ObjectiveMonitoring biomarkers using machine learning (ML) may determine glioblastoma treatment response. We systematically reviewed quality and performance accuracy of recently published studies.MethodsFollowing Preferred Reporting Items for Systematic Reviews and Meta-Analysis: Diagnostic Test Accuracy, we extracted articles from MEDLINE, EMBASE and Cochrane Register between 09/2018–01/2021. Included study participants were adults with glioblastoma having undergone standard treatment (maximal resection, radiotherapy with concomitant and adjuvant temozolomide), and follow-up imaging to determine treatment response status (specifically, distinguishing progression/recurrence from progression/recurrence mimics, the target condition). Using Quality Assessment of Diagnostic Accuracy Studies Two/Checklist for Artificial Intelligence in Medical Imaging, we assessed bias risk and applicability concerns. We determined test set performance accuracy (sensitivity, specificity, precision, F1-score, balanced accuracy). We used a bivariate random-effect model to determine pooled sensitivity, specificity, area-under the receiver operator characteristic curve (ROC-AUC). Pooled measures of balanced accuracy, positive/negative likelihood ratios (PLR/NLR) and diagnostic odds ratio (DOR) were calculated. PROSPERO registered (CRD42021261965).ResultsEighteen studies were included (1335/384 patients for training/testing respectively). Small patient numbers, high bias risk, applicability concerns (particularly confounding in reference standard and patient selection) and low level of evidence, allow limited conclusions from studies. Ten studies (10/18, 56%) included in meta-analysis gave 0.769 (0.649-0.858) sensitivity [pooled (95% CI)]; 0.648 (0.749-0.532) specificity; 0.706 (0.623-0.779) balanced accuracy; 2.220 (1.560-3.140) PLR; 0.366 (0.213-0.572) NLR; 6.670 (2.800-13.500) DOR; 0.765 ROC-AUC.ConclusionML models using MRI features to distinguish between progression and mimics appear to demonstrate good diagnostic performance. However, study quality and design require improvement.

Background An external ventricular drain (EVD) is typically indicated in the presence of hydrocephalus and increased intracranial pressure (ICP). Procedural challenges have prompted the development of different methods to improve accuracy, safety, and logistics. Objectives EVD placement and complications rates were compared using two surgical techniques; the standard method (using a 14-mm trephine burrhole with the EVD tunnelated through the skin) was compared to a less invasive method (EVD placed through a 2.7-3.3-mm twist drill burrhole and fixed to the bone with a bolt system). Methods Retrospective observational study in a single-centre setting between 2008 and 2018. EVD placement was assessed using the Kakarla scoring system. We registered postoperative complications, surgery duration and number of attempts to place the EVD. Results Two hundred seventy-two patients received an EVD (61 bolt EVDs, 211 standard EVDs) in the study period. Significant differences between the bolt system and the standard method were observed in terms of revision surgeries (8.2% vs. 21.5%, p = 0.020), surgery duration (mean 16.5 vs. 28.8 min, 95% CI 7.64, 16.8, p < 0.001) and number of attempts to successfully place the first EVD (mean 1.72 ± 1.2 vs. 1.32 ± 0.8, p = 0.017). There were no differences in accuracy of placement or complication rates. Conclusions The two methods show similar accuracy and postoperative complication rates. Observed differences in both need for revisions and surgery duration favoured the bolt group. Slightly, more attempts were needed to place the initial EVD in the bolt group, perhaps reflecting lower flexibility for angle correction with a twist drill approach. Keywords Bolt drain. External ventricular drain. Hydrocephalus. Kakarla score. Tunnelated drain. Neurosurgery This article is part of the Topical Collection on Neurosurgical technique evaluation

No abstract

Background: Neuroimaging shows considerable promise in generating sensitive and objective outcome measures for therapeutic trials across a range of neurodegenerative conditions. For volumetric measures the current gold standard is manual delineation, which is unfeasible for samples sizes required for large clinical trials.Methods: Using a cohort of early Huntington’s disease (HD) patients (n = 46) and controls (n = 35), we compared the performance of four automated segmentation tools (FIRST, FreeSurfer, STEPS, MALP-EM) with manual delineation for generating cross-sectional caudate volume, a region known to be vulnerable in HD. We then examined the effect of each of these baseline regions on the ability to detect change over 15 months using the established longitudinal Caudate Boundary Shift Integral (cBSI) method, an automated longitudinal pipeline requiring a baseline caudate region as an input.Results: All tools, except Freesurfer, generated significantly smaller caudate volumes than the manually derived regions. Jaccard indices showed poorer levels of overlap between each automated segmentation and manual delineation in the HD patients compared with controls. Nevertheless, each method was able to demonstrate significant group differences in volume (p &lt; 0.001). STEPS performed best qualitatively as well as quantitively in the baseline analysis. Caudate atrophy measures generated by the cBSI using automated baseline regions were largely consistent with those derived from a manually segmented baseline, with STEPS providing the most robust cBSI values across both control and HD groups.Conclusions: Atrophy measures from the cBSI were relatively robust to differences in baseline segmentation technique, suggesting that fully automated pipelines could be used to generate outcome measures for clinical trials.