Predictive spread of obsessive-compulsive disorder pathology using the network diffusion model.

An increasing body of studies propose that structural abnormalities begin with focal brain regions then propagate to other regions following the architecture of healthy brain network in neuropsychiatric disorders. However, these findings are untested in obsessive-compulsive disorder (OCD), also showing widespread structural brain abnormalities. In this study, we aimed to investigate whether healthy functional brain network contributed to structural brain abnormalities in OCD. The gray matter morphological abnormalities were obtained in 98 patients with OCD in relative to matched healthy controls (n = 130, HCs). The network diffusion model (NDM) was conducted to identify putative seed regions and patterns of disease propagation from seed regions to other brain regions along the functional network in OCD. The NDM has been proved to succeeded in capturing the trans-neuronal propagation of pathology and even in predicting future longitudinal progression of pathology in neurodegenerative diseases. In this study, when seeding at the right anterior cingulate cortex, the NDM best recapitulated the patterns of gray matter morphological abnormalities, suggesting this region was the most likely seed region. Further analyses revealed that pathology preferentially propagated to higher order brain systems from seed region. For non-seed regions, the arrival time of pathology was negatively correlated with their shortest functional paths to the seed (r = -0.46, p < 0.001). These results suggest that gray matter morphological abnormalities are constrained by healthy brain network and reveal temporal sequencing of pathology progression in OCD.

Cervical cell extraction network based on optimized yolo.

Early screening for cervical cancer is a common form of cancer prevention. In the microscopic images of cervical cells, the number of abnormal cells is small, and some abnormal cells are heavily stacked. How to solve the segmentation of highly overlapping cells and realize the identification of single cells from overlapping cells is still a heavy task. Therefore, this paper proposes an object detection algorithm of Cell_yolo to effectively and accurately segment overlapping cells. Cell_yolo adopts a simplified network structure and improves the maximum pooling operation, so that the information of the image is preserved to the greatest extent during the model pooling process. Aiming at the characteristics of many overlapping cells in cervical cell images, a non-maximum suppression method of center distance is proposed to prevent the overlapping cell detection frame from being deleted by mistake. At the same time, the loss function is improved and the focus loss function is added to alleviate the imbalance of positive and negative samples in the training process. Experiments are conducted on a private dataset (BJTUCELL). Experiments have verified that the Cell_yolo model has the advantages of low computational complexity and high detection accuracy, and it is superior to common network models such as YOLOv4 and Faster_RCNN.