Prediction of heart transplant rejection from routine pathology slides with self-supervised deep learning
Standard
Prediction of heart transplant rejection from routine pathology slides with self-supervised deep learning. / Seraphin, Tobias Paul; Luedde, Mark; Roderburg, Christoph; van Treeck, Marko; Scheider, Pascal; Buelow, Roman D; Boor, Peter; Loosen, Sven H; Provaznik, Zdenek; Mendelsohn, Daniel; Berisha, Filip; Magnussen, Christina; Westermann, Dirk; Luedde, Tom; Brochhausen, Christoph; Sossalla, Samuel; Kather, Jakob Nikolas.
In: European heart journal. Digital health, Vol. 4, No. 3, 05.2023, p. 265-274.Research output: SCORING: Contribution to journal › SCORING: Journal article › Research › peer-review
Harvard
APA
Vancouver
Bibtex
}
RIS
TY - JOUR
T1 - Prediction of heart transplant rejection from routine pathology slides with self-supervised deep learning
AU - Seraphin, Tobias Paul
AU - Luedde, Mark
AU - Roderburg, Christoph
AU - van Treeck, Marko
AU - Scheider, Pascal
AU - Buelow, Roman D
AU - Boor, Peter
AU - Loosen, Sven H
AU - Provaznik, Zdenek
AU - Mendelsohn, Daniel
AU - Berisha, Filip
AU - Magnussen, Christina
AU - Westermann, Dirk
AU - Luedde, Tom
AU - Brochhausen, Christoph
AU - Sossalla, Samuel
AU - Kather, Jakob Nikolas
N1 - © The Author(s) 2023. Published by Oxford University Press on behalf of the European Society of Cardiology.
PY - 2023/5
Y1 - 2023/5
N2 - AIMS: One of the most important complications of heart transplantation is organ rejection, which is diagnosed on endomyocardial biopsies by pathologists. Computer-based systems could assist in the diagnostic process and potentially improve reproducibility. Here, we evaluated the feasibility of using deep learning in predicting the degree of cellular rejection from pathology slides as defined by the International Society for Heart and Lung Transplantation (ISHLT) grading system.METHODS AND RESULTS: We collected 1079 histopathology slides from 325 patients from three transplant centres in Germany. We trained an attention-based deep neural network to predict rejection in the primary cohort and evaluated its performance using cross-validation and by deploying it to three cohorts. For binary prediction (rejection yes/no), the mean area under the receiver operating curve (AUROC) was 0.849 in the cross-validated experiment and 0.734, 0.729, and 0.716 in external validation cohorts. For a prediction of the ISHLT grade (0R, 1R, 2/3R), AUROCs were 0.835, 0.633, and 0.905 in the cross-validated experiment and 0.764, 0.597, and 0.913; 0.631, 0.633, and 0.682; and 0.722, 0.601, and 0.805 in the validation cohorts, respectively. The predictions of the artificial intelligence model were interpretable by human experts and highlighted plausible morphological patterns.CONCLUSION: We conclude that artificial intelligence can detect patterns of cellular transplant rejection in routine pathology, even when trained on small cohorts.
AB - AIMS: One of the most important complications of heart transplantation is organ rejection, which is diagnosed on endomyocardial biopsies by pathologists. Computer-based systems could assist in the diagnostic process and potentially improve reproducibility. Here, we evaluated the feasibility of using deep learning in predicting the degree of cellular rejection from pathology slides as defined by the International Society for Heart and Lung Transplantation (ISHLT) grading system.METHODS AND RESULTS: We collected 1079 histopathology slides from 325 patients from three transplant centres in Germany. We trained an attention-based deep neural network to predict rejection in the primary cohort and evaluated its performance using cross-validation and by deploying it to three cohorts. For binary prediction (rejection yes/no), the mean area under the receiver operating curve (AUROC) was 0.849 in the cross-validated experiment and 0.734, 0.729, and 0.716 in external validation cohorts. For a prediction of the ISHLT grade (0R, 1R, 2/3R), AUROCs were 0.835, 0.633, and 0.905 in the cross-validated experiment and 0.764, 0.597, and 0.913; 0.631, 0.633, and 0.682; and 0.722, 0.601, and 0.805 in the validation cohorts, respectively. The predictions of the artificial intelligence model were interpretable by human experts and highlighted plausible morphological patterns.CONCLUSION: We conclude that artificial intelligence can detect patterns of cellular transplant rejection in routine pathology, even when trained on small cohorts.
U2 - 10.1093/ehjdh/ztad016
DO - 10.1093/ehjdh/ztad016
M3 - SCORING: Journal article
C2 - 37265858
VL - 4
SP - 265
EP - 274
JO - European heart journal. Digital health
JF - European heart journal. Digital health
SN - 2634-3916
IS - 3
ER -