Structured crowdsourcing enables convolutional segmentation of histology images
| Autor: | Mustafijur Rahman, Hazem S. E. Salem, Nada M. Elgazar, Anas M. Saad, Rokia Adel Sakr, Abo-Alela F. Younes, Mohamed Amgad, Mariam M. Khalaf, Inas A. Ruhban, Ahmad M. Elkashash, Ali Abdulkarim, Habiba Elfandy, Yahya Alagha, David E. Manthey, Ahmed F. Ismail, Mohamed Hosny Osman, David A. Gutman, Hagar Hussein, Duaa M. Younes, Ahmed M. Alhusseiny, Ahmed Gadallah, Lamees A. Atteya, Mai A. T. Elsebaie, Joumana Ahmed, Jonathan D. Beezley, Deepak Roy Chittajallu, Basma M. Zaki, Lee Cooper, Lamia S. Abo Elnasr, Maha A. T. Elsebaie, Salma Y. Fala |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2019 |
| Předmět: |
Statistics and Probability
Computer science Breast Neoplasms Crowdsourcing computer.software_genre Biochemistry 03 medical and health sciences Annotation 0302 clinical medicine Breast cancer medicine Humans Segmentation Molecular Biology 030304 developmental biology 0303 health sciences Contextual image classification business.industry Histological Techniques medicine.disease Original Papers Computer Science Applications Computational Mathematics ComputingMethodologies_PATTERNRECOGNITION Computational Theory and Mathematics Annotated Tissue 030220 oncology & carcinogenesis Artificial intelligence business Bioimage Informatics computer Natural language processing Algorithms |
| Zdroj: | Bioinformatics |
| ISSN: | 1367-4811 1367-4803 |
| Popis: | Motivation While deep-learning algorithms have demonstrated outstanding performance in semantic image segmentation tasks, large annotation datasets are needed to create accurate models. Annotation of histology images is challenging due to the effort and experience required to carefully delineate tissue structures, and difficulties related to sharing and markup of whole-slide images. Results We recruited 25 participants, ranging in experience from senior pathologists to medical students, to delineate tissue regions in 151 breast cancer slides using the Digital Slide Archive. Inter-participant discordance was systematically evaluated, revealing low discordance for tumor and stroma, and higher discordance for more subjectively defined or rare tissue classes. Feedback provided by senior participants enabled the generation and curation of 20 000+ annotated tissue regions. Fully convolutional networks trained using these annotations were highly accurate (mean AUC=0.945), and the scale of annotation data provided notable improvements in image classification accuracy. Availability and Implementation Dataset is freely available at: https://goo.gl/cNM4EL. Supplementary information Supplementary data are available at Bioinformatics online. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|