Next Article in Journal
Three-Dimensional Organotypic Cultures Reshape the microRNAs Transcriptional Program in Breast Cancer Cells
Next Article in Special Issue
Novel Human Artificial Intelligence Hybrid Framework Pinpoints Thyroid Nodule Malignancy and Identifies Overlooked Second-Order Ultrasonographic Features
Previous Article in Journal
Two-Front War on Cancer—Targeting TAM Receptors in Solid Tumour Therapy
 
 
Article

iMIL4PATH: A Semi-Supervised Interpretable Approach for Colorectal Whole-Slide Images

1
Institute for Systems and Computer Engineering, Technology and Science (INESC TEC), 4200-465 Porto, Portugal
2
Faculty of Engineering, University of Porto (FEUP), 4200-465 Porto, Portugal
3
IMP Diagnostics, 4150-146 Porto, Portugal
4
School of Medicine and Biomedical Sciences, University of Porto (ICBAS), 4050-313 Porto, Portugal
5
Cancer Biology and Epigenetics Group, IPO-Porto, 4200-072 Porto, Portugal
6
Department of Pathology, IPO-Porto, 4200-072 Porto, Portugal
*
Author to whom correspondence should be addressed.
These authors contributed equally to this work.
Academic Editor: David Wong
Cancers 2022, 14(10), 2489; https://doi.org/10.3390/cancers14102489
Received: 22 April 2022 / Revised: 13 May 2022 / Accepted: 17 May 2022 / Published: 18 May 2022
(This article belongs to the Special Issue Image Analysis and Computational Pathology in Cancer Diagnosis)
Nowadays, colorectal cancer is the third most incident cancer worldwide and, although it can be detected by imaging techniques, diagnosis is always based on biopsy samples. This assessment includes neoplasia grading, a subjective yet important task for pathologists. With the growing availability of digital slides, the development of robust and high-performance computer vision algorithms can help to tackle such a task. In this work, we propose an approach to automatically detect and grade lesions in colorectal biopsies with high sensitivity. The presented model attempts to support slide decision reasoning in terms of the spatial distribution of lesions, focusing the pathologist’s attention on key areas. Thus, it can be integrated into clinical practice as a second opinion or as a flag for details that may have been missed at first glance.
Colorectal cancer (CRC) diagnosis is based on samples obtained from biopsies, assessed in pathology laboratories. Due to population growth and ageing, as well as better screening programs, the CRC incidence rate has been increasing, leading to a higher workload for pathologists. In this sense, the application of AI for automatic CRC diagnosis, particularly on whole-slide images (WSI), is of utmost relevance, in order to assist professionals in case triage and case review. In this work, we propose an interpretable semi-supervised approach to detect lesions in colorectal biopsies with high sensitivity, based on multiple-instance learning and feature aggregation methods. The model was developed on an extended version of the recent, publicly available CRC dataset (the CRC+ dataset with 4433 WSI), using 3424 slides for training and 1009 slides for evaluation. The proposed method attained 90.19% classification ACC, 98.8% sensitivity, 85.7% specificity, and a quadratic weighted kappa of 0.888 at slide-based evaluation. Its generalisation capabilities are also studied on two publicly available external datasets. View Full-Text
Keywords: weakly supervised learning; semi-supervised learning; multiple-instance learning; interpretability; computational pathology; colorectal cancer weakly supervised learning; semi-supervised learning; multiple-instance learning; interpretability; computational pathology; colorectal cancer
Show Figures

Figure 1

MDPI and ACS Style

Neto, P.C.; Oliveira, S.P.; Montezuma, D.; Fraga, J.; Monteiro, A.; Ribeiro, L.; Gonçalves, S.; Pinto, I.M.; Cardoso, J.S. iMIL4PATH: A Semi-Supervised Interpretable Approach for Colorectal Whole-Slide Images. Cancers 2022, 14, 2489. https://doi.org/10.3390/cancers14102489

AMA Style

Neto PC, Oliveira SP, Montezuma D, Fraga J, Monteiro A, Ribeiro L, Gonçalves S, Pinto IM, Cardoso JS. iMIL4PATH: A Semi-Supervised Interpretable Approach for Colorectal Whole-Slide Images. Cancers. 2022; 14(10):2489. https://doi.org/10.3390/cancers14102489

Chicago/Turabian Style

Neto, Pedro C., Sara P. Oliveira, Diana Montezuma, João Fraga, Ana Monteiro, Liliana Ribeiro, Sofia Gonçalves, Isabel M. Pinto, and Jaime S. Cardoso. 2022. "iMIL4PATH: A Semi-Supervised Interpretable Approach for Colorectal Whole-Slide Images" Cancers 14, no. 10: 2489. https://doi.org/10.3390/cancers14102489

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop