Differences in Facial Expression Recognition Between Unipolar and Bipolar Depression

Abstract

PurposeTo explore the differences in facial emotion recognition among patients with unipolar depression (UD), bipolar depression (BD), and normal controls.MethodsThirty patients with UD and 30 patients with BD, respectively, were recruited in Zhumadian Second People’s Hospital from July 2018 to August 2019. Fifteen groups of facial expressions including happiness, sadness, anger, surprise, fear, and disgust were identified.ResultsA single-factor ANOVA was used to analyze the facial expression recognition results of the three groups, and the differences were found in the happy-sad (P = 0.009), happy-angry (P = 0.001), happy-surprised (P = 0.034), and disgust-surprised (P = 0.038) facial expression groups. The independent sample T-test analysis showed that compared with the normal control group, there were differences in the happy-sad (P = 0.009) and happy-angry (P = 0.009) groups in patients with BD, and the accuracy of facial expression recognition was lower than the normal control group. Compared with patients with UD, there were differences between the happy-sad (P = 0.005) and happy-angry (P = 0.002) groups, and the identification accuracy of patients with UD was higher than that of patients with BD. The time of facial expression recognition in the normal control group was shorter than that in the patient group. Using happiness-sadness to distinguish unipolar and BDs, the area under the ROC curve (AUC) is 0.933, the specificity is 0.889, and the sensitivity is 0.667. Using happiness-anger to distinguish unipolar and BD, the AUC was 0.733, the specificity was 0.778, and the sensitivity was 0.600.ConclusionPatients with UD had lower performance in recognizing negative expressions and had longer recognition times. Those with BD had lower accuracy in recognizing positive expressions and longer recognition times. Rapid facial expression recognition performance may be as a potential endophenotype for early identification of unipolar and BD.

Working memory updating in individuals with bipolar and unipolar depression: fMRI study

“…The distractors were introduced to examine the participants’ ability to inhibit irrelevant emotionally salient (happy and fearful faces) and emotionally ambiguous (neutral faces) information. Previous research suggested that individuals with BD and UD differ in their response to emotional faces [ 39 , 40 ] with the former showing reduced performance when processing positive emotional stimuli and the latter showing reduced performance when processing negative emotional stimuli [ 40 ]. The differences in emotion processing could differentially affect the ability of individuals with BD and UD to resolve interference from emotional stimuli especially during a difficult working memory task.…”

Deep Learning on Facial Expression Detection : Artificial Neural Network Model Implementation

“…Effective communication is essential to human-machine interaction, which is why the categorization of feelings is so vitally crucial. Classifying expressions can be helpful for enhancing fields as different as the diagnosis of autism [13]- [15] and depression [16]- [18] , consumer and retail satisfaction surveys [19], [20] , and classroom education [21]- [23] . [13]- [15] and [16]- [18] are only couple examples.…”

“…Classifying expressions can be helpful for enhancing fields as different as the diagnosis of autism [13]- [15] and depression [16]- [18] , consumer and retail satisfaction surveys [19], [20] , and classroom education [21]- [23] . [13]- [15] and [16]- [18] are only couple examples. The field of emotion recognition is gaining popularity among academics and is seeing an uptick in the amount of studies that are being conducted and published on the subject.…”