Classification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methods
Abstract Tissue microarray (TMA) core images are a treasure trove for artificial intelligence applications. However, a common problem of TMAs is multiple sectioning, which can change the content of the intended tissue core and requires re-labelling. Here, we investigate different ensemble methods fo...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2021-01-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-021-81352-y |
| _version_ | 1830144039076233216 |
|---|---|
| author | Huu-Giao Nguyen Annika Blank Heather E. Dawson Alessandro Lugli Inti Zlobec |
| author_facet | Huu-Giao Nguyen Annika Blank Heather E. Dawson Alessandro Lugli Inti Zlobec |
| author_sort | Huu-Giao Nguyen |
| collection | DOAJ |
| description | Abstract Tissue microarray (TMA) core images are a treasure trove for artificial intelligence applications. However, a common problem of TMAs is multiple sectioning, which can change the content of the intended tissue core and requires re-labelling. Here, we investigate different ensemble methods for colorectal tissue classification using high-throughput TMAs. Hematoxylin and Eosin (H&E) core images of 0.6 mm or 1.0 mm diameter from three international cohorts were extracted from 54 digital slides (n = 15,150 cores). After TMA core extraction and color enhancement, five different flows of independent and ensemble deep learning were applied. Training and testing data with 2144 and 13,006 cores included three classes: tumor, normal or “other” tissue. Ground-truth data were collected from 30 ngTMA slides (n = 8689 cores). A test augmentation is applied to reduce the uncertain prediction. Predictive accuracy of the best method, namely Soft Voting Ensemble of one VGG and one CapsNet models was 0.982, 0.947 and 0.939 for normal, “other” and tumor, which outperformed to independent or ensemble learning with one base-estimator. Our high-accuracy algorithm for colorectal tissue classification in high-throughput TMAs is amenable to images from different institutions, core sizes and stain intensity. It helps to reduce error in TMA core evaluations with previously given labels. |
| first_indexed | 2024-12-17T09:59:14Z |
| format | Article |
| id | doaj.art-b59e9d18919b4f2499e27497bb7e8341 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-12-17T09:59:14Z |
| publishDate | 2021-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-b59e9d18919b4f2499e27497bb7e83412022-12-21T21:53:21ZengNature PortfolioScientific Reports2045-23222021-01-0111111110.1038/s41598-021-81352-yClassification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methodsHuu-Giao Nguyen0Annika Blank1Heather E. Dawson2Alessandro Lugli3Inti Zlobec4Institute of Pathology, University of BernInstitute of Pathology, University of BernInstitute of Pathology, University of BernInstitute of Pathology, University of BernInstitute of Pathology, University of BernAbstract Tissue microarray (TMA) core images are a treasure trove for artificial intelligence applications. However, a common problem of TMAs is multiple sectioning, which can change the content of the intended tissue core and requires re-labelling. Here, we investigate different ensemble methods for colorectal tissue classification using high-throughput TMAs. Hematoxylin and Eosin (H&E) core images of 0.6 mm or 1.0 mm diameter from three international cohorts were extracted from 54 digital slides (n = 15,150 cores). After TMA core extraction and color enhancement, five different flows of independent and ensemble deep learning were applied. Training and testing data with 2144 and 13,006 cores included three classes: tumor, normal or “other” tissue. Ground-truth data were collected from 30 ngTMA slides (n = 8689 cores). A test augmentation is applied to reduce the uncertain prediction. Predictive accuracy of the best method, namely Soft Voting Ensemble of one VGG and one CapsNet models was 0.982, 0.947 and 0.939 for normal, “other” and tumor, which outperformed to independent or ensemble learning with one base-estimator. Our high-accuracy algorithm for colorectal tissue classification in high-throughput TMAs is amenable to images from different institutions, core sizes and stain intensity. It helps to reduce error in TMA core evaluations with previously given labels.https://doi.org/10.1038/s41598-021-81352-y |
| spellingShingle | Huu-Giao Nguyen Annika Blank Heather E. Dawson Alessandro Lugli Inti Zlobec Classification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methods Scientific Reports |
| title | Classification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methods |
| title_full | Classification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methods |
| title_fullStr | Classification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methods |
| title_full_unstemmed | Classification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methods |
| title_short | Classification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methods |
| title_sort | classification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methods |
| url | https://doi.org/10.1038/s41598-021-81352-y |
| work_keys_str_mv | AT huugiaonguyen classificationofcolorectaltissueimagesfromhighthroughputtissuemicroarraysbyensembledeeplearningmethods AT annikablank classificationofcolorectaltissueimagesfromhighthroughputtissuemicroarraysbyensembledeeplearningmethods AT heatheredawson classificationofcolorectaltissueimagesfromhighthroughputtissuemicroarraysbyensembledeeplearningmethods AT alessandrolugli classificationofcolorectaltissueimagesfromhighthroughputtissuemicroarraysbyensembledeeplearningmethods AT intizlobec classificationofcolorectaltissueimagesfromhighthroughputtissuemicroarraysbyensembledeeplearningmethods |