Fig. 2: Testing a rhythmic hypothesis. | Translational Psychiatry

Fig. 2: Testing a rhythmic hypothesis.

From: Testing frameworks for personalizing bipolar disorder

Fig. 2

a Mood scores were transformed into the frequency domain using spectral density estimation. A rhythmic mood course would result in a significant peak in the spectral density. b Frequencies were tested for significance and then the resulting P-values were averaged across patients to evaluate the collective evidence that depressive and manic scores in the weekly (N = 178) and bimonthly (N = 15) datasets oscillate at a particular frequency. Significance is a mean P-value that dips below the dashed lines, which mark 5th quantiles for the mean of N uniform random variables between 0 and 1 with N = 178 for the bimonthly dataset and N = 15 for the weekly dataset

Back to article page