Automated brain tumour identification using magnetic resonance imaging: a systematic review and meta-analysis

Background: Automated brain tumour identification facilitates diagnosis and treatment planning.We evaluate the performance of traditional machine learning (TML) and deep learning (DL) in brain tumour detection and segmentation, using MRI.Methods: In this systematic review and meta-analysis, we searched PubMed, Scopus, and Web of Science for studies from Jan 1, 2000, to May 8, 2021.Studies that developed or validated an automated method for brain tumour detection or segmentation using MRI were included.Study quality was assessed using the Checklist for Artificial Intelligence in Medical Imaging (CLAIM).Detection metaanalysis was performed using a unified hierarchical model, reported as pooled sensitivity and false positive rate (FPR).Segmentation studies were evaluated using a random effects model, reported as pooled dice similarity coefficient score (DSC).Sensitivity analysis was performed for studies undertaking external validation.The study was registered with PROSPERO, CRD42021247925.Results: Of 224 studies included in the systematic review, 46 segmentation and 38 detection studies were eligible for meta-analysis.In detection, DL achieved a lower FPR compared to TML; 0.018 (95% CI 0.011-0.028)and 0.048 (0.032-0.072) (p<0.001),respectively.In segmentation, DL had a higher DSC, particularly for tumour core; 0.80 (0.77-0.83) and 0.63 (0.56-0.71) (p<0.001),persisting on sensitivity analysis.Both manual and automated whole tumour (WT) segmentation had "good" (DSC>0.70)performance.Manual tumour core segmentation was superior to automated; 0.78 (0.69-0.86) and 0.64 (0.53-0.74) (p¼0.014),respectively.Only 30% of studies reported external validation.Conclusions: The comparable performance of automated to manual WT segmentation supports its integration into clinical practice.However, manual outperformance for sub-compartmental segmentation highlights the need for further development of automated methods in this area.Compared to TML, DL provided superior performance for detection and sub-compartmental segmentation.Improvements in the quality and design of studies, including external validation, are required for the interpretability and generalisability of automated models.

Abstract licence: CC BY-NC-ND

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.