Impact of intensity standardisation and ComBat batch size on clinical-radiomic prognostic models performance in a multi-centre study of patients with glioblastoma
Question ComBat harmonisation of RFs and intensity standardisation of MRI have not been thoroughly evaluated in multicentre, heterogeneous GBM data. Findings The addition of ComBat and ISTs can improve discrimination, relative model fit, and explained variance but degrades the calibration and stability of survival models. Clinical relevance Radiomics risk prediction models in real-world, multicentre contexts could be improved by ComBat and ISTs, however, this degrades calibration and prediction stability and this must be thoroughly investigated before patients can be accurately separated into different risk groups.
Authors
- Andrew Scarsbrook (ORCID: https://orcid.org/0000-0002-4243-032X)
- Kavi Fatania (ORCID: https://orcid.org/0000-0003-2421-1083)
- Hitesh Mistry (ORCID: https://orcid.org/0000-0002-9683-1903)
- Russell Frood (ORCID: https://orcid.org/0000-0003-2681-9922)
- Stuart Currie (ORCID: https://orcid.org/0000-0002-0154-6588)
- Susan Short (ORCID: https://orcid.org/0000-0003-4423-7256)
- James P.B. O’Connor (ORCID: https://orcid.org/0000-0002-4044-8497)
Institutions
- University of Leeds (GB)
- Institute of Cancer Research (GB)
- Leeds Teaching Hospitals NHS Trust (GB)
- University of Manchester (GB)
- Cancer Research UK Manchester Institute (GB)
- The Christie Hospital (GB)
Publication Details
- Journal
- European Radiology
- Published
- 2024-11-28
- DOI
- https://doi.org/10.1007/s00330-024-11168-7
- Primary Topic
- Radiomics and Machine Learning in Medical Imaging
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Wellcome Trust
- Cancer Research UK