Deep learning-based quantification of NAFLD/NASH progression in human liver biopsies
Abstract Non-alcoholic fatty liver disease (NAFLD) affects about 24% of the world's population. Progression of early stages of NAFLD can lead to the more advanced form non-alcoholic steatohepatitis (NASH), and ultimately to cirrhosis or liver cancer. The current gold standard for diagnosis and...
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2022-11-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-022-23905-3 |
| _version_ | 1817970324482818048 |
|---|---|
| author | Fabian Heinemann Peter Gross Svetlana Zeveleva Hu Sheng Qian Jon Hill Anne Höfer Danny Jonigk Anna Mae Diehl Manal Abdelmalek Martin C. Lenter Steven S. Pullen Paolo Guarnieri Birgit Stierstorfer |
| author_facet | Fabian Heinemann Peter Gross Svetlana Zeveleva Hu Sheng Qian Jon Hill Anne Höfer Danny Jonigk Anna Mae Diehl Manal Abdelmalek Martin C. Lenter Steven S. Pullen Paolo Guarnieri Birgit Stierstorfer |
| author_sort | Fabian Heinemann |
| collection | DOAJ |
| description | Abstract Non-alcoholic fatty liver disease (NAFLD) affects about 24% of the world's population. Progression of early stages of NAFLD can lead to the more advanced form non-alcoholic steatohepatitis (NASH), and ultimately to cirrhosis or liver cancer. The current gold standard for diagnosis and assessment of NAFLD/NASH is liver biopsy followed by microscopic analysis by a pathologist. The Kleiner score is frequently used for a semi-quantitative assessment of disease progression. In this scoring system the features of active injury (steatosis, inflammation, and ballooning) and a separated fibrosis score are quantified. The procedure is time consuming for pathologists, scores have limited resolution and are subject to variation. We developed an automated deep learning method that provides full reproducibility and higher resolution. The system was established with 296 human liver biopsies and tested on 171 human liver biopsies with pathologist ground truth scores. The method is inspired by the way pathologist's analyze liver biopsies. First, the biopsies are analyzed microscopically for the relevant histopathological features. Subsequently, histopathological features are aggregated to a per-biopsy score. Scores are in the identical numeric range as the pathologist’s ballooning, inflammation, steatosis, and fibrosis scores, but on a continuous scale. Resulting scores followed a pathologist's ground truth (quadratic weighted Cohen’s κ on the test set: for steatosis 0.66, for inflammation 0.24, for ballooning 0.43, for fibrosis 0.62, and for the NAFLD activity score (NAS) 0.52. Mean absolute errors on a test set: for steatosis 0.29, for inflammation 0.53, for ballooning 0.61, for fibrosis 0.78, and for the NAS 0.77). |
| first_indexed | 2024-04-13T20:32:45Z |
| format | Article |
| id | doaj.art-d03e21bdb66a47cf83dafba82490bd38 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-13T20:32:45Z |
| publishDate | 2022-11-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-d03e21bdb66a47cf83dafba82490bd382022-12-22T02:31:08ZengNature PortfolioScientific Reports2045-23222022-11-0112111110.1038/s41598-022-23905-3Deep learning-based quantification of NAFLD/NASH progression in human liver biopsiesFabian Heinemann0Peter Gross1Svetlana Zeveleva2Hu Sheng Qian3Jon Hill4Anne Höfer5Danny Jonigk6Anna Mae Diehl7Manal Abdelmalek8Martin C. Lenter9Steven S. Pullen10Paolo Guarnieri11Birgit Stierstorfer12Drug Discovery Sciences, Boehringer Ingelheim Pharma GmbH & Co. KGDrug Discovery Sciences, Boehringer Ingelheim Pharma GmbH & Co. KGCardiometabolic Diseases Research, Boehringer Ingelheim Pharmaceuticals, Inc.Cardiometabolic Diseases Research, Boehringer Ingelheim Pharmaceuticals, Inc.Global Computational Biology and Digital Sciences, Boehringer Ingelheim Pharmaceuticals, Inc.Institute of Pathology, Hannover Medical School, and the German Center for Lung Research (DZL), Biomedical Research in Endstage and Obstructive Lung Disease Hannover (BREATH)Institute of Pathology, Hannover Medical School, and the German Center for Lung Research (DZL), Biomedical Research in Endstage and Obstructive Lung Disease Hannover (BREATH)Duke Department of Medicine, GastroenterologyDuke Department of Medicine, GastroenterologyDrug Discovery Sciences, Boehringer Ingelheim Pharma GmbH & Co. KGCardiometabolic Diseases Research, Boehringer Ingelheim Pharmaceuticals, Inc.Cardiometabolic Diseases Research, Boehringer Ingelheim Pharmaceuticals, Inc.Drug Discovery Sciences, Boehringer Ingelheim Pharma GmbH & Co. KGAbstract Non-alcoholic fatty liver disease (NAFLD) affects about 24% of the world's population. Progression of early stages of NAFLD can lead to the more advanced form non-alcoholic steatohepatitis (NASH), and ultimately to cirrhosis or liver cancer. The current gold standard for diagnosis and assessment of NAFLD/NASH is liver biopsy followed by microscopic analysis by a pathologist. The Kleiner score is frequently used for a semi-quantitative assessment of disease progression. In this scoring system the features of active injury (steatosis, inflammation, and ballooning) and a separated fibrosis score are quantified. The procedure is time consuming for pathologists, scores have limited resolution and are subject to variation. We developed an automated deep learning method that provides full reproducibility and higher resolution. The system was established with 296 human liver biopsies and tested on 171 human liver biopsies with pathologist ground truth scores. The method is inspired by the way pathologist's analyze liver biopsies. First, the biopsies are analyzed microscopically for the relevant histopathological features. Subsequently, histopathological features are aggregated to a per-biopsy score. Scores are in the identical numeric range as the pathologist’s ballooning, inflammation, steatosis, and fibrosis scores, but on a continuous scale. Resulting scores followed a pathologist's ground truth (quadratic weighted Cohen’s κ on the test set: for steatosis 0.66, for inflammation 0.24, for ballooning 0.43, for fibrosis 0.62, and for the NAFLD activity score (NAS) 0.52. Mean absolute errors on a test set: for steatosis 0.29, for inflammation 0.53, for ballooning 0.61, for fibrosis 0.78, and for the NAS 0.77).https://doi.org/10.1038/s41598-022-23905-3 |
| spellingShingle | Fabian Heinemann Peter Gross Svetlana Zeveleva Hu Sheng Qian Jon Hill Anne Höfer Danny Jonigk Anna Mae Diehl Manal Abdelmalek Martin C. Lenter Steven S. Pullen Paolo Guarnieri Birgit Stierstorfer Deep learning-based quantification of NAFLD/NASH progression in human liver biopsies Scientific Reports |
| title | Deep learning-based quantification of NAFLD/NASH progression in human liver biopsies |
| title_full | Deep learning-based quantification of NAFLD/NASH progression in human liver biopsies |
| title_fullStr | Deep learning-based quantification of NAFLD/NASH progression in human liver biopsies |
| title_full_unstemmed | Deep learning-based quantification of NAFLD/NASH progression in human liver biopsies |
| title_short | Deep learning-based quantification of NAFLD/NASH progression in human liver biopsies |
| title_sort | deep learning based quantification of nafld nash progression in human liver biopsies |
| url | https://doi.org/10.1038/s41598-022-23905-3 |
| work_keys_str_mv | AT fabianheinemann deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT petergross deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT svetlanazeveleva deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT hushengqian deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT jonhill deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT annehofer deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT dannyjonigk deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT annamaediehl deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT manalabdelmalek deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT martinclenter deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT stevenspullen deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT paologuarnieri deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies AT birgitstierstorfer deeplearningbasedquantificationofnafldnashprogressioninhumanliverbiopsies |