| ORIGINAL ARTICLE |
|
| Year : 2020 | Volume
: 11
| Issue : 1 | Page : 28 |
|
|
Colorectal cancer detection based on deep learning
Lin Xu1, Blair Walker2, Peir-In Liang3, Yi Tong1, Cheng Xu1, Yu Chun Su1, Aly Karsan4
1 GenerationsE Software Solutions, Inc., Surrey, Canada
2 Department of Pathology, St. Paul's Hospital, University of British Columbia, Vancouver, Canada
3 Kaohsiung Medical University Hospital, Kaohsiung Medical University, Kaohsiung, Taiwan
4 Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
Correspondence Address:
Dr. Aly Karsan
Room 9-111, 675 W 10th Ave, Vancouver, BC V5Z 1L3
Canada
 Source of Support: None, Conflict of Interest: None  | Check |
DOI: 10.4103/jpi.jpi_68_19
|
|
|
Introduction: The initial point in the diagnostic workup of solid tumors remains manual, with the assessment of hematoxylin and eosin (H&E)-stained tissue sections by microscopy. This is a labor-intensive step that requires attention to detail. In addition, diagnoses are influenced by an individual pathologist's knowledge and experience and may not always be reproducible between pathologists. Methods: We introduce a deep learning-based method in colorectal cancer detection and segmentation from digitized H&E-stained histology slides. Results: In this study, we demonstrate that this neural network approach produces median accuracy of 99.9% for normal slides and 94.8% for cancer slides compared to pathologist-based diagnosis on H&E-stained slides digitized from clinical samples. Conclusion: Given that our approach has very high accuracy on normal slides, use of neural network algorithms may provide a screening approach to save pathologist time in identifying tumor regions. We suggest that this new method may be a powerful assistant for colorectal cancer diagnostics.
|
|
|
|
| [FULL TEXT] [PDF]* |
|
 |
|
