Radiomics analysis of non-enhancing lesions after Bevacizumab administration in recurrent glioblastoma
Motivation: Identifying non-contrast-enhancing tumor (nCET) in glioblastoma (GBM) is crucial for surgical treatment. However, T2/FLAIR images are unable to reliably distinguish nCET from vasogenic edema (VE). Goal(s): To develop an MRI-based method to identify nCET via radiomic features on areas where CET transformed to nCET after BEV administration. Approach: Radiomic features were analyzed in an exploratory cohort under BEV treatment (n=24) and a validation cohort who underwent 11C-methionine positron emission tomography (n=23). Results: A combination of three imaging features distinguished nCET from VE (AUC=0.91) with high positive predictive accuracy. Reconstructed MRI images also provided reasonable nCET visualization in newly diagnosed GBMs. Impact: Radiomic features of lesion where CET transformed into nCET after BEV administration may provide valuable information in predicting nCET in GBM.
Authors
- Yonehiro Kanemura (ORCID: https://orcid.org/0000-0002-1894-1824)
- Masato Saito (ORCID: https://orcid.org/0000-0001-9276-8882)
- Mishie Ann Tanino (ORCID: https://orcid.org/0000-0003-3370-0452)
- Haruhiko Kishima (ORCID: https://orcid.org/0000-0002-9041-2337)
- Satoru Hiroshima (ORCID: https://orcid.org/0000-0001-9868-7557)
- Nobuyuki Mitsui (ORCID: https://orcid.org/0000-0002-0788-7690)
- Yoshiko Okita (ORCID: https://orcid.org/0000-0002-2943-238X)
- Hideyuki Arita (ORCID: https://orcid.org/0000-0002-1126-8155)
- Manabu Kinoshita (ORCID: https://orcid.org/0000-0001-7923-6902)
- Takeshi Shimizu (ORCID: https://orcid.org/0000-0002-1344-7210)
- Hirotaka Sato (ORCID: https://orcid.org/0000-0003-2873-6514)
- Takahiro Sanada (ORCID: https://orcid.org/0000-0001-6910-1959)
Publication Details
- Journal
- Proceedings on CD-ROM - International Society for Magnetic Resonance in Medicine. Scientific Meeting and Exhibition/Proceedings of the International Society for Magnetic Resonance in Medicine, Scientific Meeting and Exhibition
- Published
- 2025-09-16
- DOI
- https://doi.org/10.58530/2025/2558
- Primary Topic
- Radiomics and Machine Learning in Medical Imaging
- Type
- article
- Field-Weighted Citation Impact
- 0.00