The selection of an appropriate virtual reference schema is pivotal in determining the outcomes of Event-Related Potential (ERP) studies, particularly within the widely utilized Talk/Listen ERP paradigm, which is employed to non-invasively explore the corollary discharge (CD) phenomenon in the speech-auditory system. This research centers on examining the effects of prevalent EEG reference schemas—Linked Mastoids (LM), Common Average Reference (CAR), and Reference Electrode Standardization Technique (REST)—through statistical analysis, Statistical Parametric Scalp Mapping (SPSM), and source localization techniques. Our ANOVA findings indicate significant main effects for both the reference and the experimental condition on the amplitude of N1 ERPs. Depending on the reference used, the polarity and amplitude of the N1 ERPs demonstrate systematic variations: LM is associated with pronounced frontocentral activity, whereas both CAR and REST exhibit patterns of frontocentral and occipitotemporal activity. The significance of SPSM results is confined to regions exhibiting prominent N1 activity for each reference schema. Source analysis provides corroborative evidence more aligned with the SPSM results for CAR than for REST or LM, suggesting that results under CAR are more objective and reliable. Therefore, the CAR reference is recommended for future studies involving Talk/Listen ERP paradigms.
Cannabinoids can induce clinically significant psychotic episodes immediately following exposure, that can persist well beyond the duration of intoxication, and that require clinical intervention: cannabinoid induced acute and persistent psychosis (CIAPP). CIAPP may represent a distinct subtype of psychotic disorder that in our current nomenclature is subsumed under schizophrenia. The importance of parsing out different subtypes of chronic recurrent psychoses that are currently lumped under schizophrenia, is to more accurately diagnose and estimate prognosis, tailor disorder-specific treatments, and to minimize iatrogenic harm. Furthermore, characterizing the specific biological correlates of subtypes at the genetic and cellular levels, will be critical in validating the subtypes. However, the sporadic occurrence of CIAPP makes it is harder to study. Every year, there is a predictable spate of hospitalizations for CIAPP that occur at the Central Institute of Psychiatry, Ranchi, India following certain Indian festivals during which there is the ritual use of cannabis. Since the date of these festivals is predetermined, the capture and study of CIAPP cases can occur in a planned manner. In addition to festival-related cases, there are also CIAPP cases hospitalized throughout the year. In this prospective study, hospitalized cases of CIAPP were compared to individuals who 1) were hospitalized for psychosis unrelated to cannabis, 2) hospitalized for cannabis use disorder, and 3) were healthy. Demographic information, personal and family history of mental illness and drug use, psychosis (PANSS), and cognition (Cogstate battery) and psychophysiological indices of information processing (Auditory Steady State Response and noise [LZC]) were assessed 1) at admission, 2) mid-hospitalization, 3) around discharge, and 4) within 6 months post-discharge. In this prospective study, 50 consecutive hospitalized cases of CIAPP with toxicological confirmation of cannabis exposure have been studied and compared to 3 control groups. Preliminary results suggest that cases of CIAPP have a distinct phenomenological presentation including equivalent scores of psychosis, marked affective symptoms, lower schizotypy, and better performance on some cognitive tasks. Furthermore, CIAPP have reduced connectivity but preserve gamma band power. At the time of discharge, and 4–6 months later, CIAPP cases have fewer residual symptoms. Two cases have relapsed after resuming the consumption of cannabis, supporting a role for cannabis in the expression of psychosis. If these results are confirmed in a larger sample, it would suggest that CIAPP may represent a distinct subtype of psychotic disorder that has characteristic behavioral, cognitive and psychophysiological features. Longer longitudinal studies are warranted to understand the course and prognosis of CIAPP.
GPT-4 and PaLM 2 are advanced large language models (LLMs) that have made remarkable progress and are widely used in various human applications. The extent to which these artificial intelligence systems possess social intelligence is pivotal for their seamless integration into everyday human interactions. In this study, we used the SOCRATIS tool to measure and compare the social cognitive abilities in first-order theory of mind, second-order theory of mind, faux pas recognition, and attributional styles between GPT-4 based Microsoft Bing, PaLM 2 based Google Bard, and healthy humans. Our results showed a significant difference in second-order theory of mind, faux pas recognition, and attributional styles among the groups. Specifically, post-hoc comparisons revealed that Bing, based on GPT-4, performed significantly better than Bard, based on PaLM 2, and healthy humans in second-order theory of mind and faux pas recognition tasks. Regarding attributional styles, Bard exhibited a higher personalizing bias compared to both Bing and healthy humans. Microsoft Bing on the other hand had significantly lower externalizing bias score compared to other two groups.
Abstract The selection of an appropriate virtual reference schema is pivotal in determining the outcomes of Event-Related Potential (ERP) studies, particularly within the widely utilized Talk/Listen ERP paradigm, which is employed to non-invasively explore the corollary discharge (CD) phenomenon in the speech-auditory system. This research centers on examining the effects of prevalent EEG reference schemas—Linked Mastoids (LM), Common Average Reference (CAR), and Reference Electrode Standardization Technique (REST)—through statistical analysis, Statistical Parametric Scalp Mapping (SPSM), and source localization techniques. Our ANOVA findings indicate significant main effects for both the reference and the experimental condition on the amplitude of N1 ERPs. Depending on the reference used, the polarity and amplitude of the N1 ERPs demonstrate systematic variations: LM is associated with pronounced frontocentral activity, whereas both CAR and REST exhibit patterns of frontocentral and occipitotemporal activity. The significance of SPSM results is confined to regions exhibiting prominent N1 activity for each reference schema. Source analysis provides corroborative evidence more aligned with the SPSM results for CAR than for REST or LM, suggesting that results under CAR are more objective and reliable. Therefore, the CAR reference is recommended for future studies involving Talk/Listen ERP paradigms.
Background: Social cognition deficit is one of the marked characteristics of schizophrenia. Accumulated evidence suggests that social cognition and interaction training (SCIT) is associated with improved performance in social cognition and social skills in patients diagnosed with psychotic disorders. The cultural influence on social cognition is quite considerable. So, studies in the area of social cognition domains need to adapt and use culturally appropriate tools and measures to see the effectiveness. This study aimed to validate the materials used in SCIT training in Indian setting. Materials and Methods: The original script of video clips was translated into Hindi and was reshot, and the images were remade. A panel of experts rated the videos and images on a 5-point Likert scale. Furthermore, the content validity and internal consistency of the materials were calculated. Results: The content validity ratio (CVR) critical value was 0.357, and all the videos and images received more than the CVR critical value. The intraclass correlation coefficient for videos was 0.974, for SCIT photographs was 0.971, for “spotting character” was 0.975, and for “emotion shaping” was 0.965, indicating good internal consistency. Discussion: The majority of the experts in the panel found the videos and images adequate and appropriate for the Indian setting. In addition, the videos and photographs both yielded good internal consistency.
Objective: The objective of this study was to evaluate the feasibility of using transfer learning with state- of-the-art Convolutional Neural Networks (CNNs) to classify human figure drawings of patients diagnosed with Mania and Schizophrenia.Methods: We collected a total of 230 images (108 from patients with Mania and 122 from patients with Schizophrenia). We experimented with several CNN models, including ResNet-50, InceptionV3, DenseNet-121, VGG19, and Inception-ResNet-v2, by utilizing their pre-trained weights on the ImageNet dataset to extract features from the human figure images. These features were then used to train a dense neural network head for binary classification. We employed 15-fold cross-validation for internal validation.Results: The Inception-ResNet-v2 model demonstrated the best performance, achieving an accuracy of 73.6 ± 9.4%, an AUC score of 0.72 ± 0.09, and an F1 score of 73 ± 20% in classifying Schizophrenia and Mania. Additionally, qualitative analysis revealed that features associated with psychosis and mania, as documented in existing literature, were frequently identified in high-probability classifications.Conclusion: This study underscores the potential of CNNs in classifying human figure drawings of psychiatric patients and highlights the need for continued exploration in this domain with larger and more diverse datasets.
The schizophrenia syndrome likely encompasses multiple clinically related illness manifestations that result from distinct etio-pathological processes converging onto fewer final common pathways. Exposure to cannabis is known to result in a syndrome that clinically mimics schizophrenia-psychosis, outlasts the acute intoxication, persists for days to weeks, requires clinical intervention, and may recur with cannabis exposure. Characterizing the vulnerability to Cannabinoid Induced Acute and Persistent Psychosis (CIAPP), its clinical and neuro-physiological correlates, and its relationship to schizophrenia may enhance our understanding of the neurobiology of schizophrenia in general and specifically, this subtype. Deep learning is an extremely powerful approach to classify (e.g., cancerous vs healthy cells) and use complex stimuli to anticipate future outcomes (e.g., hurricane path) with high accuracy. Thus, deep learning seems a promising approach to differentiate subtypes of complex syndromes such as schizophrenia. In a prospective case-control study at Central Institute of Psychiatry, Ranchi, India, we compared hospitalized cases of CIAPP with two control groups which included: 1) hospitalized cases with psychosis unrelated to cannabis, and (PUC) 2) healthy controls (HC). Demographic and substance use variables, and familial loading for psychiatric illnesses (FIGS) were evaluated at baseline. The following assessments were carried out at four time points - baseline, mid hospitalization, at discharge, and at 6 months post discharge: 1) measures of psychosis (PANSS), mood (YMRS and Calgary depression scale) and cognition (Cogstate battery); 2) psychophysiological variables including resting and Auditory Steady State Response (ASSR) EEG. Electrophysiological and behavioral data were integrated into subject-specific neurobehavioral “fingerprints”, i.e. graph-like objects depicting EEG (2 second epochs) and behavioral information. These fingerprints (~200 per subject) were used to train the classification apparatus of a deep convolutional network (Inception-Res v2) pre-trained for image classification. The trained network was tested on a validation data set from an independent sub-sample of subjects. Data has been collected for 50 consecutive CIAPP cases and 25 controls (15 PUC, 10 HC) and a part of the sample has completed the sixth month follow up. Interim analysis of the data from baseline, mid-hospitalization and at discharge time points suggests that in comparison to PUC, CIAPP has a distinct profile with equivalent psychosis symptoms but more mania-like symptoms and lower pre-morbid schizotypy scores. They also have lower scores on cognitive tests at baseline in specific neurocognitive domains including working memory and recall tests but had better performance in paired associate learning and social cognition tests. Electrophysiological data showed that CIAPP and PUC had reduced gamma-band neural connectivity compared to HC while CIAPP showed levels of gamma-band power comparable to HC. Post-training, the neural network identified and classified neurobehavioral fingerprints from CIAPP, PUC, and HC with >98% accuracy in both the training and independent validation data sets. The preliminary results of this investigation suggest that CIAPP represents a subtype of schizophrenia with distinct neuro-behavioral correlates. Furthermore, deep learning has shown to be useful to classify such disease subtypes. Using this approach in larger training and validation data sets, and inclusion of the longitudinal data from 6-month follow-up may improve the robustness of the neural net classifier.
'%3e%3ccircle r='20' fill='%23008'/%3e%3ccircle r='17.5' fill='%23fff'/%3e%3ccircle r='3.5' fill='%23008'/%3e%3cg id='d'%3e%3cg id='c'%3e%3cg id='b'%3e%3cg id='a' fill='%23008'%3e%3ccircle r='.875' transform='rotate(7.5 -8.75 133.5)'/%3e%3cpath d='M0 17.5.6 7 0 2l-.6 5L0 17.5z'/%3e%3c/g%3e%3cuse xlink:href='%23a' transform='rotate(15)'/%3e%3c/g%3e%3cuse xlink:href='%23b' transform='rotate(30)'/%3e%3c/g%3e%3cuse xlink:href='%23c' transform='rotate(60)'/%3e%3c/g%3e%3cuse xlink:href='%23d' transform='rotate(120)'/%3e%3cuse xlink:href='%23d' transform='rotate(-120)'/%3e%3c/g%3e%3c/svg%3e)
