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

Neto, Pedro C.; Oliveira, Sara P.; Montezuma, Diana; Fraga, Jorge Gerardo Pereira; Monteiro, Ana Raquel; Ribeiro, Liliana; Gonçalves, Sofia; Pinto, Isabel; Cardoso, Jaime S.

doi:10.3390/cancers14102489

Cited by 13 publications

(21 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…and had a high consensus with experts. Subsequent studies have also applied MIL for Gleason grading [ 80 , 81 ], grading dysplasia of various cancers (i.e., normal, low-grade, and high-grade) [ 82 , 83 ], and grading of colorectal cancer (i.e., low-grade and high-grade) [ 84 ].…”

Section: Related Workmentioning

confidence: 99%

Contrastive Multiple Instance Learning: An Unsupervised Framework for Learning Slide-Level Representations of Whole Slide Histopathology Images without Labels

Tavolara

Gürcan

Niazi

2022

Cancers

View full text Add to dashboard Cite

Recent methods in computational pathology have trended towards semi- and weakly-supervised methods requiring only slide-level labels. Yet, even slide-level labels may be absent or irrelevant to the application of interest, such as in clinical trials. Hence, we present a fully unsupervised method to learn meaningful, compact representations of WSIs. Our method initially trains a tile-wise encoder using SimCLR, from which subsets of tile-wise embeddings are extracted and fused via an attention-based multiple-instance learning framework to yield slide-level representations. The resulting set of intra-slide-level and inter-slide-level embeddings are attracted and repelled via contrastive loss, respectively. This resulted in slide-level representations with self-supervision. We applied our method to two tasks— (1) non-small cell lung cancer subtyping (NSCLC) as a classification prototype and (2) breast cancer proliferation scoring (TUPAC16) as a regression prototype—and achieved an AUC of 0.8641 ± 0.0115 and correlation (R2) of 0.5740 ± 0.0970, respectively. Ablation experiments demonstrate that the resulting unsupervised slide-level feature space can be fine-tuned with small datasets for both tasks. Overall, our method approaches computational pathology in a novel manner, where meaningful features can be learned from whole-slide images without the need for annotations of slide-level labels. The proposed method stands to benefit computational pathology, as it theoretically enables researchers to benefit from completely unlabeled whole-slide images.

show abstract

Section: Related Workmentioning

confidence: 99%

Contrastive Multiple Instance Learning: An Unsupervised Framework for Learning Slide-Level Representations of Whole Slide Histopathology Images without Labels

Tavolara

Gürcan

Niazi

2022

Cancers

View full text Add to dashboard Cite

show abstract

“…On the other hand, it is easier to label the dataset at the slide level. The inclusion of detailed spatial annotations on approximately 10% of the dataset has been shown to positively impact the performance of deep learning algorithms [17,31]. To fully leverage the potential of spatial and slide labels, we propose a deep learning pipeline, based on previous approaches [17,31], using mixed supervision.…”

Section: Problem Definitionmentioning

confidence: 99%

“…The inclusion of detailed spatial annotations on approximately 10% of the dataset has been shown to positively impact the performance of deep learning algorithms [17,31]. To fully leverage the potential of spatial and slide labels, we propose a deep learning pipeline, based on previous approaches [17,31], using mixed supervision. Each slide, S is composed of a set of tiles T s,n , where s represents the index of the slide and n ∈ {1, • • • , n s } the tile number.…”

Section: Problem Definitionmentioning

confidence: 99%

“…The spectrum of large-scale CRC/CRS datasets is slowly increasing due to the contributions of several researchers. Two datasets that have been recently introduced in the literature are the CRS1K [17] and CRS4K [31] datasets. Since the latter is an extension of the former with roughly four times more slides, it will be the baseline dataset for the remaining of this document.…”

Section: Datasetsmentioning

confidence: 99%

“…After this analysis, the five most severe tiles are selected for training. The number of selected tiles was chosen in accordance with previous studies [31]. These five tiles per slide are used to train the proposed model for one more epoch.…”

Section: Weakly-supervised Learningmentioning

confidence: 99%

See 2 more Smart Citations

A CAD System for Colorectal Cancer from WSI: A Clinically Validated Interpretable ML-based Prototype

Neto¹,

Montezuma²,

Oliveira³

et al. 2023

Preprint

View full text Add to dashboard Cite

The integration of Artificial Intelligence (AI) and Digital Pathology has been increasing over the past years. Nowadays, applications of deep learning (DL) methods to diagnose cancer from whole-slide images (WSI) are, more than ever, a reality within different research groups. Nonetheless, the development of these systems was limited by a myriad of constraints regarding the lack of training samples, the scaling difficulties, the opaqueness of DL methods, and, more importantly, the lack of clinical validation. As such, we propose a system designed specifically for the diagnosis of colorectal samples. The construction of such a system consisted of four stages: (1) a careful data collection and annotation process, which resulted in one of the largest WSI colorectal samples datasets; (2) the design of an interpretable

show abstract

HER2GAN: Overcome the Scarcity of HER2 Breast Cancer Dataset Based on Transfer Learning and GAN Model

Mirimoghaddam,

Majidpour,

Pashaei

et al. 2024

Clinical Breast Cancer

View full text Add to dashboard Cite

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

Cited by 13 publications

References 29 publications

Contrastive Multiple Instance Learning: An Unsupervised Framework for Learning Slide-Level Representations of Whole Slide Histopathology Images without Labels

Contrastive Multiple Instance Learning: An Unsupervised Framework for Learning Slide-Level Representations of Whole Slide Histopathology Images without Labels

A CAD System for Colorectal Cancer from WSI: A Clinically Validated Interpretable ML-based Prototype

HER2GAN: Overcome the Scarcity of HER2 Breast Cancer Dataset Based on Transfer Learning and GAN Model

Contact Info

Product

Resources

About