Background and Aim: In the era of deep learning, numerous models have emerged in the literature and various application domains. Transformer architectures, particularly, have gained popularity in deep learning, with diverse transformer-based computer vision algorithms. Attention convolutional neural networks (CNNs) have been introduced to enhance image classification capabilities. In this context, we propose a novel attention convolutional model with the primary objective of detecting bipolar disorder using optical coherence tomography (OCT) images. Materials and Methods: To facilitate our study, we curated a unique OCT image dataset, initially comprising two distinct cases. For the development of an automated OCT image detection system, we introduce a new attention convolutional neural network named “TurkerNeXt”. This proposed Attention TurkerNeXt encompasses four key modules: (i) the patchify stem block, (ii) the Attention TurkerNeXt block, (iii) the patchify downsampling block, and (iv) the output block. In line with the swin transformer, we employed a patchify operation in this study. The design of the attention block, Attention TurkerNeXt, draws inspiration from ConvNeXt, with an added shortcut operation to mitigate the vanishing gradient problem. The overall architecture is influenced by ResNet18. Results: The dataset comprises two distinctive cases: (i) top to bottom and (ii) left to right. Each case contains 987 training and 328 test images. Our newly proposed Attention TurkerNeXt achieved 100% test and validation accuracies for both cases. Conclusions: We curated a novel OCT dataset and introduced a new CNN, named TurkerNeXt in this research. Based on the research findings and classification results, our proposed TurkerNeXt model demonstrated excellent classification performance. This investigation distinctly underscores the potential of OCT images as a biomarker for bipolar disorder.

Aim: The aim of this study was to examine the relationship between peripheral inflammatory markers and aggression and impulsivity in schizophrenia patients with and without criminal histories. Materials and Methods: The study was conducted with patients with schizophrenia involved in crimes and hospitalized in the Forensic Psychiatry ward of Elazığ Fethi Sekin City Hospital and patients with schizophrenia not involved in crimes and hospitalized in the psychiatry ward of Elazığ Mental Health and Diseases Hospital. All participants completed the Buss–Waren Aggression Scale (BWAS), the Barratt Impulsiveness Scale-11 (BIS-11), and the Positive and Negative Symptom Scale (PANSS). Before treatment, venous blood samples were taken for laboratory measurements on the first day of hospitalization. Results: All participants were male. The mean age of those involved in a crime was 39 ± 9.7 years, while the mean age of those not involved in a crime was 41.2 ± 10.7 years. The PANSS all subscale and total scores of the patients with schizophrenia who were involved in a crime were significantly higher than the group who were not involved (p values were p < 0.001, p = 0.001, p = 0.043, p = 0.001, respectively). The BWAS—physical aggression (p = 0.007) and total scores of the scale (p = 0.046) and BIS-11—inability to plan (p = 0.002) scores of the group involved in a crime were higher than the group not involved. As for laboratory parameters, MCH, MCHC, PDW, eosinophils, basophils, RDW-CV, and RDW-SD values were significantly higher in those involved in crime, while MPV, creatinine, albumin, and LDH values were lower. CRP and CRP/albumin values were significantly higher, while neutrophil/albumin values were significantly lower in those who committed murder in the first degree than those who committed other crimes. Conclusion: Based on our results, we found that inflammatory agents were significantly increased in forensic schizophrenia patients with high aggression scores. Significant correlations between some inflammatory factors and impulsivity and aggression scores and differences in these factors according to crime types showed that these factors might be related to violence and criminal behavior. There is a need for further large-scale studies on this subject at different stages of the disease.

It has been thought that oxidative damage may occur in the pathophysiology of schizophrenia; metallothioneins (MT) have strong antioxidant functions. In this study, we aimed to measure MT-1 levels in schizophrenia patients. A total of 52 patients diagnosed with schizophrenia and 38 healthy controls were included in the study. Serum MT-1 concentrations were measured using the Human Metallothionein-1 ELISA Kit. In addition, Cu and Zn levels were measured. PANSS (Positive and Negative Syndrome Scale) was used to determine the disease severity of patients with schizophrenia. The MT-1 levels of the schizophrenia group were lower than the MT-1 levels of the control group. When the correlation analyses were examined, a positive correlation was found between MT-1 and illness duration and Cu/Zn. A negative correlation was found between MT-1 levels and PANSS total scores and PANSS positive scores. In the regression analysis, it was seen that the decrease in MT-1 levels poses a risk for schizophrenia. It was observed that a decrease of 1 ng/mL in MT-1 levels increased the risk of schizophrenia 1.115 times. The low concentration of MT-1 is likely to cause a deficiency in antioxidant defense in patients with schizophrenia. MT-1 may be a useful biomarker for predicting schizophrenia.

Background: Electroencephalography (EEG) signal-based machine learning models are among the most cost-effective methods for information retrieval. In this context, we aimed to investigate the cortical activities of psychotic criminal subjects by deploying an explainable feature engineering (XFE) model using an EEG psychotic criminal dataset. Methods: In this study, a new EEG psychotic criminal dataset was curated, containing EEG signals from psychotic criminal and control groups. To extract meaningful findings from this dataset, we presented a new channel-based feature extraction function named Zipper Pattern (ZPat). The proposed ZPat extracts features by analyzing the relationships between channels. In the feature selection phase of the proposed XFE model, an iterative neighborhood component analysis (INCA) feature selector was used to choose the most distinctive features. In the classification phase, we employed an ensemble and iterative distance-based classifier to achieve high classification performance. Therefore, a t-algorithm-based k-nearest neighbors (tkNN) classifier was used to obtain classification results. The Directed Lobish (DLob) symbolic language was used to derive interpretable results from the identities of the selected feature vectors in the final phase of the proposed ZPat-based XFE model. Results: To obtain the classification results from the ZPat-based XFE model, leave-one-record-out (LORO) and 10-fold cross-validation (CV) methods were used. The proposed ZPat-based model achieved over 95% classification accuracy on the curated EEG psychotic criminal dataset. Moreover, a cortical connectome diagram related to psychotic criminal detection was created using a DLob-based explainable artificial intelligence (XAI) method. Conclusions: In this regard, the proposed ZPat-based XFE model achieved both high classification performance and interpretability. Thus, the model contributes to feature engineering, psychiatry, neuroscience, and forensic sciences. Moreover, the presented ZPat-based XFE model is one of the pioneering XAI models for investigating psychotic criminal/criminal individuals.

Major Depressive Disorder (MDD) is a worldwide common disease with a high risk of becoming chronic, suicidal, and recurrence, with serious consequences such as loss of workforce. Objective tests such as EEG, EKG, brain MRI, and Doppler USG are used to aid diagnosis in MDD detection. With advances in artificial intelligence and sample data from objective testing for depression, an early depression detection system can be developed as a way to reduce the number of individuals affected by MDD. In this study, MDD was tried to be diagnosed automatically with a deep learning-based approach using EEG signals. In the study, 3-channel modma dataset was used as a dataset. Modma dataset consists of EEG signals of 29 controls and 26 MDD patients. ResNet18 convolutional neural network was used for feature extraction. The ReliefF algorithm is used for feature selection. In the classification phase, kNN was preferred. The accuracy was yielded 95.65% for Channel 1, 87.00% for Channel 2, and 86.94% for Channel 3.