Search for a command to run...
Introduction: Midline Shift (MLS) quantifies mass effect and herniation risk in acute neurological emergencies. Manual MLS measurements are time-consuming and variable, delaying critical intervention. This study evaluated the performance of Rapid MLS (iSchemaView, Inc.), automated software that quantifies MLS and visualizes displacement on non-contrast CT scans. Methods: In this retrospective multicenter-study, patients with intracranial pathology causing MLS were identified from NCCT databases. Automated MLS measurements from Rapid MLS were compared to those by three expert neuroradiologists. The primary outcome assessed non-inferiority of Rapid MLS to the average mean absolute error (MAE) of experts across pairwise comparisons (one-sided α=0.025). Secondary analyses included Passing-Bablok regression and a Bland-Altman plot. Results: Of 157 initial cases, 153 (mean age 68.7 ± 16.3 years; 59 M, 89 F, 5 unknown) were analyzed after four exclusions due to non-axial images, excessive streak artifact, mass at septum pellucidum, and hemicraniectomy. The cases covered a range of pathologies and expected MLS values outlined in Table 1 and Figure 1 respectively. The primary endpoint analysis demonstrated the MAE of Rapid (0.8mm; 95% CI: 0.7-1.0 mm) was non-inferior to the average pairwise MAE of experts (0.9mm; 95% CI: 0.8-1.0mm) (p<0.0001). Passing-Bablok regression analysis showed high agreement (intercept=0.29; slope=0.90) with minimal bias (difference of 0.2; p=0.0516)(Figure 2). We note that Rapid MLS progressively underestimates values above 10mm where clinical concern is already high. Further investigation of the bias with a Bland-Altman analysis determined a small mean difference of 0.2mm (95% CI: 0.0-0.4mm) between Rapid MLS and expert measurements. Conclusion: Rapid MLS provides fast, reproducible MLS measurements on NCCT, with accuracy comparable to expert neuroradiologists. This tool may enhance triage and decision-making in stroke, traumatic brain injury, and other emergencies by reducing time to diagnosis and intervention.