Intraoperative confocal laser scanning microscopy and artificial intelligence for optimized surgical excision of basal cell carcinoma

• Applicants: Professorin Dr. Daniela Hartmann, Ph.D.; Privatdozent Dr. Tobias Lasser
• Subject Area: Medical Informatics and Medical Bioinformatics ; Medical Physics, Biomedical Technology
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 536381342

The goal of this proposal is to improve surgical excision of basal cell carcinoma (BCC) by combining intraoperative confocal laser scanning microscopy (CLSM) with artificial intelligence (AI) algorithms. The use of AI-assisted CLSM may improve the accuracy and efficiency of histological analysis during surgery, resulting in better patient outcomes. In comparison to traditional histology, ex vivo CLSM is a non-invasive imaging technique that allows for direct histological examination of skin tissue in the operating theater without the need for complex sample preparation, harsh chemicals, or a dermatopathologist. Trained surgeons can quickly evaluate CLSM high-resolution images, which can therefore assist in tissue-sparing surgeries, especially in the face and neck region. Such tissue sparing techniques include the widely used micrographically controlled surgery and Mohs surgery. In micrographically controlled surgery, the excised tissue is sent to a histology laboratory for evaluation to ensure complete removal of the tumor. Hence, patients may have to undergo surgery for several days. In Mohs surgery, which is performed by specially trained surgeons in histological analysis, the excised tissue is analyzed immediately in the operating theater. By incorporating AI-based algorithms into CLSM, this proposal aims to improve image analysis, especially for tissue sparing surgeries, to accurately identify tumor and inflammatory cells, and reduce the need for time-consuming interpretation by dermatopathologists. This can also reduce the extent of invasive surgeries, allowing surgeons to target treatment specifically to the affected area while preserving healthy tissue. Another benefit of combining CLSM and AI-based algorithms is that it may reduce the scope of invasive surgeries. By accurately identifying cancerous cells, surgeons can target treatment directly to the affected area, minimizing the amount of healthy tissue that needs to be removed. Furthermore, it also leads to the finalization of the surgical treatment of patients more quickly as operations may be carried out in just one session, and reduces the time for wound healing as well as the risk of complications, such as wound infections. This would improve the patient experience and overall outcomes, allow tissue-sparing operations, more gentle operation techniques, and assist optimal wound healing. The project's goals include training AI algorithms to distinguish tumor and inflammatory tissues using ex vivo CLSM data and verifying them with histology. The trained algorithms will then be applied during BCC excision surgeries using intraoperative CLSM, with classical histology used for validation. By incorporating AI into the surgical process, surgeons can receive real-time feedback and automated image analysis, resulting in a more targeted and efficient approach to tissue analysis.

• DFG Programme: Research Grants