). We introduce an algorithm enabling real-time localization of LC-OCT acquisitions on a previously acquired dermoscopy image. Additionally, an artificial intelligence model provides a BCC probability score based on LC-OCT images. Together, the co-localization algorithm and AI BCC model generate a color-coded visualization of the tumor within the dermoscopy image, allowing precise pre-operative in vivo margin assessment. Results: We found our protocol, the implementation of the co-localization tool and the AI model, to be quick to apply, easy to learn and helpful regarding the initial determination of BCC tumor margins. Patients responded positively to the recognizable visualization of the disease. Conclusions: Pre- and postoperative margin mapping using LC-OCT imaging appears to be effective and feasible and could reduce time, costs, resources, excision sizes and patient burden by sparing additional excision steps in micrographic surgery. The integration of real-time co-localization and the AI-calculated probability score represent meaningful and practical enhancements for routine clinical use. To further evaluate the efficacy and safety of the BCC-One-Stop-Shop-Method and the newly introduced device features, larger-scale studies are warranted and are currently being conducted.