Sensitivity and specificity of a facial emotion recognition test in classifying patients with schizophrenia

Abnormal ambiguous facial expression recognition in Chinese patients with schizophrenia

BACKGROUND

Patients with schizophrenia (SCZ) exhibit difficulties deficits in recognizing facial expressions with unambiguous valence. However, only a limited number of studies have examined how these patients fare in interpreting facial expressions with ambiguous valence (for example, surprise). Thus, we aimed to explore the influence of emotional background information on the recognition of ambiguous facial expressions in SCZ.

METHODS

A 3 (emotion: negative, neutral, and positive) × 2 (group: healthy controls and SCZ) experimental design was adopted in the present study. The experimental materials consisted of 36 images of negative emotions, 36 images of neutral emotions, 36 images of positive emotions, and 36 images of surprised facial expressions. In each trial, a briefly presented surprised face was preceded by an affective image. Participants (36 SCZ and 36 healthy controls (HC)) were required to rate their emotional experience induced by the surprised facial expressions. Participants' emotional experience was measured using the 9-point rating scale. The experimental data have been analyzed by conducting analyses of variances (ANOVAs) and correlation analysis.

RESULTS

First, the SCZ group reported a more positive emotional experience under the positive cued condition compared to the negative cued condition. Meanwhile, the HC group reported the strongest positive emotional experience in the positive cued condition, a moderate experience in the neutral cued condition, and the weakest in the negative cue condition. Second, the SCZ (vs. HC) group showed longer reaction times (RTs) for recognizing surprised facial expressions. The severity of schizophrenia symptoms in the SCZ group was negatively correlated with their rating scores for emotional experience under neutral and positive cued condition.

CONCLUSIONS

Recognition of surprised facial expressions was influenced by background information in both SCZ and HC, and the negative symptoms in SCZ. The present study indicates that the role of background information should be fully considered when examining the ability of SCZ to recognize ambiguous facial expressions.

A method for diagnosing depression: Facial expression mimicry is evaluated by facial expression recognition

BACKGROUND

Considerable evidence has shown that facial expression mimicry is impaired in patients with depression. We aimed to evaluate voluntary expression mimicry by facial expression recognition for diagnosing depression.

METHODS

A total of 168 participants performed voluntary expression mimicry task, posing anger, disgust, fear, happiness, neutrality, sadness, and surprise. 9 healthy raters performed facial expression recognition task through the observer scoring method, and evaluated seven expressions imitated by participants. Emotional scores were calculated to measure any differences between two groups of participants and provided a basis for clinical diagnosis of depression.

RESULTS

Compared with the control group, the depression group had lower accuracy in imitating happiness. Compared with the control group, the depression group imitated a higher neutrality bias for sadness, surprise, happiness and disgust, while sadness and surprise had a lower happiness bias; for imitating happiness, the depression group showed higher anger, disgust, fear, neutrality, and surprise bias; for imitating neutrality, the depression group showed higher sadness bias, and lower happiness bias. Compared with the control group, the raters had a higher reaction time to recognize the happiness imitated by depression group, and it was positively correlated with severity of depression. The severity of depression was also negatively correlated with accuracy in imitating happiness, and positively correlated with neutrality bias of imitating surprise.

LIMITATIONS

The ecological effectiveness of static stimulus materials is lower than that of dynamic stimuli. Without synchronized functional imaging, there is no way to link brain activation patterns.

CONCLUSION

The ability of patients with depression to voluntarily imitate facial expressions declines, which is mainly reflected in accuracy, bias and recognizability. Our experiment has discovered deficits in these aspects of patients with depression, which will be used as a method for diagnosising depression.

Specific facial emotion recognition deficits across the course of psychosis: A comparison of individuals with low-risk, high-risk, first-episode psychosis and multi-episode schizophrenia-spectrum disorders

Our study aimed to explore the recognition of specific emotions across the course of psychosis. A visual task representing the six basic emotions was used to assess facial emotion recognition (FER) in 204 healthy controls classified into 152 low-risk (LR) and 52 high-risk for psychosis (HR), following a psychometric risk approach; and 100 patients: 44 with first-episode psychosis (FEP) and 56 with multi-episode schizophrenia-spectrum disorders (MES). First, we performed a MANCOVA to compare the four conditions. Next, we conducted a logistic regression to explore whether specific FER deficits predicted the presence of psychosis. Finally, we investigated the relationships of FER with psychosis-like experiences (PLEs) and psychotic symptoms. Global FER, anger and fear recognition were impaired in HR, FEP and MES. No differences between HR and FEP appeared. Moreover, fear and anger correctly classified 83% of individuals into LR or psychosis. FER was associated with PLEs and psychotic symptoms. Concluding, FER is early impaired in HR individuals and increases along the psychosis continuum. However, fear recognition is similarly impaired throughout the illness, suggesting a possible vulnerability marker. Furthermore, deficits in anger and fear recognition predicted the presence of psychosis. Therefore, we suggest that FER may be essential in detecting psychosis risk.

Facial expression-based indicators of schizophrenia: Evidence from recent research

Impaired ability to recognize emotion in other's face (decoding) or to express emotion through the face (encoding) are considered critical in schizophrenia. The topic of research draws considerable attention since clinicians rely heavily on the patient's facial expressions for diagnosis and on the patient's ability to understand the clinician's communicative intent. While most researchers argue in favor of a generalized emotion deficit, others indicate an emotion-specific deficit in schizophrenia. An early review (Mandal et al., 1998) indicated a possible breakdown in perception-expression-experience link of emotion; later reviews (Kohler et al., 2010; Chan et al., 2010) pointed to a generalized emotion processing deficit due to perceptual deficits in schizophrenia. The present review (2010-2022) revisits this controversy with 47 published studies (37 decoding, 10 encoding) conducted on 2364 patients in 20 countries. Schizophrenia is characterized by reduced emotion processing ability, especially with negative symptoms and at an acute state of illness. It is however still unclear whether this dysfunction is independent of a generalized face perception deficit or of subjective experience of emotion in schizophrenia.

Theory of Mind and social functioning among neuropsychiatric disorders: A transdiagnostic study

Social dysfunction is commonly present in neuropsychiatric disorders of schizophrenia (SZ) and Alzheimer's disease (AD). Theory of Mind (ToM) deficits have been linked to social dysfunction in disease-specific studies. Nevertheless, it remains unclear how ToM is related to social functioning across these disorders, and which factors contribute to this relationship. We investigated transdiagnostic associations between ToM and social functioning among SZ/AD patients and healthy controls, and explored to what extent these associations relate to information processing speed or facial emotion recognition capacity. A total of 163 participants were included (SZ: n=56, AD: n=50 and age-matched controls: n=57). Social functioning was assessed with the Social Functioning Scale (SFS) and the De Jong-Gierveld Loneliness Scale (LON). ToM was measured with the Hinting Task. Information processing speed was measured by the Digit Symbol Substitution Test (DSST) and facial emotion recognition capacity by the facial emotion recognition task (FERT). Case-control deficits in Hinting Task performance were larger in AD (rrb = -0.57) compared to SZ (rrb = -0.35). Poorer Hinting Task performance was transdiagnostically associated with the SFS (βHinting-Task = 1.20, p<0.01) and LON (βHinting-Task = -0.27, p<0.05). DSST, but not FERT, reduced the association between the SFS and Hinting Task performance, however the association remained significant (βHinting-Task = 0.95, p<0.05). DSST and FERT performances did not change the association between LON and Hinting Task performance. Taken together, ToM deficits are transdiagnostically associated with social dysfunction and this is partly related to reduced information processing speed.

Improving Facial Emotion Recognition Using Residual Autoencoder Coupled Affinity Based Overlapping Reduction

Emotion recognition using facial images has been a challenging task in computer vision. Recent advancements in deep learning has helped in achieving better results. Studies have pointed out that multiple facial expressions may present in facial images of a particular type of emotion. Thus, facial images of a category of emotion may have similarity to other categories of facial images, leading towards overlapping of classes in feature space. The problem of class overlapping has been studied primarily in the context of imbalanced classes. Few studies have considered imbalanced facial emotion recognition. However, to the authors’ best knowledge, no study has been found on the effects of overlapped classes on emotion recognition. Motivated by this, in the current study, an affinity-based overlap reduction technique (AFORET) has been proposed to deal with the overlapped class problem in facial emotion recognition. Firstly, a residual variational autoencoder (RVA) model has been used to transform the facial images to a latent vector form. Next, the proposed AFORET method has been applied on these overlapped latent vectors to reduce the overlapping between classes. The proposed method has been validated by training and testing various well known classifiers and comparing their performance in terms of a well known set of performance indicators. In addition, the proposed AFORET method is compared with already existing overlap reduction techniques, such as the OSM, ν-SVM, and NBU methods. Experimental results have shown that the proposed AFORET algorithm, when used with the RVA model, boosts classifier performance to a greater extent in predicting human emotion using facial images.

Using machine learning to improve the discriminative power of the FERD screener in classifying patients with schizophrenia and healthy adults

Background Facial emotion recognition deficit (FERD) seems to be an obvious feature of patients with schizophrenia and has great potential for classifying patients and non-patients. The FERD screener was previously developed to classify patients from healthy adults. However, an obvious drawback of this screener is that the recommended cut-off scores could enhance either sensitivity or specificity (about 0.92) only, while the other one is at an only acceptable level (about 0.66). Machine learning (ML) algorithms are famous for their feature extraction and data classification abilities, which are promising for improving the discriminative power of screeners. This study aimed to improve the discriminative power of the FERD screener using an ML algorithm. Methods The data were extracted from a previous study. Artificial neural networks were generated to estimate the probability of being a patient with schizophrenia or a healthy adult based on the examinee's responses on the FERD screener (168 items). The performances of the ML-FERD screener were examined using a stratified five-fold cross-validation method. Results Across the five subsets of data, the ML-FERD screener showed extremely high areas under the receiver operating characteristic curve of 0.97-0.99. With the optimized cut-off scores, the average sensitivity and specificity of the ML-FERD screener were 0.90 (0.85-0.93) and 0.93 (0.86-1.00), respectively. Limitations The characteristics of patients were not representative, and the age was mismatched to control group. Conclusion The ML-FERD screener appears to have a better discriminative power to classify patients with schizophrenia and healthy adults than does the FERD screener.

The Relationship Between Facial Expression and Cognitive Function in Patients With Depression

Objective: Considerable evidence has shown that facial expression recognition ability and cognitive function are impaired in patients with depression. We aimed to investigate the relationship between facial expression recognition and cognitive function in patients with depression.

Methods: A total of 51 participants (i.e., 31 patients with depression and 20 healthy control subjects) underwent facial expression recognition tests, measuring anger, fear, disgust, sadness, happiness, and surprise. The Chinese version of the MATRICS Consensus Cognitive Battery (MCCB), which assesses seven cognitive domains, was used.

Results: When compared with a control group, there were differences in the recognition of the expressions of sadness (p = 0.036), happiness (p = 0.041), and disgust (p = 0.030) in a depression group. In terms of cognitive function, the scores of patients with depression in the Trail Making Test (TMT; p < 0.001), symbol coding (p < 0.001), spatial span (p < 0.001), mazes (p = 0.007), the Brief Visuospatial Memory Test (BVMT; p = 0.001), category fluency (p = 0.029), and continuous performance test (p = 0.001) were lower than those of the control group, and the difference was statistically significant. The accuracy of sadness and disgust expression recognition in patients with depression was significantly positively correlated with cognitive function scores. The deficits in sadness expression recognition were significantly correlated with the TMT (p = 0.001, r = 0.561), symbol coding (p = 0.001, r = 0.596), maze (p = 0.015, r = 0.439), and the BVMT (p = 0.044, r = 0.370). The deficits in disgust expression recognition were significantly correlated with impairments in the TMT (p = 0.005, r = 0.501) and symbol coding (p = 0.001, r = 0.560).

Conclusion: Since cognitive function is impaired in patients with depression, the ability to recognize negative facial expressions declines, which is mainly reflected in processing speed, reasoning, problem-solving, and memory.