A transfer learning approach to facilitate ComBat-based harmonization of multicentre radiomic features in new datasets

Abstract

Purpose To facilitate the demonstration of the prognostic value of radiomics, multicenter radiomics studies are needed. Pooling radiomic features of such data in a statistical analysis is however challenging, as they are sensitive to the variability in scanner models, acquisition protocols and reconstruction settings, which is often unavoidable in a multicentre retrospective analysis. A statistical harmonization strategy called ComBat was utilized in radiomics studies to deal with the “center-effect”. The goal of the present work was to integrate a transfer learning (TL) technique within ComBat—and recently developed alternate versions of ComBat with improved flexibility (M-ComBat) and robustness (B-ComBat)–to allow the use of a previously determined harmonization transform to the radiomic feature values of new patients from an already known center. Material and methods The proposed TL approach were incorporated in the four versions of ComBat (standard, B, M, and B-M ComBat). The proposed approach was evaluated using a dataset of 189 locally advanced cervical cancer patients from 3 centers, with magnetic resonance imaging (MRI) and positron emission tomography (PET) images, with the clinical endpoint of predicting local failure. The impact performance of the TL approach was evaluated by comparing the harmonization achieved using only parts of the data to the reference (harmonization achieved using all the available data). It was performed through three different machine learning pipelines. Results The proposed TL technique was successful in harmonizing features of new patients from a known center in all versions of ComBat, leading to predictive models reaching similar performance as the ones developed using the features harmonized with all the data available. Conclusion The proposed TL approach enables applying a previously determined ComBat transform to new, previously unseen data.

Healthy Organs Uptake on Baseline 18F-FDG PET/CT as an Alternative to Total Metabolic Tumor Volume to Predict Event-Free Survival in Classical Hodgkin's Lymphoma

“…Although procedures were proposed to harmonize image quality ( 31 ), a dedicated reconstruction requiring raw data storage would be needed and would mostly be not feasible in a retrospective setting ( 15 ). To counteract this batch effect, the ComBat harmonization method, initially introduced in the field of genomics ( 32 ), has been proposed ( 15 ) and used ( 33 ). ComBat is a data-driven method that does not require phantom acquisitions to estimate the scanner effect but requires data from the different sites with sufficient sample size.…”

18F-FDG texture analysis predicts the pathological Fuhrman nuclear grade of clear cell renal cell carcinoma

“…On the other hand, PET radiomics features are very sensitive to changes in acquisition methods, image reconstruction algorithms, number of iterations or subsets, and acquisition time after injection [33][34][35], which largely limits the development of multi-center studies. Although there are currently some methods to eliminate the influence of multiple sites and different scanners on texture features [36,37], such as using conditional generative adversarial networks (cGANs) or ComBat, they are still in the research stage and there is no authoritative standardized guideline. Therefore, the sample size of patients who meet the research conditions is small.…”

Radiomics in Oncological PET Imaging: A Systematic Review—Part 2, Infradiaphragmatic Cancers, Blood Malignancies, Melanoma and Musculoskeletal Cancers

“…The retrospective nature of most of the available studies (>90%) and the lack of conservation of raw data prevent the performance of a standardized dedicated reconstruction protocol for radiomic purposes [ 2 , 168 ] and may limit the external validity of the proposed models. However, some solutions such as the ComBat harmonization method are starting to be used, with positive results [ 169 ].…”