[Poster] Prediction of Oncotype DX using Distributional Regression Forests
Date:
Work conducted in collaboration with Zeina Al Masry, Clément Dombry & Christine Devalland.
Introduction Oncotype DX (ODX) is a multi-gene expression signature for breast cancer that provides prognostic and predictive breast cancer recurrence information for estrogen receptor (ER)-positive and HER2 negative patients. This test can predict if a chemotherapy treatment would be beneficial. However, this test is expensive and several studies have shown its link with certain clinico-pathological data. The result of the test is a score between 0 and 100 and guides clinicians regarding prescription of chemotherapy (e.g. for score higher than 25).
Aim - Predict the ODX score absed on histopathological variables.
Proposed methodology - Probabilistic forecast using Distributional Regression Forest (DRF).
This poster was presented at the Journées Math Bio Santé 2022 workshop.
Support of the presentation : Download See on HAL
Associated preprint : A new methodology to predict the oncotype scores based on clinico-pathological data with similar tumor profiles, Al Masry et al. (2023) link