A personalized semi-automatic sleep spindle detection (PSASD) framework.

BACKGROUND: Sleep spindles are distinct electroencephalogram (EEG) patterns of brain activity that have been posited to play a critical role in development, learning, and neurological disorders. Manual scoring for sleep spindles is labor-intensive and tedious but could supplement automated algorithms to resolve challenges posed with either approaches alone. NEW METHODS: A Personalized Semi-Automatic Sleep Spindle Detection (PSASD) framework was developed to combine the strength of automated detection algorithms and visual expertise of human scorers. The underlying model in the PSASD framework assumes a generative model for EEG sleep spindles as oscillatory components, optimized to EEG amplitude, with remaining signals distributed into transient and low-frequency components. RESULTS: A single graphical user interface (GUI) allows both manual scoring of sleep spindles (model training data) and verification of automatically detected spindles. A grid search approach allows optimization of parameters to balance tradeoffs between precision and recall measures. COMPARISON WITH EXISTING METHODS: PSASD outperformed DETOKS in F1-score by 19% and 4% on the DREAMS and P-DROWS-E datasets, respectively. It also outperformed YASA in F1-score by 25% in the P-DROWS-E dataset. Further benchmarking analysis showed that PSASD outperformed four additional widely used sleep spindle detectors in F1-score in the P-DROWS-E dataset. Titration analysis revealed that four 30-second epochs are sufficient to fine-tune the model parameters of PSASD. Associations of frequency, duration, and amplitude of detected sleep spindles matched those previously reported with automated approaches. CONCLUSIONS: Overall, PSASD improves detection of sleep spindles in EEG data acquired from both younger healthy and older adult patient populations.

Income Disparities in Outcomes of Horizontal Strabismus Surgery in a Pediatric Population.

PURPOSE: To examine postoperative outcomes in pediatric patients undergoing strabismus surgery to determine the potential impact of socioeconomic disparities on ophthalmic outcomes. METHODS: This study included 284 children undergoing strabismus surgery at a tertiary institution with at least 11 months of follow-up and no prior strabismus surgery or other neurologic or ophthalmologic conditions. Demographics, insurance, operative parameters, and appointments scheduled/attended were collected via chart review. Ocular alignment was recorded preoperatively and postoperatively at 3, 12, and 24 months. Two-sided t tests and chi-squared analyses were used to compare demographic and operative parameters. Logistic regression was employed to determine predictive factors for ophthalmic outcomes. RESULTS: There was no difference in failure rates between patients with Medicaid and patients with private insurance 24 months postoperatively (45.9% vs 50.5%, respectively, P = .46). Patients with Medicaid were more likely to not follow up postoperatively (28.2% vs 9.6%, respectively, P < .01), whereas patients with private insurance were more likely to complete more than three follow-up appointments in 24 months (21.5% vs 39.0%, respectively, P < .01). Postoperative attendance was linked to Medicaid status (P < .01) but not travel time, neighborhood income levels, or social deprivation index factors. CONCLUSIONS: There was no difference in failure rates between patients with Medicaid and patients with private insurance. Medicaid status was significantly predictive of loss to follow-up. [J Pediatr Ophthalmol Strabismus. 2022;59(3):156-163.].

Reliability of the Neck Disability Index and Japanese Orthopedic Association questionnaires in adult cervical radiculopathy and myelopathy patients when administered by telephone or via online format.

BACKGROUND: The internal validity of long-term studies is significantly affected by the high loss to follow-up in the spine surgery population (>20%). Phone and email-based administration of patient-reported outcomes instruments is a less cumbersome approach for increasing response rates and assessment frequency while potentially decreasing follow-up burden on patients and physicians. PURPOSE: This study sought to validate simultaneous administration of the Neck Disability Index (NDI) and Japanese Orthopedic Association (JOA) questionnaires in patients with either cervical myelopathy and/or radiculopathy. STUDY DESIGN/SETTING: This is a single-center, randomized crossover phone and email validation of legacy outcome measures for cervical myelopathy and/or radiculopathy patients. PATIENT SAMPLE: The study included nonsurgical along with pre- and postsurgical cervical myelopathy and/or radiculopathy patients presenting to a tertiary spine care center. OUTCOME MEASURES: NDI and JOA. METHODS: Two-hundred and six patients (mean age: 58.5 years) were randomized in a 1:4 ratio to either email completion of the NDI and JOA before or after in-office completion, or to phone completion before or after in-office completion. An interval of 1 to 4 weeks was established between the administration of questionnaires. The difference between written in-office and corresponding email and phone versions was assessed with a paired t test. Homogeneity was assessed using intraclass correlation coefficients. Test-retest reliabilities were independently examined for postoperative patients (n=145). Recall bias was assessed in postoperative patients by calculating intraclass correlation coefficients for those with days between assessments lesser than the mean and for those greater. Differences in response rates between phone and email versions were assessed with McNemar's and Cochran-Mantel-Haenszel tests. RESULTS: There was no significant difference between email and in-office versions (n=85) of the NDI (p=.17, Mean Paired Difference=1.34) and JOA (p=.64, Mean Paired Difference=0.11). No significant difference was seen between phone followed by in-office administration (n=32) of the NDI (p=.88, Mean Paired Difference=0.22) and JOA (p=.38, Mean Paired Difference=-0.22), nor between in-office administration followed by phone (n=44) for the NDI (p=.10, Mean Paired Difference=2.79) and JOA (p=.37, Mean Paired Difference=0.27). Intraclass coefficients (ICCs) of the email versions of the NDI and JOA were 0.88 and 0.78, respectively; of the phone-before-office versions of the NDI and JOA were 0.91 and 0.82; of the office-before-phone versions were 0.86 and 0.78. Similarly, strong intraclass correlation coefficients indicative of a lack of recall bias were found for "In-Office" and external assessments completed by postsurgical patients with days between assessments lesser and greater than the mean (ICC range 0.63-0.92). No significant difference was seen in completion rates between email and in-office questionnaire completion (p=.13) and phone-before-office and in-office questionnaire completion (p=.31). However, a significant difference was found in completion rates for phone-after-office questionnaires (p<.001). CONCLUSIONS: Administration of the NDI and JOA over phone and email in patients with cervical myelopathy or radiculopathy is valid with strong test-retest reliability and internal consistency. Phone and email administration of the NDI and JOA can reduce nonresponse rates and decrease the burden of follow-up and data acquisition. Follow-up phone reminders increase the response rate for administration of these patient-reported outcomes (PROs) via email.