Applications of Artificial Intelligence Based on Medical Imaging in Glioma: Current State and Future Challenges

Abstract

Glioma is one of the most fatal primary brain tumors, and it is well-known for its difficulty in diagnosis and management. Medical imaging techniques such as magnetic resonance imaging (MRI), positron emission tomography (PET), and spectral imaging can efficiently aid physicians in diagnosing, treating, and evaluating patients with gliomas. With the increasing clinical records and digital images, the application of artificial intelligence (AI) based on medical imaging has reduced the burden on physicians treating gliomas even further. This review will classify AI technologies and procedures used in medical imaging analysis. Additionally, we will discuss the applications of AI in glioma, including tumor segmentation and classification, prediction of genetic markers, and prediction of treatment response and prognosis, using MRI, PET, and spectral imaging. Despite the benefits of AI in clinical applications, several issues such as data management, incomprehension, safety, clinical efficacy evaluation, and ethical or legal considerations, remain to be solved. In the future, doctors and researchers should collaborate to solve these issues, with a particular emphasis on interdisciplinary teamwork.

Radiogenomics of adult intracranial gliomas after the 2021 World Health Organisation classification: a review of changes, challenges and opportunities

“…For example, the T2-fluid-attenuated inversion recovery (FLAIR) mismatch sign (Figure 1), is strongly predictive of an IDH mut , 1p/19q-intact status in a histological grade II-III glioma (4-7), and is the most specific conventional radiogenomic feature across all diffuse glioma types and grades (3). Earlier studies into conventional MRI features have been followed by studies into advanced MRI techniques, and subsequently by research into predicting genotype using artificial intelligence (AI) techniques, including radiomics (8) and deep learning (9).…”

“…This is analogous to the prediction of IDH mutation, which will vary depending on patient age, histological grade and the results of R132H-IDH1 immunohistochemistry (42). Further challenges with AI include reproducibility, given the inherent risk of over-fitting, and lack of transparency, due to the "black box" nature of most AI algorithms (9). Key strategies for improving translation of AI methods include obtaining multi-centre datasets, performing external validation, improving explainable AI methodologies and prospectively evaluating the incorporation of these techniques into clinical practice (8,9,43,44).…”