AUTHOR=Navarro Jorge , Fernández Rosell Mercedes , Castellanos Angel , del Moral Raquel , Lahoz-Beltra Rafael , Marijuán Pedro C. 

TITLE=Plausibility of a Neural Network Classifier-Based Neuroprosthesis for Depression Detection via Laughter Records

JOURNAL=Frontiers in Neuroscience

VOLUME=Volume 13 - 2019

YEAR=2019

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2019.00267

DOI=10.3389/fnins.2019.00267

ISSN=1662-453X

ABSTRACT=The application of neuroprostheses in neuropsychiatry has to date been limited to the treatment of certain diseases related to nervous system malfunction, such as epilepsy, depression, and sensorimotor disorders. At present, the use of this kind of prosthesis has not been explored regarding the diagnosis and onset of mental disorders, for instance depression, Parkinson, or bipolar disorder. The present work explores the diagnostic performance for depression of different neural network classifiers analyzing directly the sound structures of laughter as registered from clinical patients of depression as well as from healthy controls. The main methodological novelty of this work is that simple sound variables are used as input signals, instead of electrophysiological signals or local field potentials (LFPs) which are the usual protocol to date. In the present study, involving 934 laughs from 30 patients and 20 controls, four different neural networks models were tested for sensitivity analysis and were additionally trained for depression detection. Several sound fundamental variables were extracted from the records: time, fundamental frequency mean, first three formants, average power, and the Shannon-Wiener entropy. In the results obtained, two of the neural networks show a diagnostic discrimination capability of 93.02% and 91.15% respectively, while the third and fourth ones have an 87.96 and 82.40% percentage of success. Remarkably, entropy turns out to be a fundamental variable to distinguish between patients and controls. In more general terms, this experimental survey of laughter to be used as an auxiliary diagnostic tool in depression shows the possibility of bona fide inferring qualities of mental states out from the sound structures of emotional utterances, as seems to be true in the case of laughter. In biomedical terms, our neural network classifier-based neuroprosthesis opens up the possibility of applying the methodology we have developed to other mental health and neuropsychiatric pathologies. Exploring the application of laughter in the early detection and prognosis of Alzheimer and Parkinson represents an enticing possibility. It may look paradoxical, but regarding the man-machine interface, laughter represents a most robust bridge in between the consistency of intelligence processes and the ad hoc onset of emotional reactions.