Screening programs for early detection of cancer such as colorectal and cervical cancer have led to an increased demand for histopathological analysis of biopsies. Advanced image analysis with Deep Learning has shown the potential to automate cancer detection in digital pathology whole-slide images. Particularly, techniques of weakly supervised learning can achieve whole-slide image classification without the need for tedious, manual annotations, using only slide-level labels. Here, we used data from n=12,580 whole-slide images from n=9,141 tissue blocks to train and validate a deep learning approach based on Neural Image Compression with Attention (NIC-A) and show how it can be leveraged to pre-screen (pre)malignant lesions in colorectal and cervical biopsies and to analyze duodenal biopsies for celiac disease. Our NIC-A classifies normal tissue, low-grade dysplasia, high-grade dysplasia and cancer in colon and uterine cervix, and identifies celiac disease in duodenal biopsies. We validated NIC-A for colon and cervix against a panel of four and three pathologists, respectively, on cohorts from two European centers. We show that the proposed approach reaches pathologist-level performance at detecting and classifying abnormalities, suggesting its potential to assist pathologists in pre-screening workflows by reducing workload in digital pathology routine diagnostics.

JvdL was a member of the advisory boards of Philips, the Netherlands and ContextVision, Sweden, and received research funding from Philips, the Netherlands, ContextVision, Sweden, and Sectra, Sweden in the last five years. He is chief scientific officer (CSO) and shareholder of Aiosyn BV, the Netherlands. FC was Chair of the Scientific and Medical Advisory Board of TRIBVN Healthcare, France, and received advisory board fees from TRIBVN Healthcare, France in the last five years. He is shareholder of Aiosyn BV, the Netherlands. Witali Aswolinskiy is currently employed at Paicon GmbH, Germany. All other authors declare no conflict of interest.

This project was funded by the European Unions Horizon 2020 research and innovation programme under grant agreement No 825292 (ExaMode, htttp://www.examode.eu/).

I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

Yes

The details of the IRB/oversight body that provided approval or exemption for the research described are given below:

The use of the patient data for the study was approved by the Committee on Research Involving Human Subjects of the Radboud University Medical Center (dossier number 2018-4764) and by the Ethical Committee of the Cannizzaro Hospital (approval number 4428, 12/12/2018). During their cancer treatment, patients were informed that left-over tissue could be used for research. Those, who did not authorize the use of their data for scientific research (i.e., decided to opt-out) were not included in the study. This research was performed following the Declaration of Helsinki.

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

Yes

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

Yes

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

Yes