Background: Impedance cardiography (ICG) offers a potential alternative for hemodynamic assessment in pulmonary arterial hypertension (PAH) as a non-invasive technique. Methods: A total of 132 patients who underwent right heart catheterization (RHC) were included. Cardiac output (CO) and stroke volume (SV) measured by thermodilution during RHC (COTD) and ICG (COICG) were compared. The capacity of ICG in PAH risk stratification and clinical deterioration prediction was also analyzed. Results: Ninety-three pre-capillary pulmonary hypertension patients were enrolled, 54 (58.06%) patients belong to Group 1 PAH, and 39 (41.94%) patients were diagnosed with chronic thromboembolic pulmonary hypertension. The mean COTD was 4.93 ± 1.06 L/min, while the COICG was 4.41 ± 1.23 L/min, showing a moderate correlation (r = 0.49, P < .001). In Group 1 PAH patients, the COTD was 5.13 ± 1.10 L/min, and COICG was 4.57 ± 1.22 L/min (r = 0.52, P < .001). Bland–Altman analysis indicated a mean difference of 0.52 L/min and limits of agreement from −1.76 to 2.80 L/min. The mean SVTD was 64.63 ± 17.10 mL, and the SVICG was 60.94 ± 18.03 mL (r = 0.53, P < .001) with a mean difference of 3.69 mL. After a 1-year follow-up, the CIICG and SVIICG showed potential power in predicting clinical deterioration in PAH patients, with area under the curves of 0.76 and 0.81, respectively. Conclusion: Impedance cardiography measured CO and SV presented an acceptable correlation with RHC in PAH patients. Stroke volume index and cardiac index measured by ICG is potent to identify the low-risk status and predict clinical deterioration in PAH patients.

This paper proposes Temporal-decomposition and Attention-based Denoising Network (TADNet) for magnetocardiography (MCG) signal denoising. The raw time-series data are decomposed into trend and seasonal components. The trend component is modeled using a multi-layer perceptron to capture the trend of the data. For seasonal component, a multi-layer convolutional neural network is employed to perform segmentation and extract local temporal features. These fine-grained features are processed by a Transformer-based self-attention mechanism to capture long-range temporal dependencies. Finally, deconvolution operations are applied to reconstruct the signal, achieving effective noise reduction for MCG data. This paper conducts in-depth studies on effects of time, wavelet, and Fourier style attention on model performance, enhancing the interpretability of the model. Extensive experiments on clinical MCG data under various noise conditions demonstrate that TADNet outperforms traditional algorithms, proving its effectiveness and robustness. The research results indicate that TADNet provides a highly efficient and accurate solution for MCG signal denoising.

Accurate detection and risk assessment are crucial for the prognosis of patients with pulmonary hypertension (PH). This study aimed to develop a detection model by employing ultra‐high‐sensitivity magnetocardiography (MCG) combined with machine learning while exploring the potential of MCG features in enhancing prognostic evaluation. PH patients were allocated into the exploratory and external validation cohort according to enrollment period in a single‐center. The control group comprises patients with symptoms including dyspnea, chest tightness, chest pain, or fatigue in the absence of PH. Outpatient follow‐up is performed for PH patients in the exploratory cohort. Seven machine learning algorithms and 15 out of 40 MCG features were employed to develop PH detection models. Cox regression is used to construct a risk assessment model. The exploratory cohort includes 312 subjects (156 PH patients), whereas the external validation cohort includes 162 subjects (81 PH patients). The Random Forest algorithm demonstrates superior performance, achieving sensitivities of 79.6% and 84.0% and specificities of 92.4% and 91.4% in different datasets, respectively. During the average follow‐up period of 33 weeks, 14 patients experienced clinical deterioration events. Compared to the baseline model developed using three parameters from the guideline‐recommended model, the combined model incorporating 2 MCG features performs better in this small cohort to detect deterioration events. In conclusion, the Random Forest‐based MCG model demonstrated robust detection accuracy for PH. The integration of 2 MCG features with clinical baseline parameters may improve short‐term risk assessment.

Background: Exploring accurate and noninvasive methods for detecting pulmonary hypertension (PH) has always been a focal point of research. Owing to its exceptional spatiotemporal resolution, magnetocardiography (MCG) has demonstrated potential value in cardiovascular diseases. Aims: This exploratory study aims to investigate the characteristics of MCG variations in PH patients and evaluate their potential utility in distinguishing healthy subjects from those with PH. Methods: This study analyzed 175 PH patients and 333 healthy subjects who underwent MCG examination. The training cohort consisted of patients with PH previously diagnosed by right heart catheterization (RHC) and age‐frequency‐matched healthy controls (HC). The testing cohort comprised age‐ and frequency‐matched HC and PH patients who underwent both MCG and RHC on the same day. Nine MCG parameters were included. Logistic regression was used to screen for significant parameters and develop a model. Results: By comparing the pseudo‐current density maps, it was found that the current vector at the R‐wave peak of HC points toward the lower‐left quadrant, whereas in PH patients, it points toward the lower‐right quadrant. The MCG detection model demonstrated robust performance, achieving a sensitivity of 86.1% and a specificity of 94.1% in the testing cohort. Compared to the ECG of PH patients, MCG demonstrated greater sensitivity; however, it exhibited slightly lower specificity. Furthermore, MCG can detect PH in patients with normal ECG findings. Conclusion: MCG demonstrates highly promising potential for the noninvasive detection of PH.

This study aimed to characterize cortical dysfunction and frequency-specific network reorganization following optic nerve injury using spin-exchange relaxation-free magnetoencephalography (SERF-MEG), and to assess the potential of MEG-derived multiscale features as sensitive functional biomarkers for clinical evaluation. In this prospective case–control study, SERF-MEG recordings were acquired during a pattern-reversal visual stimulation paradigm. Time-domain evoked components (M100/M135), global electrophysiological indices, energy-based metrics, and alpha- and beta-band phase-based functional connectivity were extracted. Network topology was quantified using graph-theoretical measures, including global and local efficiency, clustering coefficient, and assortativity. Group-level differences between patients and healthy controls were statistically analyzed. Patients showed significantly reduced M100/M135 amplitudes, prolonged M100 latency, and a lower early-component energy ratio. Functional connectivity was significantly decreased in the alpha and beta bands, accompanied by reduced global and local efficiency, mean strength, and clustering coefficient. Seed-based analyses revealed reduced connectivity predominantly in occipito-parietal and occipito-temporal pathways. SERF-MEG provides sensitive identification of cortical- and network-level functional impairments following optic nerve damage. MEG has significant clinical potential for disease diagnosis and therapy monitoring, providing a novel objective assessment tool for neuro-ophthalmological disorders.

Non-invasive methods used for PH detection in clinical practice have several limitations. The combination of high spatiotemporal sensitivity magnetocardiography (MCG) and artificial intelligence algorithms may offer an accurate approach for PH detection. In this study, we develop a convolutional neural network (CNN) model based on the 64-channel MCG time-series data. This exploratory study enrolled patients undergoing 64-channel MCG, including right-heart-catheterization confirmed PH patients and symptomatic controls with low echocardiographic probability of PH. After data preprocessing, a temporal CNN integrating MCG signals with age, sex, and body mass index was developed and compared with conventional machine learning models. The CNN model achieved strong discrimination, with area under the curve (AUC) values of 0.939 (95% confidence interval [CI]: 0.913–0.961) in the development out-of-fold evaluation and 0.974 (95% CI: 0.944–0.994) in the hold-out test set, outperforming conventional machine learning models. Decision curve analysis showed the greatest net benefit at clinically relevant thresholds. Attribution analysis indicated that spatial QRS morphology redistribution contributed substantially to PH classification. The temporal CNN model based on raw 64-channel MCG signals showed promising performance for non-invasive PH detection and outperformed conventional machine learning approaches in this exploratory single-center cohort enriched for PAH and CTEPH.