Machine learning and glioma imaging biomarkers

Abstract

Aim: To review how machine learning (ML) is applied to imaging biomarkers in neuro-oncology, in particular for diagnosis, prognosis, and treatment response monitoring.

Materials and methods: The PubMed and MEDLINE databases were searched for articles published before September 2018 using relevant search terms. The search strategy focused on articles applying ML to high-grade glioma biomarkers for treatment response monitoring, prognosis, and prediction.

Results: Magnetic resonance imaging (MRI) is typically used throughout the patient pathway because routine structural imaging provides detailed anatomical and pathological information and advanced techniques provide additional physiological detail. Using carefully chosen image features, ML is frequently used to allow accurate classification in a variety of scenarios. Rather than being chosen by human selection, ML also enables image features to be identified by an algorithm. Much research is applied to determining molecular profiles, histological tumour grade, and prognosis using MRI images acquired at the time that patients first present with a brain tumour. Differentiating a treatment response from a post-treatment-related effect using imaging is clinically important and also an area of active study (described here in one of two Special Issue publications dedicated to the application of ML in glioma imaging).

Conclusion: Although pioneering, most of the evidence is of a low level, having been obtained retrospectively and in single centres. Studies applying ML to build neuro-oncology monitoring biomarker models have yet to show an overall advantage over those using traditional statistical methods. Development and validation of ML models applied to neuro-oncology require large, well-annotated datasets, and therefore multidisciplinary and multi-centre collaborations are necessary.

Figure 1

The phases of radiomics are shown using explicit feature engineering. Some pre-processing steps are shown here: manual segmentation of hyperintense voxels associated with a glioblastoma in a T2-weighted image is performed. A mask is extracted, which undergoes quantisation. Some feature estimation steps are shown here: in this example, the pixels are made into three features that are topological descriptors of image heterogeneity (area is the number of white pixels = 1; perimeter around a white pixel = 4; genus is the number of rings subtracted from number of holes = 0). Note that deep learning uses implicit feature engineering and some of the feature estimation steps may not be required.

Figure 2

A longitudinal series of T1-weighted images after gadolinium administration. On the left is an image demonstrating a glioblastoma 1 month after surgery before chemoradiotherapy. In the middle is an image demonstrating the appearances 2 months after radiotherapy and concomitant chemotherapy. On the right is an image demonstrating the appearances 4 months after radiotherapy and concomitant chemotherapy. There was no new treatment between 2 and 4 months therefore this shows pseudoprogression occurred at 2 months.