Fast intraoperative detection of primary CNS lymphoma and differentiation from common CNS tumors using stimulated Raman histology and deep learning
David Reinecke,Nader Maarouf,Andrew Smith,Daniel Alber,John Markert,Nicolas K. Goff,Todd C. Hollon,Asadur Chowdury,Cheng Jiang,Xinhai Hou,Anna-Katharina Meissner,Gina Fuertjes,Maximilian I. Ruge,Daniel Ruess,Thomas Stehle,Abdulkader Al-Shughri,Lisa I. Koerner,Georg Widhalm,Thomas Roetzer-Pejrimovsky,John G. Golfinos,Matija Snuderl,Volker Neuschmelting,Daniel A. Orringer
DOI: https://doi.org/10.1101/2024.08.25.24312509
2024-08-26
Abstract:Accurate intraoperative diagnosis is crucial for differentiating between primary CNS lymphoma (PCNSL) and other CNS entities, guiding surgical decision-making, but represents significant challenges due to overlapping histomorphological features, time constraints, and differing treatment strategies. We combined stimulated Raman histology (SRH) with deep learning to address this challenge. We imaged unprocessed, label-free tissue samples intraoperatively using a portable Raman scattering microscope, generating virtual H&E-like images within less than three minutes. We developed a deep learning pipeline called RapidLymphoma based on a self-supervised learning strategy to (1) detect PCNSL, (2) differentiate from other CNS entities, and (3) test the diagnostic performance in a prospective international multicenter cohort and two additional independent test cohorts. We trained on 54,000 SRH patch images sourced from surgical resections and stereotactic-guided biopsies, including various CNS tumor/non-tumor lesions. Training and test data were collected from four tertiary international medical centers. The final histopathological diagnosis served as ground-truth. In the prospective test cohort of PCNSL and non-PCNSL entities (n=160), RapidLymphoma achieved an overall balanced accuracy of 97.81%, non-inferior to frozen section analysis in detecting PCNSL (100% vs. 78.94%). The additional test cohorts (n=420, n=59) reached balanced accuracy rates of 95.44% and 95.57% in differentiating IDH-wildtype diffuse gliomas and various brain metastasis from PCNSL. Visual heatmaps revealed capabilities to detect class-specific histomorphological key features.
RapidLymphoma proves reliable and valid for intraoperative PCNSL detection and differentiation from other CNS entities. It provides visual feedback within three minutes, enabling fast clinical decision-making and subsequent treatment strategy planning.