GlioGuide: Deep Learning Platform for Therapy Response Prediction in Glioblastoma
Overview
De novo Glioblastoma (GBM), a highly aggressive cancer of the brain, claims over 200,000 lives worldwide annually. Tumor progression occurs in over 60% of cases and the incidence of pseudoprogression (a false appearance of tumor progression) is at 36%. However, physicians lack objective metrics to distinguish between these occurrences early on. Administering treatment for tumors experiencing pseudoprogression results in unnecessary harm to the patient while, contrarily, no treatment for patients experiencing true progression of GBM results in irreversible tumor progression. The goal of GlioGuide, therefore, was to be a platform for physicians to analyze the efficacy of standard-of-care glioblastoma treatment - surgery followed by chemoradiation - using only preoperative MRI scans, enabling a patient-centric adaptive treatment pathway. Multimodal data was compiled, including pre-and post-operative structural and perfusion MRI scans, genetic mutation statuses, clinical information (gender, age, and histopathology assessments), and extracted radiomic features. Patients in the dataset were moreover treated with temozolomide. This alkylating agent aims to inhibit tumor growth, but patients often confer resistance - thus a secondary goal for GlioGuide was to predict temozolomide resistance. GlioGuide utilized Fourier transform-based analysis for enhanced pre-operative MRI resolution, radiomic feature maps post-dimensionality reduction, and patient-specific growth kinetics (the tumor’s cellular proliferation and invasion). These components were integrated into a parallel-convolutional neural network to detect standard-of-care treatment responses, including temozolomide resistance. GlioGuide yielded an accuracy of over 95%, significantly higher than the current standard, and is the first accurate platform to steer treatment in De novo Glioblastoma using only preoperative metrics.
Competition history
- AJAS 2024
Related projects
ISEF · 2018
GlioVision: A Platform for the Automatic Assessment of Glioblastoma Tumor Features, Molecular Identity, and Gene Methylation from Histopathological Images Using Deep Learning
JSHS · 2020
GLIA-Deep: Glioblastoma Image Analysis using Deep Learning Convolutional Neural Networks to Accurately Classify Gene Methylation and Predict Drug Effectiveness Viraj Mehta
ISEF · 2023
Personalized Neurooncology Radiogenomic Classification and Recommendation Model With AI and MRI
ISEF · 2024
GB-NeuroAnalysis AI: Early Diagnosis of Glioblastoma Multiforme in MRI Images Through a Novel End-to-End Deep Learning Pipeline
ISEF · 2022
Artificial Intelligence Solution for Effective Treatment Planning for Brain Tumor Patients
ISEF · 2022
Creation of a Novel Machine Learning Model To Predict MGMT Promoter Methylation Status Using Multimodal MRI Images
ISEF · 2023
Spatiotemporal-Aware Glioblastoma Multiforme Tumor Growth Modeling With Deep Encoder-Decoder Networks
ISEF · 2024
G.L.O.W.: A Novel Hybrid Neural Network Approach for Glioblastoma Localization Using Carcinogenic Oxidative Stress Biomarkers
Closest projects by meaning, across every fair and year in the corpus.
Browse more like this
Source: AAAS Annual Meeting (Confex) / American Junior Academy of Science