Universität zu Köln

Reinecke, David ORCID: 0000-0002-3298-9517, von Spreckelsen, Niklas, Mawrin, Christian, Ion-Margineanu, Adrian, Fuertjes, Gina, Juenger, Stephanie T., Khalid, Florian, Freudiger, Christian W., Timmer, Marco, Ruge, Maximilian, I, Goldbrunner, Roland and Neuschmelting, Volker ORCID: 0000-0001-7527-6990 (2022). Novel rapid intraoperative qualitative tumor detection by a residual convolutional neural network using label-free stimulated Raman scattering microscopy. Acta Neuropathol. Commun., 10 (1). LONDON: BMC. ISSN 2051-5960

Full text not available from this repository.

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% (kappa = 0.671). An excellent internal consistency was found among the random areas with 90.2% (C alpha = 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.

Item Type:	Journal Article
Creators:	Creators Email ORCID ORCID Put Code Reinecke, David UNSPECIFIED orcid.org/0000-0002-3298-9517 UNSPECIFIED von Spreckelsen, Niklas UNSPECIFIED UNSPECIFIED UNSPECIFIED Mawrin, Christian UNSPECIFIED UNSPECIFIED UNSPECIFIED Ion-Margineanu, Adrian UNSPECIFIED UNSPECIFIED UNSPECIFIED Fuertjes, Gina UNSPECIFIED UNSPECIFIED UNSPECIFIED Juenger, Stephanie T. UNSPECIFIED UNSPECIFIED UNSPECIFIED Khalid, Florian UNSPECIFIED UNSPECIFIED UNSPECIFIED Freudiger, Christian W. UNSPECIFIED UNSPECIFIED UNSPECIFIED Timmer, Marco UNSPECIFIED UNSPECIFIED UNSPECIFIED Ruge, Maximilian, I UNSPECIFIED UNSPECIFIED UNSPECIFIED Goldbrunner, Roland UNSPECIFIED UNSPECIFIED UNSPECIFIED Neuschmelting, Volker UNSPECIFIED orcid.org/0000-0001-7527-6990 UNSPECIFIED
URN:	urn:nbn:de:hbz:38-670282
DOI:	10.1186/s40478-022-01411-x
Journal or Publication Title:	Acta Neuropathol. Commun.
Volume:	10
Number:	1
Date:	2022
Publisher:	BMC
Place of Publication:	LONDON
ISSN:	2051-5960
Language:	English
Faculty:	Unspecified
Divisions:	Unspecified
Subjects:	no entry
Uncontrolled Keywords:	Keywords Language CANCER-SURGERY; ULTRASOUND; RESECTION; GLIOMAS Multiple languages Neurosciences Multiple languages
URI:	http://kups.ub.uni-koeln.de/id/eprint/67028