Figure 12.

Workflow of AI-assisted GB analysis: (1) Acquisition of radiological image data through MRI scanning of GB patients. The images undergo pre-processing and standardization before being labeled by radiologists. (2) Data augmentation and pre-processing for ML models involves the use of different techniques to enhance data generalizability. The data undergo optional pre-processing for ML modeling, including feature extraction to extract explicit features and filter out unnecessary data. (3) ML modeling and training involve feeding the augmented and pre-processed data into different ML models for GB radiomic analysis training and validation. (4) AI-assisted clinical diagnosis/deployment involves predictions from the ML models for different medical goals, such as differential diagnosis and survival estimation [80].