Major differences exist between men and women in both physiology and pathophysiology. Dissecting the underlying processes and contributing mechanisms of sex differences in health and disease represents a crucial step towards precision medicine. Considering the significant differences between men and women in the response to pharmacotherapies, our aim was to develop an in silico model able to predict sex-specific drug responses in a large-scale. For this purpose, we focused on cardiovascular effects because of their high morbidity and mortality. Our model predicted several drugs (including acebutolol and tacrine) with significant differences in the heart between men and women. To validate the sex-specific drug responses identified by our model, acebutolol was selected to lower blood pressure in spontaneous hypertensive rats (SHR), tacrine was used to assess cardiac injury in mice and metformin as control for a non-sex-specific response. As our model predicted, acebutolol exhibited a stronger decrease in heart rate and blood pressure in female than male SHRs. Tacrine lowered heart rate in male but not in female mice, induced higher plasma cTNI level and increased cardiac superoxide (DHE staining) generation in female than male mice, indicating stronger cardiac toxicity in female than male mice. To validate our model in humans, we employed two Chinese cohorts, which showed that among patients taking a beta-receptor blocker (metoprolol), women reached significantly lower diastolic blood pressure than men. We conclude that our in silico model could be translated into clinical practice to predict sex-specific drug responses, thereby contributing towards a more appropriate medical care for both men and women.

As the basic units of the human body structure and function, cells have a considerable influence on maintaining the normal work of the human body. In medical diagnosis, cell examination is an important part of understanding the human function. Incorporating cell examination into medical diagnosis would greatly improve the efficiency of pathological research and patient treatment. In addition, cell segmentation and identification technology can be used to quantitatively analyze and study cellular components at the molecular level. It is conducive to the study of the pathogenesis of diseases and to the formulation of highly effective disease treatment programs. However, because cells are of diverse types, their numbers are huge, and they exist in the order of micrometers, detecting and identifying cells without using a deep learning-based computer program are extremely difficult. Therefore, the use of computers to study and analyze cells has a certain practical value. In this work, target detection theory using deep learning is applied to cell detection. A target recognition network model is built based on the faster region-based convolutional neural network (R-CNN) algorithm, and the anchor box is designed in accordance with the characteristics of the data set. Different design methods influence cell detection results. Using the object detection method based on our novel faster R-CNN framework to detect the cell image can help improve the speed and accuracy of cell detection. The method has considerable advantages in dealing with the identification of flowing cells.

In recent years, the research on electroencephalography (EEG) has focused on the feature extraction of EEG signals. The development of convenient and simple EEG acquisition devices has produced a variety of EEG signal sources and the diversity of the EEG data. Thus, the adaptability of EEG classification methods has become significant. This study proposed a deep network model for autonomous learning and classification of EEG signals, which could self-adaptively classify EEG signals with different sampling frequencies and lengths. The artificial design feature extraction methods could not obtain stable classification results when analyzing EEG data with different sampling frequencies. However, the proposed depth network model showed considerably better universality and classification accuracy, particularly for EEG signals with short length, which was validated by two datasets.

A critical step for gene selection algorithms using rough set theory is the establishment of a gene evaluation function to assess the classification ability of candidate gene subsets. The concept of dependency in a classic neighborhood rough set model plays the role of this evaluation function. This criterion only notes the information provided by the lower approximation and omits the upper approximation, which may result in the loss of some important information. This paper proposes gene selection algorithms within a single-cell gene decision space by employing self-information, taking into account both lower and upper approximations. Initially, the distance between gene expression values within each subspace is defined to establish the tolerance relation on the cell set. Subsequently, self-information is introduced through the lens of tolerance classes. The relationship between these measures and their respective properties is then examined in detail. For gene expression data, the proposed self-information metric demonstrates superiority over other measures by accounting for both lower and upper approximations, thereby facilitating the selection of optimal gene subsets. Finally, gene selection algorithms within a single-cell gene decision space are developed based on the proposed self-information metric, and experiments conducted on 10 publicly available single-cell datasets indicate that the classification performance of the proposed algorithms can be enhanced through the selection of genes pertinent to classification. The results demonstrate that $Fi-SI$ achieves an average classification accuracy of 93.7% (KNN) while selecting 48.3% fewer genes than Fisher’s score.