Annonce en lien avec l’Action/le Réseau : aucun
Laboratoire/Entreprise : ICube / Université de Strasbourg
Durée : 36 mois
Contact : firstname.lastname@example.org
Date limite de publication : 2019-12-31
Digital Pathology is currently regarded as one of the most promising avenues of diagnostic medicine. With the recent advent of
Whole-Slide Imaging (WSI), the field of digital pathology produces daily a massive amount of images with related metadata (e.g.
patient information, diagnosis, treatment). In the context of colon cancer, such images could be used for both diagnosis and to
find some prognostic biomarkers. For example immune infiltrate are associated with better prognosis while high stromal contain
or tumor budding or poor differentiation status are associated with poorer outcome. Making these quantitative analysis is time
consuming for pathologist and frequently lack of reproducibility. To assist experts, automatic analysis of whole-slide images
(WSI) has been recently studied to predict survival outcomes or making tumor classification. The AiCOLO project aims to
contribute to the development of new artificial intelligence techniques trained on a large cohort of clinical annotated colon
cancer patients with a twofold objective. On the one hand, we will develop an innovative prognostic aid tool to automatically
classify tissues constituting WSI pathological slides and to enumerate the various known prognostic markers like TILS immune
infiltration, stromal contain or eosinophil count in the different areas. The method will be trained to find image patterns in tumor
tissue related to patients’ outcome. On the other hand, we will also propose a resolutely new approach to predict RAS and
BRAF genetic status from WSI. We aim to determine if artificial intelligence could detect patterns associated with such genetic
features and could outperform clinical or immune infiltrate variables.The idea is to study the activation layers of a deep network
trained to classify the patients in order to extract information to explain its decision.
The Engineering science, computer science and imaging laboratory (ICube, Strasbourg, France), associated with the Institute of Research in Computer Science, Mathematics, Automatics and Signal processing (IRIMAS, Mulhouse, France), opens a postdoctoral position for a computer scientist, in the field of artificial intelligence and histopathological whole slide images analysis, with a duration of 36 months (2019/12/01 – 2022/11/30).
In the context of the AiCOLO project described above, the appointee will work in close collaboration with the three partners of the project to develop the methodology for WSI analysis, spatial patterns extraction and the machine learning approach to classify
automatically the genetic mutation from HES images.
More specifically, the objective is to develop a complete methodology
enabling to assign a label to each image region. This problem
will be tackled by two complementary approaches: a pixel-based
method, in order to obtain a cartography of regions of interest for
the studied pathologies (inflammatory zones or tumors area) and
an object-based approach enabling to compute a list of biological
objects with their contour, localization and attributes. The main
workflow must be automatic in order to operate in an unsupervised
manner, which constitutes a crucial and challenging aspect of this
task. The WSI analysis system will rely both on previous work2, 3 and on novel techniques based on levellines decomposition of
an image4 and connected operators from mathematical morphology5 based on hierarchical representations. These latter methods
enable to analyze an image at the level of the connected components of its threshold sets (or other increasing transformation).
These methods are relevant in this context since (i) they enable to process an image in a contrast invariant way; (ii) they prevent to alter the contour of objects; (iii) they permit to compute object based attributes.
The work will also consists on the identification of genetic prognostic/predictive markers on HES slide. BRAF and RAS mutational status are mandatory required for the treatment of metastatic colon cancer. These markers are both prognostic and predictive of response to anti EGFR therapies. Recent data in lung cancer make the demonstration that gene mutations could affect the pattern of tumor cells on a lung cancer whole-slide image6. Training network using the presence or absence of mutated genes as a label revealed that there are certain genes whose mutational status can be predicted from image data alone: EGFR, STK11, FAT1, SETBP1, KRAS, and TP53 with good accuracy. The ability to quickly and inexpensively predict both the type of cancer and the gene mutations from histopathology images could be beneficial to the treatment of patients with cancer given the importance and impact of mutation in patient care. We propose to perform similar work on colon cancer and try to isolate feature detected by neural network to detect these mutation.
Profil du candidat :
The candidate should hold a PhD in computer science (preferably in computer vision or machine learning) and have excellent knowledge of the English language.
Formation et compétences requises :
Demonstrated experience in Python programming and Keras/Tensorflow libraries will be considered as an advantage.
Adresse d’emploi :
300 boulevard Sébastien Brant
Document attaché : 2019___AiCOLO_postdoc_position-compressé.pdf