Currently submitted to: JMIR Mental Health
Date Submitted: Sep 23, 2017
Open Peer Review Period: Sep 24, 2017 - Nov 19, 2017 (Under Review)
Lamotrigine therapy for bipolar depression: Analysis of self-reported patient data
Lamotrigine is an anticonvulsant used for the treatment of bipolar disorder, especially for depressive episodes. A randomized placebo controlled trial was conducted and self-reported mood scores were collected over a period of 52 weeks. Different features were computed from the weekly time series produced by each subject. The coefficient of variation (σ/μ) and detrended fluctuation analysis scaling exponent (α), a measure of scaling and long-range persistence (long memory) in time series, were selected as having the most explanatory power. Data from patients taking lamotrigine show a general decrease in depression score in comparison to the patients taking placebo. The time series of lamotrigine patients tend to be rougher than the placebo group. A classifier was built and, based on the two chosen metrics, we are able to achieve a classification accuracy of more than 60% in predicting the treatment mode.