Constructive Fuzzy Cognitive Map for Depression Severity Estimation

Abstract

Depression is a common and serious medical disorder that negatively affects the mood and the emotions of people, especially adolescents. In this paper, a novel framework for automatically creating Fuzzy Cognitive Maps (FCMs) is proposed. It is applied for the estimation of the severity of depression among adolescents, based on their electroencephalogram (EEG). The introduced Constructive FCM (CFCM) utilizes features extracted by a Constructive Fuzzy Representation Model (CFRM), which conduces to detect in a more intuitive way the cause-and-effect relationships between the brain activity and depression. CFCM contributes to limiting the participation of experts, and the manual interventions in the traditional construction of FCMs, it provides an embedded mechanism for dimensionality reduction, and it constitutes an inherently interpretable approach to decision making, while being uncertainty-aware and simple to implement. The results of the experiments, using a recent publicly available dataset, demonstrate the effectiveness of the proposed framework and highlight its advantages.

Interpretable EEG-Based Emotion Recognition Using Fuzzy Cognitive Maps

“…For example, as can be observed from Figure 1, HAHV includes positive emotions, e.g., happiness and amusement. In addition, to examine the emotional state of the brain, a segmentation of the EEG electrode positions into brain regions was performed, based on [4], and adapted to the dataset used.…”

“…In our experiments, ߣ corresponds to positive and negative emotion. The fuzzy set construction is performed, aiming to characterize linguistically the calculated brain electrical activity, based on [4]. Specifically, a clustering algorithm is applied to group ‫ܨ‬ ǡఒ ௪ into a set of ‫ܯ‬ clusters with ‫ܯ‬ ൏ ܸ ఒ , where ܸ ఒ corresponds to the total number of volunteers participating in the emotion recognition problem.…”

“…The proposed FCM model iteratively calculates its states until convergence, based on ‫ܣ‬ ௧ାଵ ൌ ݄൫‫ܣ‬ ௧ σ ‫ݓ‬ ‫ܣ‬ ௧ ୀଵǡஷ ൯ , where ‫ݐ‬ ൌ ͳ ǥ Ǥ ǡ ܶ is the iteration numberǡ ‫ݓ‬ is the weight matrix of ‫ܥ‬ to ‫ܥ‬ , and ݄ is a transfer function. The values of the initial state vector ‫ܣ‬ ൌ ሺ‫ܣ‬ ଵ ǡ ǥ ǡ ‫ܣ‬ ሻ are estimated based on [4].…”

“…During the last decades, the use of fuzzy network structures has shown great potential in various fields, including medicine [3]. Specifically, FCMs are graph-based methods, which have received considerable attention from researchers, due to their simplicity, effectiveness, and high ability to deal with uncertainties [4]. However, a limitation of FCMs is that there is a need for human participation to determine the structure of the graph.…”

“…However, a limitation of FCMs is that there is a need for human participation to determine the structure of the graph. Recent modified FCMs have been effectively applied to various medical problems, including Constructive Fuzzy Representation Model (CFRM) for heart disease classification [5], and Constructive FCM (CFCM) for depression severity estimation [4]. In this paper, we introduce an FCM model for interpretable emotion recognition, based on EEG signals.…”