1.
Systematic Review of Artificial Intelligence for Abnormality Detection in High-volume Neuroimaging and Subgroup Meta-analysis for Intracranial Hemorrhage Detection
Agarwal, S., Wood, D., Grzeda, M., Suresh, C., Din, M., Cole, J., Modat, M., Booth, T. C.
Clinical neuroradiology. 2023;:1-14
Abstract
PURPOSE Most studies evaluating artificial intelligence (AI) models that detect abnormalities in neuroimaging are either tested on unrepresentative patient cohorts or are insufficiently well-validated, leading to poor generalisability to real-world tasks. The aim was to determine the diagnostic test accuracy and summarise the evidence supporting the use of AI models performing first-line, high-volume neuroimaging tasks. METHODS Medline, Embase, Cochrane library and Web of Science were searched until September 2021 for studies that temporally or externally validated AI capable of detecting abnormalities in first-line computed tomography (CT) or magnetic resonance (MR) neuroimaging. A bivariate random effects model was used for meta-analysis where appropriate. This study was registered on PROSPERO as CRD42021269563. RESULTS Out of 42,870 records screened, and 5734 potentially eligible full texts, only 16 studies were eligible for inclusion. Included studies were not compromised by unrepresentative datasets or inadequate validation methodology. Direct comparison with radiologists was available in 4/16 studies and 15/16 had a high risk of bias. Meta-analysis was only suitable for intracranial hemorrhage detection in CT imaging (10/16 studies), where AI systems had a pooled sensitivity and specificity 0.90 (95% confidence interval [CI] 0.85-0.94) and 0.90 (95% CI 0.83-0.95), respectively. Other AI studies using CT and MRI detected target conditions other than hemorrhage (2/16), or multiple target conditions (4/16). Only 3/16 studies implemented AI in clinical pathways, either for pre-read triage or as post-read discrepancy identifiers. CONCLUSION The paucity of eligible studies reflects that most abnormality detection AI studies were not adequately validated in representative clinical cohorts. The few studies describing how abnormality detection AI could impact patients and clinicians did not explore the full ramifications of clinical implementation.
2.
The accuracy of aneurysm size in predicting rebleeding after subarachnoid hemorrhage: a meta-analysis
Yu Z, Zheng J, Guo R, Li M, Li H, Ma L, You C
Neurological sciences : official journal of the Italian Neurological Society and of the Italian Society of Clinical Neurophysiology. 2020
Abstract
BACKGROUND Aneurysmal subarachnoid hemorrhage (SAH) is a severe cerebrovascular disease. Rebleeding is an independent predictor of unfavorable outcome after aneurysmal SAH. However, the accuracy of aneurysm size for predicting rebleeding after aneurysmal SAH is still unclear. Hence, we conducted this meta-analysis to evaluate the predictive accuracy of large aneurysm for rebleeding after SAH. METHODS We performed a literature search in PubMed and Embase. Original studies about aneurysm size and rebleeding after SAH were included. Two reviewers completed data extraction and quality assessment. Pooled sensitivity, specificity, and their 95% confidence intervals (CIs) of large aneurysm for predicting rebleeding were calculated and shown in a forest plot. The overall accuracy of large aneurysm for predicting rebleeding after SAH was shown using a summary receiver operating characteristic (SROC) curve. Publication bias were assessed using Deeks' funnel plot. RESULTS A total of ten studies with 3889 patients met eligibility criteria and were included in this meta-analysis. The pooled sensitivity and specificity of large aneurysm for predicting rebleeding were 0.39 (95% CI 0.25-0.56) and 0.75 (95% CI 0.67-0.82), respectively. The area under SROC curve was 0.67 (95% CI 0.62-0.71). Deeks' funnel plot did not show obvious publication bias among included studies in this meta-analysis. CONCLUSION The specificity of large aneurysm for predicting rebleeding after SAH is relatively high. However, its overall accuracy for predicting aneurysm rebleeding is not very satisfying. A comprehensive model should be developed to improve the accuracy of rebleeding prediction after SAH.