Citations

Abstract Introduction The SleepImage is an HSAT device (sHSAT) that uses photoplethysmography and oxygen saturation data to generate an apnea hypopnea index (AHI). This sHSAT has been FDA cleared as a software-as-a-medical-device for use in children 2 years and older for diagnosing obstructive sleep apnea (OSA). To our knowledge this sHSAT has not been tested on children concurrent with the gold standard polysomnography (PSG). Methods This study evaluated 50 children 2-6 years of age with tonsillar hypertrophy and OSA symptoms with sHSAT during simultaneous in-lab PSG. Aims were to determine the technical feasibility of using sHSAT in this population and to compare sHSAT results to PSG. The primary endpoint was AHI by sHSAT compared to AHI by PSG and presence of OSA. Sensitivity analysis was performed for presence or absence of OSA diagnosis based on sHSAT versus PSG. Bland-Altman and concordance correlation calculations were used for comparisons. Results 50 patients, median age 4 (IQR 3-5), met criteria; 39 completed PSG and HSAT. The sensitivity with sHSAT for OSA prevalence based on proprietary categorization was 94.1%, specificity 22.7%, LR+ 1.22, and LR- 0.26, with PSG OSA prevalence of 43.6%. There was a 9% sHSAT total failure rate, and the device frequently fell off, resulting in TST difference of -226.9 ±182.8 min between sHSAT and PSG. PSG was trimmed to include only segments where HSAT was confirmed to be on before comparisons. Concordance correlation coefficient between sAHI and AHI-PSG was 0.52. The Bland-Altman analysis comparing sAHI and PSG AHI revealed a mean difference of 0.73 (95% confidence limits: -3.47 to 4.93). Considering only obstructive events, the Bland-Altman the sAHI vs the AHI-PSG shows a mean difference of 1.52 (95%CI -5.40 to 8.44). Conclusion This pilot study demonstrated challenges with feasibility in using the sHSAT in this age group given likelihood of technical difficulties leading to absent or truncated data. When compared with PSG, sHSAT demonstrated a high false-positive rate for diagnosis of OSA, particularly concerning if guiding surgical management. This topic is of considerable importance to the practice of pediatric sleep medicine; further investigation of technological advances in this area is imperative. Support (if any)

Abstract Objective To develop machine learning models using patient and migraine features that can predict treatment responses to commonly used migraine preventive medications. Background Currently, there is no accurate way to predict response to migraine preventive medications, and the standard trial‐and‐error approach is inefficient. Methods In this cohort study, we analyzed data from the Mayo Clinic Headache database prospectively collected from 2001 to December 2023. Adult patients with migraine completed questionnaires during their initial headache consultation to record detailed clinical features and then at each follow‐up to track preventive medication changes and monthly headache days. We included patients treated with at least one of the following migraine preventive medications: topiramate, beta‐blockers (propranolol, metoprolol, atenolol, nadolol, timolol), tricyclic antidepressants (amitriptyline, nortriptyline), verapamil, gabapentin, onabotulinumtoxinA, and calcitonin gene‐related peptide (CGRP) monoclonal antibodies (mAbs) (erenumab, fremanezumab, galcanezumab, eptinezumab). We pre‐trained a deep neural network, “TabNet,” using 145 variables, then employed TabNet‐embedded data to construct prediction models for each medication to predict binary outcomes (responder vs. non‐responder). A treatment responder was defined as having at least a 30% reduction in monthly headache days from baseline. All model performances were evaluated, and metrics were reported in the held‐out test set (train 85%, test 15%). SHapley Additive exPlanations (SHAP) were conducted to determine variable importance. Results Our final analysis included 4260 patients. The responder rate for each medication ranged from 28.7% to 34.9%, and the mean time to treatment outcome for each medication ranged from 151.3 to 209.5 days. The CGRP mAb prediction model achieved a high area under the receiver operating characteristics curve (AUC) of 0.825 (95% confidence interval [CI] 0.726, 0.920) and an accuracy of 0.80 (95% CI 0.70, 0.88). The AUCs of prediction models for beta‐blockers, tricyclic antidepressants, topiramate, verapamil, gabapentin, and onabotulinumtoxinA were: 0.664 (95% CI 0.579, 0.745), 0.611 (95% CI 0.562, 0.682), 0.605 (95% CI 0.520, 0.688), 0.673 (95% CI 0.569, 0.724), 0.628 (0.533, 0.661), and 0.581 (95% CI 0.550, 0.632), respectively. Baseline monthly headache days, age, body mass index (BMI), duration of migraine attacks, responses to previous medication trials, cranial autonomic symptoms, family history of headache, and migraine attack triggers were among the most important variables across all models. A variable could have different contributions; for example, lower BMI predicts responsiveness to CGRP mAbs and beta‐blockers, while higher BMI predicts responsiveness to onabotulinumtoxinA, topiramate, and gabapentin. Conclusion We developed an accurate prediction model for CGRP mAbs treatment response, leveraging detailed migraine features gathered from a headache questionnaire before starting treatment. Employing the same methods, the model performances for other medications were less impressive, though similar to the machine learning models reported in the literature for other diseases. This may be due to CGRP mAbs being migraine‐specific. Incorporating medical comorbidities, genomic, and imaging factors might enhance the model performance. We demonstrated that migraine characteristics are important in predicting treatment responses and identified the most crucial predictors for each of the seven types of preventive medications. Our results suggest that precision migraine treatment is feasible.