Biology subjects

Lazard, T.

Publications and source records attributed to Lazard, T..

5 recordsLinked to original sources

bioRxiv2024

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

ObjectiveThe 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) DesignTranscriptomic 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). ResultsOur 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. ConclusionWe 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.

cancer biology↗

bioRxiv2023

Democratizing computational pathology: optimized Whole Slide Image representations for The Cancer Genome Atlas

Automatic analysis of hematoxylin and eosin (H&E) stained Whole Slide Images (WSI) bears great promise for computer assisted diagnosis and biomarker discovery. However, scarcity of annotated datasets leads to underperforming models. Furthermore, the size and complexity of the image data limit their integration into bioinformatic workflows and thus their adoption by the bioinformatics community. Here, we present Giga-SSL, a self-supervised method for learning WSI representations without any annotation. We show that applying a simple linear classifier on the Giga-SSL representations improves classification performance over the fully supervised alternative on five benchmarked tasks and across different datasets. Moreover, we observe a substantial performance increase for small datasets (average gain of 7 AUC point) and a doubling of the number of mutations predictable from WSIs in a pan-cancer setting (from 45 to 93). We make the WSI representations available, compressing the TCGA-FFPE images from 12TB to 23MB and enabling fast analysis on a laptop CPU. We hope this resource will facilitate multimodal data integration in order to analyze WSI in their genomic and transcriptomic context.

bioinformatics↗

bioRxiv2022

Neural network for the prediction of treatment response in triple negative breast cancer

AO_SCPLOWBSTRACTC_SCPLOWThe automatic analysis of stained histological sections is becoming increasingly popular. Deep Learning is today the method of choice for the computational analysis of such data, and has shown spectacular results for large datasets for a large variety of cancer types and prediction tasks. On the other hand, many scientific questions relate to small, highly specific cohorts. Such cohorts pose serious challenges for Deep Learning, typically trained on large datasets. In this article, we propose a modification of the standard nested cross-validation procedure for hyper-parameter tuning and model selection, dedicated to the analysis of small cohorts. We also propose a new architecture for the particularly challenging question of treatment prediction, and apply this workflow to the prediction of response to neoadjuvant chemotherapy for Triple Negative Breast Cancer.

bioinformatics↗

bioRxiv2022

Automatic grading of cervical biopsies by combining full and self-supervision

In computational pathology, predictive models from Whole Slide Images (WSI) mostly rely on Multiple Instance Learning (MIL), where the WSI are represented as a bag of tiles, each of which is encoded by a Neural Network (NN). Slide-level predictions are then achieved by building models on the agglomeration of these tile encodings. The tile encoding strategy thus plays a key role for such models. Current approaches include the use of encodings trained on unrelated data sources, full supervision or self-supervision. While self-supervised learning (SSL) exploits unlabeled data, it often requires large computational resources to train. On the other end of the spectrum, fully-supervised methods make use of valuable prior knowledge about the data but involve a costly amount of expert time. This paper proposes a framework to reconcile SSL and full supervision, showing that a combination of both provides efficient encodings, both in terms of performance and in terms of biological interpretability. On a recently organized challenge on grading Cervical Biopsies, we show that our mixed supervision scheme reaches high performance (weighted accuracy (WA): 0.945), outperforming both SSL (WA: 0.927) and transfer learning from ImageNet (WA: 0.877). We further shed light upon the internal representations that trigger classification results, providing a method to reveal relevant phenotypic patterns for grading cervical biopsies. We expect that the combination of full and self-supervision is an interesting strategy for many tasks in computational pathology and will be widely adopted by the field.

cancer biology↗

bioRxiv2021

Deep Learning identifies new morphological patterns of Homologous Recombination Deficiency in luminal breast cancers from whole slide images.

Homologous Recombination DNA-repair deficiency (HRD) is a well-recognized marker of platinum-salt and PARP inhibitor chemotherapies in ovarian and breast cancers (BC). Causing high genomic instability, HRD is currently determined by BRCA1/2 sequencing or by genomic signatures, but its morphological manifestation is not well understood. Deep Learning (DL) is a powerful machine learning technique that has been recently shown to be capable of predicting genomic signatures from stained tissue slides. However, DL is known to be sensitive to dataset biases and lacks interpretability. Here, we present and evaluate a strategy to control for biases in retrospective cohorts. We train a deep-learning model to predict the HRD in a controlled cohort with unprecedented accuracy (AUC: 0.86) and we develop a new visualization technique that allows for automatic extraction of new morphological features related to HRD. We analyze in detail the extracted morphological patterns that open new hypotheses on the phenotypic impact of HRD.

cancer biology↗