%0 Journal Article %1 Schrammen2022-yy %A Schrammen, Peter Leonard %A Ghaffari Laleh, Narmin %A Echle, Amelie %A Truhn, Daniel %A Schulz, Volkmar %A Brinker, Titus J %A Brenner, Hermann %A Chang-Claude, Jenny %A Alwers, Elizabeth %A Brobeil, Alexander %A Kloor, Matthias %A Heij, Lara R %A Jäger, Dirk %A Trautwein, Christian %A Grabsch, Heike I %A Quirke, Philip %A West, Nicholas P %A Hoffmeister, Michael %A Kather, Jakob Nikolas %D 2022 %I Wiley %J J. Pathol. %K topic_lifescience Lynch artificial cancer; colorectal computational deep digital instability intelligence; learning; microsatellite pathology; syndrome; %N 1 %P 50--60 %T Weakly supervised annotation-free cancer detection and prediction of genotype in routine histopathology %V 256 %X Deep learning is a powerful tool in computational pathology: it can be used for tumor detection and for predicting genetic alterations based on histopathology images alone. Conventionally, tumor detection and prediction of genetic alterations are two separate workflows. Newer methods have combined them, but require complex, manually engineered computational pipelines, restricting reproducibility and robustness. To address these issues, we present a new method for simultaneous tumor detection and prediction of genetic alterations: The Slide-Level Assessment Model (SLAM) uses a single off-the-shelf neural network to predict molecular alterations directly from routine pathology slides without any manual annotations, improving upon previous methods by automatically excluding normal and non-informative tissue regions. SLAM requires only standard programming libraries and is conceptually simpler than previous approaches. We have extensively validated SLAM for clinically relevant tasks using two large multicentric cohorts of colorectal cancer patients, Darmkrebs: Chancen der Verhütung durch Screening (DACHS) from Germany and Yorkshire Cancer Research Bowel Cancer Improvement Programme (YCR-BCIP) from the UK. We show that SLAM yields reliable slide-level classification of tumor presence with an area under the receiver operating curve (AUROC) of 0.980 (confidence interval 0.975, 0.984; n = 2,297 tumor and n = 1,281 normal slides). In addition, SLAM can detect microsatellite instability (MSI)/mismatch repair deficiency (dMMR) or microsatellite stability/mismatch repair proficiency with an AUROC of 0.909 (0.888, 0.929; n = 2,039 patients) and BRAF mutational status with an AUROC of 0.821 (0.786, 0.852; n = 2,075 patients). The improvement with respect to previous methods was validated in a large external testing cohort in which MSI/dMMR status was detected with an AUROC of 0.900 (0.864, 0.931; n = 805 patients). In addition, SLAM provides human-interpretable visualization maps, enabling the analysis of multiplexed network predictions by human experts. In summary, SLAM is a new simple and powerful method for computational pathology that could be applied to multiple disease contexts. \copyright 2021 The Authors. The Journal of Pathology published by John Wiley & Sons, Ltd. on behalf of The Pathological Society of Great Britain and Ireland.

@article{Schrammen2022-yy, abstract = {Deep learning is a powerful tool in computational pathology: it can be used for tumor detection and for predicting genetic alterations based on histopathology images alone. Conventionally, tumor detection and prediction of genetic alterations are two separate workflows. Newer methods have combined them, but require complex, manually engineered computational pipelines, restricting reproducibility and robustness. To address these issues, we present a new method for simultaneous tumor detection and prediction of genetic alterations: The Slide-Level Assessment Model (SLAM) uses a single off-the-shelf neural network to predict molecular alterations directly from routine pathology slides without any manual annotations, improving upon previous methods by automatically excluding normal and non-informative tissue regions. SLAM requires only standard programming libraries and is conceptually simpler than previous approaches. We have extensively validated SLAM for clinically relevant tasks using two large multicentric cohorts of colorectal cancer patients, Darmkrebs: Chancen der Verh{\"u}tung durch Screening (DACHS) from Germany and Yorkshire Cancer Research Bowel Cancer Improvement Programme (YCR-BCIP) from the UK. We show that SLAM yields reliable slide-level classification of tumor presence with an area under the receiver operating curve (AUROC) of 0.980 (confidence interval 0.975, 0.984; n = 2,297 tumor and n = 1,281 normal slides). In addition, SLAM can detect microsatellite instability (MSI)/mismatch repair deficiency (dMMR) or microsatellite stability/mismatch repair proficiency with an AUROC of 0.909 (0.888, 0.929; n = 2,039 patients) and BRAF mutational status with an AUROC of 0.821 (0.786, 0.852; n = 2,075 patients). The improvement with respect to previous methods was validated in a large external testing cohort in which MSI/dMMR status was detected with an AUROC of 0.900 (0.864, 0.931; n = 805 patients). In addition, SLAM provides human-interpretable visualization maps, enabling the analysis of multiplexed network predictions by human experts. In summary, SLAM is a new simple and powerful method for computational pathology that could be applied to multiple disease contexts. \copyright{} 2021 The Authors. The Journal of Pathology published by John Wiley \& Sons, Ltd. on behalf of The Pathological Society of Great Britain and Ireland.}, added-at = {2024-09-10T11:54:51.000+0200}, author = {Schrammen, Peter Leonard and Ghaffari Laleh, Narmin and Echle, Amelie and Truhn, Daniel and Schulz, Volkmar and Brinker, Titus J and Brenner, Hermann and Chang-Claude, Jenny and Alwers, Elizabeth and Brobeil, Alexander and Kloor, Matthias and Heij, Lara R and J{\"a}ger, Dirk and Trautwein, Christian and Grabsch, Heike I and Quirke, Philip and West, Nicholas P and Hoffmeister, Michael and Kather, Jakob Nikolas}, biburl = {https://puma.scadsai.uni-leipzig.de/bibtex/22cc2de04bd70fd6078b0767cbbd8d52c/scadsfct}, copyright = {http://creativecommons.org/licenses/by/4.0/}, interhash = {527fb8740b875b0d3db7247434800b1e}, intrahash = {2cc2de04bd70fd6078b0767cbbd8d52c}, journal = {J. Pathol.}, keywords = {topic_lifescience Lynch artificial cancer; colorectal computational deep digital instability intelligence; learning; microsatellite pathology; syndrome;}, language = {en}, month = jan, number = 1, pages = {50--60}, publisher = {Wiley}, timestamp = {2024-11-22T15:48:46.000+0100}, title = {Weakly supervised annotation-free cancer detection and prediction of genotype in routine histopathology}, volume = 256, year = 2022 }