ArcticAI: A Deep Learning Platform for Rapid and Accurate Histological Assessment of Intraoperative Tumor Margins

Abstract

Successful treatment of solid cancers relies on complete surgical excision of the tumor either for definitive treatment or before adjuvant therapy. Radial sectioning of the resected tumor and surrounding tissue is the most common form of intra-operative and post-operative margin assessment. However, this technique samples only a tiny fraction of the available tissue and therefore may result in incomplete excision of the tumor, increasing the risk of recurrence and distant metastasis and decreasing survival. Repeat procedures, chemotherapy, and other resulting treatments pose significant morbidity, mortality, and fiscal costs for our healthcare system. Mohs Micrographic Surgery (MMS) is used for the removal of basal cell and squamous cell carcinoma utilizing frozen sections for real-time margin assessment while assessing 100% of the peripheral and deep margins, resulting in a recurrence rate of less than one percent. Real-time assessment in many tumor types is constrained by tissue size and complexity and the time to process tissue and evaluate slides while a patient is under general anesthesia. In this study, we developed an artificial intelligence (AI) platform, ArcticAI, which augments the surgical workflow to improve efficiency by reducing rate-limiting steps in tissue preprocessing and histological assessment through automated mapping and orientation of tumor to the surgical specimen. Using basal cell carcinoma (BCC) as a model system, the results demonstrate that ArcticAI can provide effective grossing recommendations, accurately identify tumor on histological sections, map tumor back onto the surgical resection map, and automate pathology report generation resulting in seamless communication between the surgical pathology laboratory and surgeon. AI-augmented-surgical excision workflows may make real-time margin assessment for the excision of more complex and challenging tumor types more accessible, leading to more streamlined and accurate tumor removal while increasing healthcare delivery efficiency.

A deep learning algorithm to detect cutaneous squamous cell carcinoma on frozen sections in Mohs micrographic surgery: A retrospective assessment

“…However, cutting additional tissue sections could introduce further bottlenecks when using AI tools for intraoperative assessment. Yet, some prior research has suggested that AI might expedite the decision on when to section blocks 6,21,22 . The prevalence of low‐quality sections can differ among institution, thus influencing their impact.…”

“…For example, previous studies have shown that suboptimal specimen quality, like fragmented tissue samples, has limited impact on identifying positive margins. This is partly because it signals the Mohs surgeon to possibly extract additional sections from the tissue block or resected tissue for a more comprehensive evaluation 6,16 . However, cutting additional tissue sections could introduce further bottlenecks when using AI tools for intraoperative assessment.…”

“…This is partly because it signals the Mohs surgeon to possibly extract additional sections from the tissue block or resected tissue for a more comprehensive evaluation. 6,16 However, cutting additional tissue sections could introduce further bottlenecks when using AI tools for intraoperative assessment. Yet, some prior research has suggested that AI might expedite the decision on when to section blocks.…”

“…Despite this, our team believes that identifying holes/fragmentation is perhaps more important in this clinical context, as additional tissue sections are required if any epidermis is missing or if any significant holes or processing artefact is present. In a previous work, we developed an algorithm to detect holes on frozen section tissue with the intent of flagging slides that require further reprocessing 6 …”

“…Currently, patient demand far outweighs the capacity of the dermatology workforce (Association of American Medical Colleges, AAMC), making early treatment more difficult 5 . Machine learning models exist to detect basal cell carcinoma (BCC), 5–7 but given the complexities and variable morphologies of cSCC, similar algorithms are yet to be developed for this tumour type 8 . Our study presents an algorithm to detect cSCC on whole slide images (WSI) of frozen sections obtained in MMS 9 .…”