A temporal classification method based on behavior time series data in patients with behavioral variant of frontotemporal dementia and apathy - Sorbonne Université Access content directly
Journal Articles Journal of Neuroscience Methods Year : 2022

A temporal classification method based on behavior time series data in patients with behavioral variant of frontotemporal dementia and apathy

Abstract

Background Apathy is a common behavioral syndrome that occurs across neurological and psychiatric disorders. An influential theoretical framework defined apathy as the quantitative reduction of self-generated voluntary and purposeful behaviors. There is evidence in the literature of the multidimensional nature of apathy with cognitive, behavioral, and emotional dimensions. To date, apathy has been assessed using various scales and questionnaires. Alternative objective and ecological measurements of apathy are needed. New method We used the ECOCAPTURE protocol and an ethological approach to investigate behavior in bvFTD patients under ecological conditions (a waiting room) while they freely explored a novel environment. Data were collected by behavioral coding from 7-minute video using an ethogram and transformed into behavior time series data. We present an approach considering behavioral kinetics to assess behavior. We aimed to construct a new behavior analysis method, called ECOCAPTURE kinetics, using temporal classification for behavior time series data analysis. To develop our classifier, we retained a nonelastic Euclidian metric, combined with a convolutional approach. Results We applied the ECOCAPTURE kinetics method to a cohort of 20 bvFTD patients and 18 healthy controls. We showed that bvFTD patients can be classified according to their behavioral kinetics into three groups. Each subgroup was characterized by specific behavior disorders and neuropsychological profile. Comparison with Existing Method(s) The ECOCAPTURE kinetics method is different from those of the classical approach of measuring behavior, producing time budgets, frequency of behavior occurrences, or kinematic diagrams. Conclusions This approach can be extended to any behavioral study encoding time.

Dates and versions

hal-03699572 , version 1 (20-06-2022)

Identifiers

Cite

Caroline Peltier, François-Xavier Lejeune, Lars G.T. Jorgensen, Armelle Rametti-Lacroux, Delphine Tanguy, et al.. A temporal classification method based on behavior time series data in patients with behavioral variant of frontotemporal dementia and apathy. Journal of Neuroscience Methods, 2022, 376, pp.109625. ⟨10.1016/j.jneumeth.2022.109625⟩. ⟨hal-03699572⟩
93 View
0 Download

Altmetric

Share

Gmail Facebook X LinkedIn More