Henning Höfener1, André Homeyer2, Nick Weiss3, Jesper Molin4, Claes F Lundström5, Horst K Hahn6. 1. Fraunhofer MEVIS, Am Fallturm 1, 28359, Bremen, Germany. Electronic address: henning.hoefener@mevis.fraunhofer.de. 2. Fraunhofer MEVIS, Am Fallturm 1, 28359, Bremen, Germany. Electronic address: andre.homeyer@mevis.fraunhofer.de. 3. Fraunhofer MEVIS, Am Fallturm 1, 28359, Bremen, Germany. Electronic address: nick.weiss@mevis.fraunhofer.de. 4. Sectra AB, Teknikringen 20, 58330, Linköping, Sweden. Electronic address: Jesper.Molin@sectra.com. 5. Sectra AB, Teknikringen 20, 58330, Linköping, Sweden; Center for Medical Image Science and Visualization, Linköping University, 58183, Linköping, Sweden. Electronic address: claes.lundstrom@liu.se. 6. Fraunhofer MEVIS, Am Fallturm 1, 28359, Bremen, Germany; Jacobs University, Campus Ring 1, 28759, Bremen, Germany. Electronic address: horst.hahn@mevis.fraunhofer.de.
Abstract
BACKGROUND: Deep convolutional neural networks have become a widespread tool for the detection of nuclei in histopathology images. Many implementations share a basic approach that includes generation of an intermediate map indicating the presence of a nucleus center, which we refer to as PMap. Nevertheless, these implementations often still differ in several parameters, resulting in different detection qualities. METHODS: We identified several essential parameters and configured the basic PMap approach using combinations of them. We thoroughly evaluated and compared various configurations on multiple datasets with respect to detection quality, efficiency and training effort. RESULTS: Post-processing of the PMap was found to have the largest impact on detection quality. Also, two different network architectures were identified that improve either detection quality or runtime performance. The best-performing configuration yields f1-measures of 0.816 on H&E stained images of colorectal adenocarcinomas and 0.819 on Ki-67 stained images of breast tumor tissue. On average, it was fully trained in less than 15,000 iterations and processed 4.15 megapixels per second at prediction time. CONCLUSIONS: The basic PMap approach is greatly affected by certain parameters. Our evaluation provides guidance on their impact and best settings. When configured properly, this simple and efficient approach can yield equal detection quality as more complex and time-consuming state-of-the-art approaches.
BACKGROUND: Deep convolutional neural networks have become a widespread tool for the detection of nuclei in histopathology images. Many implementations share a basic approach that includes generation of an intermediate map indicating the presence of a nucleus center, which we refer to as PMap. Nevertheless, these implementations often still differ in several parameters, resulting in different detection qualities. METHODS: We identified several essential parameters and configured the basic PMap approach using combinations of them. We thoroughly evaluated and compared various configurations on multiple datasets with respect to detection quality, efficiency and training effort. RESULTS: Post-processing of the PMap was found to have the largest impact on detection quality. Also, two different network architectures were identified that improve either detection quality or runtime performance. The best-performing configuration yields f1-measures of 0.816 on H&E stained images of colorectal adenocarcinomas and 0.819 on Ki-67 stained images of breast tumor tissue. On average, it was fully trained in less than 15,000 iterations and processed 4.15 megapixels per second at prediction time. CONCLUSIONS: The basic PMap approach is greatly affected by certain parameters. Our evaluation provides guidance on their impact and best settings. When configured properly, this simple and efficient approach can yield equal detection quality as more complex and time-consuming state-of-the-art approaches.
Authors: Martin Hailstone; Dominic Waithe; Tamsin J Samuels; Lu Yang; Ita Costello; Yoav Arava; Elizabeth Robertson; Richard M Parton; Ilan Davis Journal: Elife Date: 2020-05-19 Impact factor: 8.140