Medical image registration (MIR) is essential for various clinical diagnoses and treatments. Despite the rapid progress in deep learning-based MIR techniques, most methods focus on directly optimizing the raw image intensity information. In this paper, we explore the usage of edge information of anatomical structures associated with the spatial location of image intensities to assist in registration, termed EdgeReg. The intuition is that the edge information can provide additional rich boundary information to the raw images, enhancing the network’s feature representation. Additionally, as the edge images are strictly spatially consistent with the raw images, additional supervised information can be added to network training. Specifically, we first extract the edge images from the raw moving and fixed images using the Sobel operator and feed these images into a lightweight feature extractor to merge the image intensity and edge information. The enriched features are subsequently input into established registration networks. Finally, similarity loss is applied to both the raw and edge images. Extensive experiments show that EdgeReg is compatible with various networks across diverse datasets and dimensions (2D and 3D), achieving superior registration performance. In particular, EdgeReg does not rely on segmentation labels and is trained in an unsupervised paradigm. Therefore, edge information is a beneficial assistance for unsupervised MIR. The code is available at https://github.com/PerceptionComputingLab/EdgeReg.

Recently, self-training strategies have shown outstanding performance in the unsupervised medical image registration field. These strategies use their own network to generate pseudo-displacement fields (PFs) to supervise network training. However, limited diversity and accuracy of these PFs hinder their effectiveness. To address these limitations, we propose a novel mutual learning registration paradigm (MutualReg), where knowledge is distilled mutually between teacher and student networks for alternate improvement via recursive training. This involves two fundamental challenges: 1) how to generate more diverse and accurate PFs; and 2) how to effectively integrate knowledge distillation from the teacher network and learning from the student network. For the former, we employ a different and powerful teacher network thanks to the decoupling nature of MutualReg. For the latter, we introduce a Voxel-wise Reliability Criterion (VRC) module to retain reliable voxel locations of knowledge distillation. In the abdominal CT registration task, MutualReg outperforms state-of-the-art competitors, demonstrating its effectiveness. Code is available from https://github.com/PerceptionComputingLab/MutualReg/.

The shortage of annotated ECG data presents a significant impediment, hampering the overall generalization capabilities of machine learning models tailored for automated ECG classification. The collective integration of multisource datasets presents a potential remedy for this challenge. However, it is crucial to underscore that the mere addition of supplementary data does not automatically guarantee performance enhancement, given the unresolved challenges associated with multisource data. In this research, we address one such challenge, namely, the issue of incomplete labels arising from the diversity of annotations within multi-source ECG datasets. First, we identified three distinct types of label missing: dataset-related label missing, supertype missing, and subtype missing. To address the supertype missing effectively, we introduce a novel approach known as offline category mapping which leverages the hierarchical relationships inherent within the categories to recover the missing supertype labels. Additionally, two complementary strategies, referred to as prediction masking and online category mapping, are proposed to mitigating the adverse effects of subtype and dataset-related label missing on model optimization. These strategies enhance the model's ability to identify missing subtypes under conditions of weak supervision. These pioneering methodologies are integrated into a deep learning-based framework designed for multilabel ECG classification. The performance of our proposed framework is rigorously evaluated using realistic multi-source datasets obtained from the PhysioNet/CinC challenge 2020/2021. The proposed learning framework exhibits a notable improvement in macro-average precision, surpassing the corresponding baseline model by more than 25 % on the test datasets. As a result, this research study makes a substantial contribution to the field of ECG classification by addressing the critical issue of incomplete labels in multisource datasets, ultimately enhancing the generalization capabilities of machine learning models in this domain.

Previous studies have shown that coronary artery occlusion can cause myocardial ischemia, which can induce ventricular tachycardia or fibrillation. According to the time sequence, myocardial ischemia can be divided into different pathological stages: the ischemia 1a stage (0-15 minutes), the ischemia 1b stage (15-45 minutes), the short-term myocardial infarction (MI) (within a few days) and the long-term MI (within a few weeks). However, few studies give attention to antiarrhythmic drugs directly acting on ion channels for the treatment of myocardial ischemia. The main reason is that the effective targets in myocardial ischemia are unclear. Therefore, based on the technology of electrophysiological simulation, the paper modeled the human ventricular cell model in myocardial ischemia. On this basis, the parameter sensitivity of the model was analyzed and the effective drug targets are selected. Firstly, human ventricular cell models were modeled in ischemia 1a, ischemia 1b, short-term MI and long-term MI based on the experimental data. Then, the sensitivity analysis of output parameters (APA, APD and RP) of the cell model in ischemia was analyzed based on the Sobol method. The results of parameter sensitivity analysis in this paper showed that APD of cell models in ischemia 1a, ischemia 1b and short-term MI was the most sensitive to the change of IKATP, and APA and RP of cells were the most sensitive to the change of [K+]o. The parameter sensitivity analysis of the cell model in long-term MI showed that APD of cell models was sensitive to IKsand ICaL, and APA was most sensitive to INa. Therefore, according to the results of parameter sensitivity analysis, the possible effective targets for the treatment of myocardial ischemia can be preliminarily selected: [K+]o, IKATP, ICaLand INa.

Arsenic trioxide $(As2\mathrm{O}_{3}$), an antileukemia drug, has been used to treat acute promyelocytic leukemia (APL) for more than fifty years, and its therapeutic effect has been elucidated at the molecular level. However, several side effects were observed in APL patients administrated with $As2\mathrm{O}_{3}$, such as long QT (LQT) syndrome, torsade de pointes tachycardia, and even sudden cardiac death. This means that the clinically relevant dosage may induce severe cardiotoxicity. Accordingly, it is essential to determine the underlying mechanisms of arrhythmia induced by $\mathrm{As}2\mathrm{O}_{3}$. Some biological experiments indicated that $\mathrm{As}2\mathrm{O}_{3}$ can impair the human ether-à-go-gorelated gene (hERG), thus inhibiting rapid delayed rectifier potassium current $(I_{Kr})$ and prolonging action potential duration (APD), which was regarded as the reason for LQT syndrome. However, previous experiments did not illuminate the deep mechanisms of $\mathrm{As}2\mathrm{O}_{3}$-induced side effects, which is important in clinical treatment. In addition, the experimental data were restricted to animal studies, so human cellular data were lacking. In this study, we investigated $\mathrm{As}2\mathrm{O}_{3}$-related cardiotoxicity through a human ventricular model study. Based on the current experimental data, the effects of $\mathrm{As}2\mathrm{O}_{3}$ on ventricular myocytes (VMs) were predicted at various $\mathrm{As}2\mathrm{O}_{3}$ concentrations. In addition, the potential hazard of $\mathrm{As}2\mathrm{O}_{3}$ was simulated and illustrated under different stimulation protocols. Moreover, electrocardiograms (ECGs) were estimated in heterogeneous ventricular cables, by which the clinical phenomenon was verified and explained. Based on the present modeling study, deep reasons for arrhythmia caused by $\mathrm{As}2\mathrm{O}_{3}$ were uncovered. $\mathrm{As}2\mathrm{O}_{3}$ not only led to a prolonged APD but also alternated action potentials and exacerbated heterogeneity among VMs. Moreover, the degree of arrhythmia risk was susceptible to $\mathrm{As}2\mathrm{O}_{3}$ dosage. These new findings may provide targets for attenuating $\mathrm{As}2\mathrm{O}_{3}$ toxicity and may help to improve the APL therapeutic regimen.

Cardiovascular obstruction could lead to myocardial ischemia and myocardial infarction (MI). MI can be further divided into short-term MI stage (several days) and long-term MI stage (several months) with the development of coronary artery obstruction, and the electrophysiological characteristics in these two MI stages vary greatly. At present, there are no relevant studies on the effects of different infarct areas (size and location) on the initialization and maintenance of reentrant waves in these two MI stages. Therefore, this study aims to investigate the differences in vulnerability to reentry between these two MI stages by computer modeling and simulation. Firstly, single cell models, based on the TP06 model were developed in two different MI stages. And simulation results on single-cells showed that the action potential duration (APD) significantly shortened and the resting potential (RP) elevated in the short-term MI stage, compared with that in the normal condition. However, APD prolonged and RP only changed little in the long-term MI stage. When MI areas in 2D annular ventricular tissues were designed with different lengths, widths and positions, the distribution of the vulnerable window (VW) in these two MI stages was investigated. The simulation results showed that the vulnerability of the two MI stages to the length and position of the infarct areas is the same. That is with the increase of the length, VW gradually increased and reached a constant value when the percentage of the length of the MI area reached 50%. And VW was the largest when the infarct area was close to the inner or outer wall. The vulnerability to the width of the infarct area in these two MI stages is different. In short-term MI, VW was the largest when the width of the infarct area was narrow or wide, while in long-term MI, VW was the largest when the width of the MI area reached half of the width of the ventricular wall. In this paper, the effect of the different infarct areas on the initialization and maintenance of reentrant waves in two different MI stages was investigated by computing simulation. This would improve the understanding of arrhythmogenicity in the MI stage and could provide new sights in arrhythmogenic mechanism of MI phases.

Myocardial ischemia and myocardial infarction (MI) are often accompanied by the occurrence of reentrant arrhythmias, which may lead to sudden cardiac death in severe cases. Previous studies show that electrical alternans can occur during myocardial ischemia and MI and may lead to arrhythmias. However, so far, the mechanism of alternans during myocardial ischemia and MI is unclear, so the related study on alternans is particularly important. Based on single-cell models previously modeled by us at three stages: ischemia 1a, 1b, and MI, the mechanism of alternans was revealed by comparing the changes in alternans at three levels: single cells, one-dimensional (1D) tissues, and two-dimensional (2D) tissues. In addition, the main factors inducing alternans were investigated, and the effect of antiarrhythmic drug glibenclamide on alternans was simulated. The simulation results on single cells of ischemia 1a, 1b and MI showed that the electrical alternans on the cell-levels were unstable electrical alternans. Simulation results in tissues showed that stable electrical alternans could occur in both 1D and 2D tissues. Simulation results showed that alternans in ischemia 1a were mainly caused by two factors: inhibition of $\mathrm{I}_{\mathrm{Na}}$ and elevation of $[\mathrm{K}^{+}]_{\mathrm{o}}$; alternans in ischemia 1b were mainly caused by two factors: inhibition of $\mathrm{I}_{\mathrm{NaK}}$ and elevation of $[\mathrm{K}^{+}]_{\mathrm{o}}$; alternans in MI were mainly caused by three factors: inhibition of $\mathrm{I}_{\mathrm{Kr}}$, inhibition of $\mathrm{I}_{\mathrm{Ks}}$, and elevation of $[\mathrm{K}^{+}]_{\mathrm{o}}$. And electrical alternans in the tissues result in a 2:1 conduction block. In addition, the simulation results showed that glibenclamide could inhibit electrical alternans in single cells and tissues. Electrical alternans during ischemia 1a, 1b and MI are caused by several currents that directly affect the action potential, and can lead to a 2:1 conduction block in tissues. Glibenclamide inhibits the occurrence of electrical alternans by inhibiting the efflux of potassium ions.