Automatic Label-free Detection of Breast Cancer Using Nonlinear Multimodal Imaging and the Convolutional Neural Network ResNet50

Title data

Ali, Nairveen ; Quansah, Elsie ; Köhler, Katarina ; Meyer, Tobias ; Schmitt, Michael ; Popp, Jürgen ; Niendorf, Axel ; Bocklitz, Thomas:
Automatic Label-free Detection of Breast Cancer Using Nonlinear Multimodal Imaging and the Convolutional Neural Network ResNet50.
In: Translational Biophotonics. Vol. 1 (2019) Issue 1-2 . - e201900003.
ISSN 2627-1850
DOI: https://doi.org/10.1002/tbio.201900003

Official URL:

Abstract in another language

Breast cancer is the main cause of all female cancer deaths worldwide. Because of the lack of early symptoms, the early detection of breast cancer becomes challenging. The detection is performed by screening techniques in organized preventive examinations. A promising imaging technology that can detect bio-molecular alterations and can support the screening technologies by enhancing their low sensitivity, is nonlinear multimodal imaging. To detect these bio-molecular alterations machine learning algorithms are utilized. Our analysis starts by preprocessing the images and comparing them to the pathological diagnosis. We trained two classification models utilizing the deep convolutional neural network ResNet50. This network was either used as feature extractor or to be finetuned. Beside these two classification approaches, two data validation techniques were investigated: the leave-one-patient-out cross-validation (LOPO-CV) and the training-test validation. The best reported result of breast cancer detection was introduced by the finetuned ResNet50 network and LOPO-CV accounting to 86.23% mean-sensitivity.

Further data

Item Type:	Article in a journal
Refereed:	Yes
Keywords:	breast cancer imaging; coherent anti-Stokes Raman scattering (CARS); computer aided diagnosis; convolutional neural network; deep learning; image analysis; Nonlinear multimodal imaging; second-harmonic generation (SHG); two-photon excited fluorescence (TPEF)
Institutions of the University:	Faculties > Faculty of Mathematics, Physics und Computer Science > Department of Computer Science > Lehrstuhl Künstliche Intelligenz in der Mikroskopie und Spektroskopie > Lehrstuhl Künstliche Intelligenz in der Mikroskopie und Spektroskopie - Univ.-Prof. Dr. Thomas Wilhelm Bocklitz
Result of work at the UBT:	No
DDC Subjects:	500 Science > 530 Physics
Date Deposited:	22 May 2023 13:00
Last Modified:	22 May 2023 13:00
URI:	https://eref.uni-bayreuth.de/id/eprint/76269