Mental Status Detection for Schizophrenia Patients via Deep Visual Perception

Lin, Bing-Jhang; Lin, Yi-Ting; Liu, Chen-Chung; Lee, Lue-En; Chuang, Chih-Yuan; Liu, An-Sheng; Hung, Shih-Han; Fu, Li‐Chen

doi:10.1109/jbhi.2022.3199575

Cited by 4 publications

(2 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Schizophrenia patients are at high risk of ending their life by suicide, due to hallucinations. Further, the existing related works [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25] predict schizophrenia with less accuracy, and older age people are not subjected to the detection. These research gaps motivate the research in detecting both younger and older persons having schizophrenia with early diagnosis of delusion and hallucination with high accuracy.…”

Section: Introductionmentioning

confidence: 99%

“…2. status detection In order to determine the patient's mental state during the learning phase, including emotions and depression severity, the author suggests a multi-task learning framework. In the learning phase, this proposed system presents a model name Cross-Modality Graph Convolutional Network (CMGCN) to effectively integrate visual features from different modalities, including the face and context 20 .…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Preclinical Diagnosis of Schizophrenia using EEGTime-Frequency Analysis During Motor Function and verbal Fluency Tests

Arikumar¹,

Tamilarasi

Baza

et al. 2023

Preprint

View full text Add to dashboard Cite

Schizophrenia is a severe mental illness that impairs the way a person perceives reality. It causes a variety of issues related to behavior, emotions, and thinking (cognition). Patients experience auditory hallucinations, delusion, and sleep deprivation. Although the Diagnostic and Statistical Manual (DSM) IV version helps in diagnosis, the lack of clinical tools available leads to delayed diagnosis. To ease preclinical diagnosis, this study conducted two tests, namely Motor Function Test (MFT) and the verbal Fluency Test (VFT). The neural activity of the brain during these two tests is captured by an Electroencephalogram(EEG). However, EEG signals are not widely used for clinical analysis, as the signals are blurred during acquisition, random, and shortcoming of the signal processing algorithm. Hence, in this paper, we propose a Preclinical Diagnosis of Schizophrenia using Multi-SynchroSqueezing Transform (MSST) (PDS-M) to analyze the blurred EEG signal with Time-frequency analysis.PDS-M produces a sharper signal and achieves perfect signal reconstruction. The mono component of MSST in each mode is termed intermediate frequency and it is used as the feature (such as lack of emotion, hallucinations, or disorganized thinking)for discriminating schizophrenic subjects from normal ones. The results show that the MFT is better than the VFT for preclinical study and this study also suggests that MSST is more suitable for time-frequency analysis of EEG signal since all the three modes multi-components are different from normal having a low p-value of p=0, p=0, and p=0.008, respectively, where p is the probability of data having occurred under the null hypothesis of a statistical test.

show abstract