Novel rapid intraoperative qualitative tumor detection by a residual convolutional neural network using label-free stimulated Raman scattering microscopy

Abstract

Determining the presence of tumor in biopsies and the decision-making during resections is often dependent on intraoperative rapid frozen-section histopathology. Recently, stimulated Raman scattering microscopy has been introduced to rapidly generate digital hematoxylin-and-eosin-stained-like images (stimulated Raman histology) for intraoperative analysis. To enable intraoperative prediction of tumor presence, we aimed to develop a new deep residual convolutional neural network in an automated pipeline and tested its validity. In a monocentric prospective clinical study with 94 patients undergoing biopsy, brain or spinal tumor resection, Stimulated Raman histology images of intraoperative tissue samples were obtained using a fiber-laser-based stimulated Raman scattering microscope. A residual network was established and trained in ResNetV50 to predict three classes for each image: (1) tumor, (2) non-tumor, and (3) low-quality. The residual network was validated on images obtained in three small random areas within the tissue samples and were blindly independently reviewed by a neuropathologist as ground truth. 402 images derived from 132 tissue samples were analyzed representing the entire spectrum of neurooncological surgery. The automated workflow took in a mean of 240 s per case, and the residual network correctly classified tumor (305/326), non-tumorous tissue (49/67), and low-quality (6/9) images with an inter-rater agreement of 89.6% (κ = 0.671). An excellent internal consistency was found among the random areas with 90.2% (Cα = 0.942) accuracy. In conclusion, the novel stimulated Raman histology-based residual network can reliably detect the microscopic presence of tumor and differentiate from non-tumorous brain tissue in resection and biopsy samples within 4 min and may pave a promising way for an alternative rapid intraoperative histopathological decision-making tool.

Technical report: surgical preparation of human brain tissue for clinical and basic research

“…The macroscopic pathological effect can be controlled by measuring the distance of the cortical access tissue to the pathologies (contrast-enhancing border or FLAIR signal) [ 23 ] and microscopically using histological methods (e.g., immuno/H&E stain, stimulated Raman histology [ 22 , 26 , 35 ]) to assess tumor infiltration (Fig. 6 e–f).…”

Live Stimulated Raman Histology for the Near-Instant Assessment of Central Nervous System Samples

“…As such, SRH imaging can be directly accepted by pathologist. Previous studies demonstrated SRH’s suitability for the analysis of freshly excised CNS surgical specimens which is particularly valuable for intraoperative pathological consultations. − ,− Indeed, the current workflow for IC requires the sample transport to a pathology department, processing of the specimen including cytological preparations (smear or touch preps) and frozen sections, both associated with rapid HE staining of the slides and their interpretation by a neuropathologist . This conventional procedure takes 20–30 min for each sample to be processed with potential negative impact for the patient since an increase surgery duration is associated with possible complications .…”

Advances, technological innovations, and future prospects in stereotactic brain biopsies

“…In a multicenter prospective clinical trial including 278 patients, Hollon et al combined SRH and artificial intelligence (i.e., deep convolutional neural networks trained on over 2.5 million SRH images) and showed that this diagnostic method was non-inferior to pathologist-based interpretation of conventional histological images (accuracy 94.6%) [ 21 ]. Using the same artificial intelligence process, Reinecke et al demonstrated that SRH can reliably detect the microscopic presence of tumor and discriminate from non-neoplastic brain tissue in stereotactic biopsy specimens [ 49 ].…”