Search

Supplemental Digital Content is Available in the Text. Background: Optic disc drusen (ODD) represent an important differential diagnosis of papilledema caused by intracranial hypertension, but their distinction may be difficult in clinical practice. The aim of this study was to train, validate, and test a dedicated deep learning system (DLS) for binary classification of ODD vs papilledema (including various subgroups within each category), on conventional mydriatic digital ocular fundus photographs collected in a large international multiethnic population. Methods: This retrospective study included 4,508 color fundus images in 2,180 patients from 30 neuro-ophthalmology centers (19 countries) participating in the Brain and Optic Nerve Study with Artificial Intelligence (BONSAI) Group. For training and internal validation, we used 857 ODD images and 3,230 papilledema images, in 1,959 patients. External testing was performed on an independent data set (221 patients), including 207 images with ODD (96 visible and 111 buried), provided by 3 centers of the Optic Disc Drusen Studies Consortium, and 214 images of papilledema (92 mild-to-moderate and 122 severe) from a previously validated study. Results: The DLS could accurately distinguish between all ODD and papilledema (all severities included): area under the receiver operating characteristic curve (AUC) 0.97 (95% confidence interval [CI], 0.96–0.98), accuracy 90.5% (95% CI, 88.0%–92.9%), sensitivity 86.0% (95% CI, 82.1%–90.1%), and specificity 94.9% (95% CI, 92.3%–97.6%). The performance of the DLS remained high for discrimination of buried ODD from mild-to-moderate papilledema: AUC 0.93 (95% CI, 0.90–0.96), accuracy 84.2% (95% CI, 80.2%–88.6%), sensitivity 78.4% (95% CI, 72.2%–84.7%), and specificity 91.3% (95% CI, 87.0%–96.4%). Conclusions: A dedicated DLS can accurately distinguish between ODD and papilledema caused by intracranial hypertension, even when considering buried ODD vs mild-to-moderate papilledema. [ABSTRACT FROM AUTHOR]

Neuroanatomical localization (NL) is a key skill in neurology, but learners often have difficulty with it. This study aims to evaluate a concise NL tool (NLT) developed to help teach and learn NL. To evaluate the NLT, an extended-matching questions (EMQ) test to assess NL was designed and validated. The EMQ was validated with fourth-year medical students and internal medicine and neurology residents. The NLT's usability was evaluated with third- and fourth-year students, and the effectiveness was evaluated with an experimental study of second-year students, using the EMQ as the outcome measure. Students were taught how to use both the NLT and textbook algorithms (control) to perform NL, then randomized into either group, and only allowed to use their assigned tool to complete the EMQ. Primary outcome was the difference in mean EMQ scores expressed as a percentage of total score. For EMQ validation, students (n = 56) scored lower than residents (n = 50) (76.7% ± 1.7 vs. 83.0% ± 1.6; mean ± standard error of mean, P < 0.009). The EMQ demonstrated good reliability (Cronbach's a 0.85) and generalizability (G-coefficient 0.85). Third- (n = 77) and fourth-year (n = 42) students found the NLT user-friendly and helpful in their learning of NL. In the experimental study, scores were significantly higher for NLT group (n = 94) than for controls (n = 101) (42.5 vs. 37.0%, P = 0.014); the effect size (Cohen's d) was 0.36. The EMQ is validated to reliably assess NL and is generalizable, feasible, practical, and of low cost. The concise and user-friendly NLT for NL was effective in aiding medical student performance of NL.