G. Thibault

Towards Generalizability and Robustness in Biological Object Detection in Electron Microscopy Images

This study improves model generalizability in biological object detection using electron microscopy images through normalization and augmentation techniques, achieving a 29% average precision boost.

Decoding the complexity of delayed wound healing following Enterococcus faecalis infection

This study uses scRNA-seq to analyze infected and uninfected wounds in mice, revealing transcriptional and metabolic changes linked to E. faecalis infection and its impact on wound healing.

Efficient semi-supervised semantic segmentation of electron microscopy cancer images with sparse annotations

The paper compares deep learning models for segmenting nuclei and nucleoli in cancer EM images using sparse labels and semi-supervised learning, showing improvements in 3D Dice scores.

Multiplex imaging of localized prostate tumors reveals changes in mast cell type composition and spatial organization of AR-positive cells in the tumor microenvironment

Robust Segmentation of Cellular Ultrastructure on Sparsely Labeled 3D Electron Microscopy Images using Deep Learning

This paper develops a deep learning method to segment cellular ultrastructures in 3D EM images, achieving high accuracy for nuclei and nucleoli in cancer biopsies.

Noise Reconstruction & Removal Network: A New Architecture to Denoise FIB-SEM Images

The paper introduces an unsupervised Noise Reconstruction and Removal Network for denoising FIB-SEM images, achieving results comparable to or better than supervised methods.

Transfer Learning for Inference of Metastatic Origin from Whole Slide Histology

This paper uses transfer learning to predict metastatic origin from H&E-stained tumor images, achieving higher accuracy than models trained on primary or metastatic tumors alone.

An Omic and Multidimensional Spatial Atlas from Serial Biopsies of an Evolving Metastatic Breast Cancer

This paper presents an OMS Atlas from serial biopsies of a metastatic breast cancer patient, linking clinical data to multi-omic and spatial profiles to study therapy-associated heterogeneity and resistance mechanisms.

Yeast FIT2 homolog is necessary to maintain cellular proteostasis by regulating lipid homeostasis

This study shows that yeast FIT homologs SCS3 and YFT2 regulate lipid homeostasis and proteostasis by modulating the UPR and HSR pathways.

ER stress sensor Ire1 deploys a divergent transcriptional program in response to lipid bilayer stress

The paper identifies a lipid bilayer stress sensing mechanism in the UPR transducer Ire1 and shows it activates divergent transcriptional programs in yeast and C. elegans.

Predicting Primary Site of Secondary Liver Cancer with a Neural Estimator of Metastatic Origin (NEMO)

The paper introduces a deep learning model to predict the metastatic origin of secondary liver cancer from H&E stained tissue sections, achieving 90.2% accuracy comparable to expert pathologists.

A workflow for visualizing human cancer biopsies using large-format electron microscopy

This paper describes a workflow for 2D and 3D imaging of human cancer biopsies using electron microscopy to reveal ultrastructural details not visible with conventional 2D EM.

Glucose increases the lifespan of post-reproductive C. elegans independently of FOXO

This study shows that high glucose diet extends the lifespan of post-reproductive C. elegans through UPR activation, independent of FOXO.

A subset of UPR-induced transmembrane proteins are prematurely degraded during lipid perturbation

The paper shows that lipid perturbation causes premature degradation of some UPR-induced transmembrane proteins, including Sbh1, through an unconventional ERAD pathway.

Lipid perturbation-activated IRE-1 modulates autophagy and lipolysis during endoplasmic reticulum stress

This study identifies IRE-1's role in modulating autophagy and lipolysis during lipid perturbation-induced ER stress in C. elegans.