Deep Learning–Enabled Diagnosis of Liver Adenocarcinoma

Albrecht, Thomas; Rossberg, Annik; Albrecht, Julia N.; Nicolay, Jan P.; Straub, Beate K.; Gerber, Tiemo S.; Albrecht, Michael von; Brinkmann, Fritz; Charbel, Alphonse; Schwab, Constantin; Schreck, Johannes; Brobeil, Alexander; Flechtenmacher, Christa; Winterfeld, Moritz von; Köhler, Bruno; Springfeld, Christoph; Mehrabi, Arianeb; Singer, Stephan; Vogel, Monika; Neumann, Olaf; Schirmacher, Peter; Weis, Cleo–Aron; Roessler, Stephanie; Kather, Jakob Nikolas; Goeppert, Benjamin

doi:10.1053/j.gastro.2023.07.026

Cited by 7 publications

(3 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Of note, we found transcriptomic classes differentially represented between surgical and biopsy cases, which highlights the importance of working with both tissue specimen. Few studies, mainly focusing on diagnostic tasks, have laid the groundwork for using biopsies and have demonstrated that encouraging deep-learning–based results can be obtained in this type of sample despite their size[18,35,36]. Thus, our model could be a useful molecular screening tool particularly in the context of biopsy in which the molecular analysis is not always possible because of the low amount of materials.…”

Section: Discussionmentioning

confidence: 99%

“…Artificial intelligence (AI) models, particularly deep neural networks are rapidly emerging in the medical field, especially in imaging [14–17]. With the development of digital pathology and wide access to digitised whole slide images (WSI), AI approaches can be used for classification tasks, as for example distinguishing cholangiocarcinoma from secondary forms of liver metastatic adenocarcinoma [18]. AI approaches can also be used to identify prognostic microscopic features and transcriptomic classification in hepatocellular carcinoma [19–23].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Self-supervised learning to predict intrahepatic cholangiocarcinoma transcriptomic classes on routine histology

Beaufrère,

Lazard,

Nicolle

et al. 2024

Preprint

View full text Add to dashboard Cite

Objective: The transcriptomic classification of intrahepatic cholangiocarcinomas (iCCA) has been recently refined from two to five classes, associated with pathological features, targetable genetic alterations and survival. Despite its prognostic and therapeutic value, the classification is not routinely used in the clinic because of technical limitations, including insufficient tissue material or the cost of molecular analyses. Here, we assessed a self-supervised learning (SSL) model for predicting iCCA transcriptomic classes on whole-slide digital histological images (WSIs). Design: Transcriptomic classes defined from RNAseq data were available for all samples. The SSL method, called Giga-SSL, was used to train our model on a discovery set of 766 biopsy slides (n=137 cases) and surgical samples (n=109 cases) from 246 patients in a five-fold cross-validation scheme. The model was validated in The Cancer Genome Atlas (TCGA) (n= 29) and a French external validation set (n=32). Results: Our model showed good to very good performance in predicting the four most frequent transcriptomic class in the discovery set (area under the curve [AUC]: 0.63-0.84), especially for the hepatic stem-like class (37% of cases, AUC 0.84). The model performed equally well in predicting these four transcriptomic classes in the two validation sets, with AUCs ranging from 0.76 to 0.80 in the TCGA set and 0.62 to 0.92 in the French external set. Conclusion: We developed and validated an SSL-based model for predicting iCCA transcriptomic classes on routine histological slides of biopsy and surgical samples, which may impact iCCA management by predicting prognosis and guiding the treatment strategy.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Self-supervised learning to predict intrahepatic cholangiocarcinoma transcriptomic classes on routine histology

Beaufrère,

Lazard,

Nicolle

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…If algorithms can reliably identify cancer cells on HE‐stained sections, this may save the high costs of immunohistochemistry, such as keratin stains on sentinel lymph nodes for breast of cervix carcinoma, 41 prostate biopsies, 41 or in identifying intrahepatic cholangiocarcinoma from colorectal liver metastases 42 . For example, in the UMC Utrecht, where HE slides of sentinel lymph nodes are assessed first, and stains are requested when these slides are morphologically negative, we spent over €30,000 on keratin stains for sentinel lymph node assessment in approximately 180 breast cancer patients.…”

Section: Pros Of Introducing Ai In Diagnostic Pathologymentioning

confidence: 99%

Pros and cons of artificial intelligence implementation in diagnostic pathology

van Diest,

Flach,

van Dooijeweert

et al. 2024

Histopathology

View full text Add to dashboard Cite

The rapid introduction of digital pathology has greatly facilitated development of artificial intelligence (AI) models in pathology that have shown great promise in assisting morphological diagnostics and quantitation of therapeutic targets. We are now at a tipping point where companies have started to bring algorithms to the market, and questions arise whether the pathology community is ready to implement AI in routine workflow. However, concerns also arise about the use of AI in pathology. This article reviews the pros and cons of introducing AI in diagnostic pathology.

show abstract

Künstliche Intelligenz in der Pathologie: Status quo und Zukunftsperspektiven

Foersch,

Schulz

2024

Onkologie

View full text Add to dashboard Cite

Deep Learning–Enabled Diagnosis of Liver Adenocarcinoma

Cited by 7 publications

References 33 publications

Self-supervised learning to predict intrahepatic cholangiocarcinoma transcriptomic classes on routine histology

Self-supervised learning to predict intrahepatic cholangiocarcinoma transcriptomic classes on routine histology

Pros and cons of artificial intelligence implementation in diagnostic pathology

Künstliche Intelligenz in der Pathologie: Status quo und Zukunftsperspektiven

Contact Info

Product

Resources

About