Interpretable survival prediction for colorectal cancer using deep learning
Abstract Deriving interpretable prognostic features from deep-learning-based prognostic histopathology models remains a challenge. In this study, we developed a deep learning system (DLS) for predicting disease-specific survival for stage II and III colorectal cancer using 3652 cases (27,300 slides)...
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2021-04-01
|
| Series: | npj Digital Medicine |
| Online Access: | https://doi.org/10.1038/s41746-021-00427-2 |
| _version_ | 1797427303487111168 |
|---|---|
| author | Ellery Wulczyn David F. Steiner Melissa Moran Markus Plass Robert Reihs Fraser Tan Isabelle Flament-Auvigne Trissia Brown Peter Regitnig Po-Hsuan Cameron Chen Narayan Hegde Apaar Sadhwani Robert MacDonald Benny Ayalew Greg S. Corrado Lily H. Peng Daniel Tse Heimo Müller Zhaoyang Xu Yun Liu Martin C. Stumpe Kurt Zatloukal Craig H. Mermel |
| author_facet | Ellery Wulczyn David F. Steiner Melissa Moran Markus Plass Robert Reihs Fraser Tan Isabelle Flament-Auvigne Trissia Brown Peter Regitnig Po-Hsuan Cameron Chen Narayan Hegde Apaar Sadhwani Robert MacDonald Benny Ayalew Greg S. Corrado Lily H. Peng Daniel Tse Heimo Müller Zhaoyang Xu Yun Liu Martin C. Stumpe Kurt Zatloukal Craig H. Mermel |
| author_sort | Ellery Wulczyn |
| collection | DOAJ |
| description | Abstract Deriving interpretable prognostic features from deep-learning-based prognostic histopathology models remains a challenge. In this study, we developed a deep learning system (DLS) for predicting disease-specific survival for stage II and III colorectal cancer using 3652 cases (27,300 slides). When evaluated on two validation datasets containing 1239 cases (9340 slides) and 738 cases (7140 slides), respectively, the DLS achieved a 5-year disease-specific survival AUC of 0.70 (95% CI: 0.66–0.73) and 0.69 (95% CI: 0.64–0.72), and added significant predictive value to a set of nine clinicopathologic features. To interpret the DLS, we explored the ability of different human-interpretable features to explain the variance in DLS scores. We observed that clinicopathologic features such as T-category, N-category, and grade explained a small fraction of the variance in DLS scores (R 2 = 18% in both validation sets). Next, we generated human-interpretable histologic features by clustering embeddings from a deep-learning-based image-similarity model and showed that they explained the majority of the variance (R 2 of 73–80%). Furthermore, the clustering-derived feature most strongly associated with high DLS scores was also highly prognostic in isolation. With a distinct visual appearance (poorly differentiated tumor cell clusters adjacent to adipose tissue), this feature was identified by annotators with 87.0–95.5% accuracy. Our approach can be used to explain predictions from a prognostic deep learning model and uncover potentially-novel prognostic features that can be reliably identified by people for future validation studies. |
| first_indexed | 2024-03-09T08:42:19Z |
| format | Article |
| id | doaj.art-da537cd907ad48cb854376055689af37 |
| institution | Directory Open Access Journal |
| issn | 2398-6352 |
| language | English |
| last_indexed | 2024-03-09T08:42:19Z |
| publishDate | 2021-04-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | npj Digital Medicine |
| spelling | doaj.art-da537cd907ad48cb854376055689af372023-12-02T16:33:12ZengNature Portfolionpj Digital Medicine2398-63522021-04-014111310.1038/s41746-021-00427-2Interpretable survival prediction for colorectal cancer using deep learningEllery Wulczyn0David F. Steiner1Melissa Moran2Markus Plass3Robert Reihs4Fraser Tan5Isabelle Flament-Auvigne6Trissia Brown7Peter Regitnig8Po-Hsuan Cameron Chen9Narayan Hegde10Apaar Sadhwani11Robert MacDonald12Benny Ayalew13Greg S. Corrado14Lily H. Peng15Daniel Tse16Heimo Müller17Zhaoyang Xu18Yun Liu19Martin C. Stumpe20Kurt Zatloukal21Craig H. Mermel22Google HealthGoogle HealthGoogle HealthMedical University of GrazMedical University of GrazGoogle HealthGoogle Health via Advanced ClinicalGoogle Health via Advanced ClinicalMedical University of GrazGoogle HealthGoogle HealthGoogle HealthGoogle HealthGoogle HealthGoogle HealthGoogle HealthGoogle HealthMedical University of GrazGoogle HealthGoogle HealthGoogle Health via Advanced ClinicalMedical University of GrazGoogle HealthAbstract Deriving interpretable prognostic features from deep-learning-based prognostic histopathology models remains a challenge. In this study, we developed a deep learning system (DLS) for predicting disease-specific survival for stage II and III colorectal cancer using 3652 cases (27,300 slides). When evaluated on two validation datasets containing 1239 cases (9340 slides) and 738 cases (7140 slides), respectively, the DLS achieved a 5-year disease-specific survival AUC of 0.70 (95% CI: 0.66–0.73) and 0.69 (95% CI: 0.64–0.72), and added significant predictive value to a set of nine clinicopathologic features. To interpret the DLS, we explored the ability of different human-interpretable features to explain the variance in DLS scores. We observed that clinicopathologic features such as T-category, N-category, and grade explained a small fraction of the variance in DLS scores (R 2 = 18% in both validation sets). Next, we generated human-interpretable histologic features by clustering embeddings from a deep-learning-based image-similarity model and showed that they explained the majority of the variance (R 2 of 73–80%). Furthermore, the clustering-derived feature most strongly associated with high DLS scores was also highly prognostic in isolation. With a distinct visual appearance (poorly differentiated tumor cell clusters adjacent to adipose tissue), this feature was identified by annotators with 87.0–95.5% accuracy. Our approach can be used to explain predictions from a prognostic deep learning model and uncover potentially-novel prognostic features that can be reliably identified by people for future validation studies.https://doi.org/10.1038/s41746-021-00427-2 |
| spellingShingle | Ellery Wulczyn David F. Steiner Melissa Moran Markus Plass Robert Reihs Fraser Tan Isabelle Flament-Auvigne Trissia Brown Peter Regitnig Po-Hsuan Cameron Chen Narayan Hegde Apaar Sadhwani Robert MacDonald Benny Ayalew Greg S. Corrado Lily H. Peng Daniel Tse Heimo Müller Zhaoyang Xu Yun Liu Martin C. Stumpe Kurt Zatloukal Craig H. Mermel Interpretable survival prediction for colorectal cancer using deep learning npj Digital Medicine |
| title | Interpretable survival prediction for colorectal cancer using deep learning |
| title_full | Interpretable survival prediction for colorectal cancer using deep learning |
| title_fullStr | Interpretable survival prediction for colorectal cancer using deep learning |
| title_full_unstemmed | Interpretable survival prediction for colorectal cancer using deep learning |
| title_short | Interpretable survival prediction for colorectal cancer using deep learning |
| title_sort | interpretable survival prediction for colorectal cancer using deep learning |
| url | https://doi.org/10.1038/s41746-021-00427-2 |
| work_keys_str_mv | AT ellerywulczyn interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT davidfsteiner interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT melissamoran interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT markusplass interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT robertreihs interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT frasertan interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT isabelleflamentauvigne interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT trissiabrown interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT peterregitnig interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT pohsuancameronchen interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT narayanhegde interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT apaarsadhwani interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT robertmacdonald interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT bennyayalew interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT gregscorrado interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT lilyhpeng interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT danieltse interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT heimomuller interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT zhaoyangxu interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT yunliu interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT martincstumpe interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT kurtzatloukal interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning AT craighmermel interpretablesurvivalpredictionforcolorectalcancerusingdeeplearning |