Radiomic Clustering in Glioblastoma and Its Link to Gene Expression
Author Information
Author(s): Wroblewski Tadeusz H., Karabacak Mert, Seah Carina, Yong Raymund L., Margetis Konstantinos
Primary Institution: SUNY Downstate Health Sciences University
Hypothesis
Consensus clustering of radiomic features will offer a non-invasive method to identify molecularly relevant GBM groups, which can be further characterized through gene expression analysis.
Conclusion
Consensus clustering of radiomic features revealed unique gene expression profiles in the GBM cohort, suggesting potential for non-invasive characterization of tumor biology.
Supporting Evidence
- Glioblastoma is an aggressive tumor with poor survival outcomes.
- Three clusters were identified based on radiomic features.
- Cluster 1 showed significant differentially expressed genes associated with GBM prognosis.
Takeaway
This study shows that by looking at MRI scans, we can group glioblastoma tumors into different types based on their genes, which might help doctors choose better treatments.
Methodology
Patients from TCGA and CPTAC datasets were analyzed using consensus clustering of MRI-based radiomic features to identify gene expression profiles.
Potential Biases
Batch effects from using data from multiple datasets were addressed but may still pose risks.
Limitations
The study had a limited sample size and used expression data from GBMs at a single time point, which may not capture changes over time.
Participant Demographics
Average age was 60 years, with 60% male and 83% White racial identity.
Statistical Information
P-Value
p<0.05
Confidence Interval
[11.7–16.7]
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website