Glioblastoma Multi‑Omics and Machine Learning for Clinical Trials | McMaster & IAG Partnership
McMaster University’s Stem Cell and Cancer Research Institute and Image Analysis Group have joined forces to develop multi‑omics‑based machine learning models for glioblastoma multiforme, aiming to improve risk stratification and support more precise treatment decisions.
Key Themes:
- Multidisciplinary collaboration across neurosurgery, oncology, radiology and computational science to study GBM biology and outcomes.
- Integrated radiomics, proteomics and transcriptomics to characterise tumour behaviour and recurrence patterns beyond standard imaging and pathology.
- Development of genotypic–phenotypic biomarkers to support risk stratification and prognostication in one of the most lethal brain tumours.
- Foundation for future GBM trials that use advanced imaging and multi‑omics to identify high‑risk patients and tailor therapy.
Reach out to our team to discuss your clinical trial imaging needs: contact@ia-grp.com
Partnership to advance glioblastoma research
McMaster University’s Stem Cell and Cancer Research Institute has partnered with Image Analysis Group to conduct a multidisciplinary research project in glioblastoma multiforme (GBM). Led by Dr Sheila Singh’s lab, the collaboration brings together expertise from neurosurgery, oncology, radiology and quantitative imaging to develop new prognostication models for this highly aggressive brain tumor.
The joint program focuses on building machine learning models that can better capture GBM’s biological heterogeneity and link it to patient outcomes, with the long‑term goal of informing trial design and treatment strategies in neuro‑oncology.
Multi‑Omics and Machine Learning Approach
The collaboration centers on a fully integrated multi‑omics approach. Quantitative radiomics features extracted from advanced imaging are combined with proteomic and transcriptomic profiles to capture the microstructural and functional characteristics of glioblastoma. By bringing these data types together, the team aims to identify biomarkers that distinguish more aggressive tumors and predict recurrence patterns more accurately than imaging or histology alone.
Radiomics contributes high‑dimensional information from MRI and other imaging modalities, while proteomics and transcriptomics provide insight into protein expression and gene activity in GBM tissue. Machine learning models integrate these signals to build prognostic signatures that can support patient stratification and early assessment of treatment efficacy.
Towards Precision Medicine in GBM
With this novel multi‑omics approach, the team aims to discover new predictive biomarkers of recurrence patterns in GBM, the most common adult primary malignant brain tumor. The work recognizes that tumor behavior is shaped by a spectrum of biological factors, including gene expression, secreted proteins and microstructural changes visible on imaging.
By studying these features individually and then integrating them, the collaboration seeks to classify glioblastomas more accurately and move towards treating each patient based on the unique characteristics of their tumor and its recurrence pattern. As Dr Faiq Shaikh, IAG’s Head of Oncology & Radiomics, notes, the goal is to harness machine learning to develop genotypic–phenotypic biomarkers that enable risk stratification and improved prognostication in GBM.
Senior neuroradiology expertise, including co‑investigators such as Dr Sotirios Bisdas, supports robust imaging protocol design and feature extraction, ensuring that multi‑omics models are grounded in high‑quality neuro‑oncology imaging data.
IAG’s role in neuro‑oncology imaging and analytics
IAG brings advanced analytics and imaging trial experience into the field of neurobiology, with the aim of improving understanding of brain diseases and accelerating drug development in GBM and other CNS tumours. The company leverages its DYNAMIKA™ platform and radiomics expertise to manage imaging data, extract quantitative features and support integration with clinical and molecular datasets.
While proteomics, genomics and radiomics have been used individually in previous oncology studies, this collaboration represents one of the first fully integrated multi‑omics efforts within IAG’s neuro‑oncology portfolio to support the understanding and management of a malignant brain disease. For sponsors, this demonstrates IAG’s capability to partner on complex biomarker programmes that sit at the intersection of imaging, omics and machine learning, and to translate these insights into practical strategies for GBM trial design and patient stratification.
Exploring ways to de‑risk your glioblastoma development programme with better imaging and biomarker strategies?
Discover how multi‑omics, radiomics and machine learning can enhance prognostication, refine patient stratification and support clearer go/no‑go decisions in GBM trials.
Book a personalized DYNAMIKA™ demo today to see how our neuro‑oncology imaging and analytics can strengthen your study design and help your teams act with greater confidence.
About Image Analysis Group (IAG)
Image Analysis Group (IAG) is a global imaging clinical research organization (iCRO) focused on de-risking and accelerating drug development through advanced imaging science and AI. IAG designs and runs imaging-centric clinical trials across oncology, immunology, neurology, rare diseases, musculoskeletal conditions, fertility and women’s health, partnering closely with biotech and pharma sponsors from early proof-of-concept through pivotal and real-world evidence studies.
Through its proprietary, cloud-native DYNAMIKA™ platform, IAG integrates centralized imaging workflows, blinded independent review, expert readers and AI-powered imaging biomarkers to deliver high-quality, regulator-ready imaging endpoints and decision-enabling insights that optimize trial design, reduce operational risk and unlock the full value of imaging data.
Media contact:
Marie Fussell
IAG Communications Lead
Press@ia-grp.com
Follow Image Analysis Group (IAG) on LinkedIn for weekly updates.
