Smartphone-based interventions for the diagnosis of epileptic seizures: A systematic review and meta-analysis

The main objective of this study was to assess the utility of smartphone-based interventions for epilepsy diagnosis.

A systematic review was performed to evaluate the use of smartphone devices to diagnose epileptic seizures compared with encephalogram (EEG), using the MEDLINE, Scopus, Web of Science, and Embase databases. We plotted pooled sensitivity and specificity estimates on forest plots and on receiver operating characteristics curves. We evaluated evidence certainty using the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) methodology. Also, we constructed a Fagan nomogram to guide clinical decision-making.

We identified 10 studies that evaluated different smartphone-based interventions to diagnose epileptic seizures, including a mobile app of a clinical survey, smartphone-based studies assessing EEG recording, heart rate variability recording and classifier, and video of the epileptic seizure using a smartphone. However, we only performed a quantitative analysis for the smartphone videos and Smartphone Brain Scanner-2. We found that smartphone videos had sensitivity of 77% (95% confidence interval [CI] = 60%-88%) and specificity of 91% (95% CI = 88%-93%) to diagnostic epileptic seizure, with an area under the curve of .91. On the other hand, Smartphone Brain Scanner-2 had sensitivity of 44% (95% CI = 34%-55%) and specificity of 94% (95% CI = 89%-96%). The Epilepsy Diagnosis Aid app had good sensitivity of 76% (95% CI = 66%-84%) and specificity 100% (95% CI = .83%-1.00%); in addition, the RRBLE6:9123 device had sensitivity of 86%.

Our systematic review and meta-analysis demonstrate high specificity with the use of smartphone videos compared with studies assessing EEG. Smartphone-based interventions show promise for diagnosing epilepsy; however, further research is needed to assess the utility of other tools like symptom survey apps and heart rate variability recorders.