A self-supervised multimodal deep learning approach to differentiate post-radiotherapy progression from pseudoprogression in glioblastoma

Abstract

Accurate differentiation of pseudoprogression (PsP) from True Progression (TP) following radiotherapy (RT) in glioblastoma patients is crucial for optimal treatment planning. However, this task remains challenging due to the overlapping imaging characteristics of PsP and TP. This study therefore proposes a multimodal deep-learning approach utilizing complementary information from routine anatomical MR images, clinical parameters, and RT treatment planning information for improved predictive accuracy. The approach utilizes a self-supervised Vision Transformer (ViT) to encode multi-sequence MR brain volumes to effectively capture both global and local context from the high dimensional input. The encoder is trained in a self-supervised upstream task on unlabeled glioma MRI datasets from the open BraTS2021, UPenn-GBM, and UCSF-PDGM datasets (n = 2317 MRI studies) to generate compact, clinically relevant representations from FLAIR and T1 post-contrast sequences. These encoded MR inputs are then integrated with clinical data and RT treatment planning information through guided cross-modal attention, improving progression classification accuracy. This work was developed using two datasets from different centers: the Burdenko Glioblastoma Progression Dataset (n = 59) for training and validation, and the GlioCMV progression dataset from the University Hospital Erlangen (UKER) (n = 20) for testing. The proposed method achieved competitive performance, with an AUC of 75.3%, outperforming the current state-of-the-art data-driven approaches. Importantly, the proposed approach relies solely on readily available anatomical MRI sequences, clinical data, and RT treatment planning information, enhancing its clinical feasibility. The proposed approach addresses the challenge of limited data availability for PsP and TP differentiation and could allow for improved clinical decision-making and optimized treatment plans for glioblastoma patients.

Fig. 1

The proposed workflow of the model. On the left-hand side is the self-supervised setup. On the right-hand side is the progression classification setup after the self-supervised training. In the inference phase, multi-parameter MRI volumes and clinical data are fed to the pre-trained ViT encoder and FC encoder, respectively.

Fig. 2

ROC curves evaluated on the external test set. Subplot (a) presents the ROC for the CNN-LSTM model (Jang et al.), while (b) displays the ROC for the CNN-SVM model (Akbari et al.). Subplot (c) illustrates the ROC for the random forest model (Sun et al.), and (d) features the proposed model. Subplot (e) highlights the mean ROC curves of the proposed model and existing baselines, and (f) shows the mean ROC curves depending upon the input modality.

Fig. 3

Mean feature importance analysis using SHAP in a 5-fold cross-validation setup.