Machine learning classification of multiple sclerosis in children using optical coherence tomography.

BACKGROUND: In children, multiple sclerosis (MS) is the ultimate diagnosis in only 1/5 to 1/3 of cases after a first episode of central nervous system (CNS) demyelination. As the visual pathway is frequently affected in MS and other CNS demyelinating disorders (DDs), structural retinal imaging such as optical coherence tomography (OCT) can be used to differentiate MS. OBJECTIVE: This study aimed to investigate the utility of machine learning (ML) based on OCT features to identify distinct structural retinal features in children with DDs. METHODS: This study included 512 eyes from 187 (neyes = 374) children with demyelinating diseases and 69 (neyes = 138) controls. Input features of the analysis comprised of 24 auto-segmented OCT features. RESULTS: Random Forest classifier with recursive feature elimination yielded the highest predictive values and identified DDs with 75% and MS with 80% accuracy, while multiclass distinction between MS and monophasic DD was performed with 64% accuracy. A set of eight retinal features were identified as the most important features in this classification. CONCLUSION: This study demonstrates that ML based on OCT features can be used to support a diagnosis of MS in children.

Teaching an experiential mind-body method to medical students to increase interpersonal skills: a pilot study.

OBJECTIVE: The authors investigate whether inner relationship focusing increases self-awareness in medical students and, in the process, to give them experience with empathic listening. METHODS: Thirteen second-year medical students were randomized into experimental and control groups and surveyed pre-course and post-course about their self-awareness and perceived comfort with clinical interpersonal skills. Subjects attended a 20-h course on inner relationship focusing, followed by 5 months of weekly sessions. Pre-course and post-course survey scores were averaged by group, and mean differences were calculated and compared using the two-sample t test. RESULTS: The experimental group showed improvement in all areas compared to the control group. Improvement in one area (comfort talking to patients about how recurring symptoms might relate to issues in their lives) reached statistical significance (P = 0.05). CONCLUSIONS: Inner relationship focusing is a potential tool to increase self-awareness and empathic listening in medical students.

Shorter psychiatry clerkship length is associated with lower NBME psychiatry shelf exam performance.

OBJECTIVE: The goal of this study was to evaluate a recent medical school curriculum change at our institution 3 years ago; specifically: shortening the Psychiatry core clerkship from 4 to 3 weeks and adding an optional 6-week core/elective combination rotation in lieu of the 3-week core. The authors aimed to determine whether clerkship length was associated with lower scores on the NBME Psychiatry shelf exam (PSE), a requirement of all third-year medical students at our institution. METHOD: Authors collected a convenience sample of 12 years of shelf-exam scores from all Mayo medical students and determined the length of each student's clerkship. Creating three groups (6-week core/elective [N=14], 4-week clerkship [N=478], and 3-week clerkship [N=24]), they determined the mean exam score for each group and compared the means across the three groups by use of one-way ANOVA. RESULTS: The mean shelf-exam score for each group was: 81.5 (6 weeks), 75.3 (4 weeks), and 75.5 (3 weeks), a significant difference between 6 versus 4 weeks and 6 versus 3 weeks, but not 4 versus 3 weeks. CONCLUSIONS: Students completing 3 weeks in psychiatry had an average shelf score almost identical to those completing 4 weeks, but both 3- and 4-week clerkship groups had exam scores significantly lower than students who completed 6 weeks of psychiatry clerkship. The authors propose that differences between groups are the result of shorter clerkship lengths, but other factors, including differences between student cohorts or differences between individual students, may also be involved.

Assessing the integration of audience response system technology in teaching of anatomical sciences.

The goals of our study were to determine the predictive value and usability of an audience response system (ARS) as a knowledge assessment tool in an undergraduate medical curriculum. Over a three year period (2006-2008), data were collected from first year didactic blocks in Genetics/Histology and Anatomy/Radiology (n = 42-50 per class). During each block, students answered clinically oriented multiple choice questions using the ARS. Students' performances were recorded and cumulative ARS scores were compared with final examination performances. Correlation coefficients between these variables were calculated to assess the existence and direction of an association between ARS and final examination score. If associations existed, univariate models were then constructed using ARS as a predictor of final examination score. Student and faculty perception of ARS difficulty, usefulness, effect on performance, and preferred use were evaluated using a questionnaire. There was a statistically significant positive correlation between ARS and final examination scores in all didactic blocks and predictive univariate models were constructed for each relationship (all P < 0.0001). Students and faculty agreed that ARS was easy to use and a reliable tool for providing real-time feedback that improved their performance and participation. In conclusion, we found ARS to be an effective assessment tool benefiting the faculty and the students in a curriculum focused on interaction and self-directed learning.