Background/Objectives: Heart failure (HF) is a growing global health burden, yet early detection remains challenging due to the limitations of traditional diagnostic tools such as electrocardiograms (ECGs). Recent advances in deep learning offer new opportunities to identify left ventricular systolic dysfunction (LVSD), a key indicator of HF, from ECG data. This study validates AiTiALVSD, our previously developed artificial intelligence (AI)-enabled ECG Software as a Medical Device, for its accuracy, transparency, and robustness in detecting LVSD. Methods: This retrospective single-center cohort study involved patients suspected of LVSD. The AiTiALVSD model, based on a deep learning algorithm, was evaluated against echocardiographic ejection fraction values. To enhance model transparency, the study employed Testing with Concept Activation Vectors (TCAV), clustering analysis, and robustness testing against ECG noise and lead reversals. Results: The study involved 688 participants and found AiTiALVSD to have a high diagnostic performance, with an AUROC of 0.919. There was a significant correlation between AiTiALVSD scores and left ventricular ejection fraction values, confirming the model’s predictive accuracy. TCAV analysis showed the model’s alignment with medical knowledge, establishing its clinical plausibility. Despite its robustness to ECG artifacts, there was a noted decrease in specificity in the presence of ECG noise. Conclusions: AiTiALVSD’s high diagnostic accuracy, transparency, and resilience to common ECG discrepancies underscore its potential for early LVSD detection in clinical settings. This study highlights the importance of transparency and robustness in AI-ECG, setting a new benchmark in cardiac care.

Objectives: Early hemorrhage detection in intensive care units (ICUs) enables timely intervention and reduces the risk of irreversible outcomes. In this study, we aimed to develop a machine learning model to predict hemorrhage by learning the patterns of continuously changing, real-world clinical data.Methods: We used the Medical Information Mart for Intensive Care databases (MIMIC-III and MIMIC-IV). A recurrent neural network was used to predict severe hemorrhage in the ICU. We developed three machine learning models with an increasing number of input features and levels of complexity: model 1 (11 features), model 2 (18 features), and model 3 (27 features). MIMIC-III was used for model training, and MIMIC-IV was split for internal validation. Using the model with the highest performance, external verification was performed using data from a subgroup extracted from the eICU Collaborative Research Database.Results: We included 5,670 ICU admissions, with 3,150 in the training set and 2,520 in the internal test set. A positive correlation was found between model complexity and performance. As a measure of performance, three models developed with an increasing number of features showed area under the receiver operating characteristic (AUROC) curve values of 0.61–0.94 according to the range of input data. In the subgroup extracted from the eICU database for external validation, an AUROC value of 0.74 was observed.Conclusions: Machine learning models that rely on real clinical data can be used to predict patients at high risk of bleeding in the ICU.

SnS₂and Sn_1−xTi_xS₂(x= 0, 0.1, 0.3, 0.5, and 0.7 mol) materials were designed using solvothermal method with the aim to enhance hydrogen production from water/methanol water photosplitting. Scanning electron microscopy revealed hexagonal plates with one side, 3.0 μm in length, in the SnS₂materials. Pure SnS₂showed absorption band edges of above 660 nm, and the absorption was shifted to low wavelengths with the insertion of Ti ions. The evolution of H₂from MeOH/H₂O (1 : 1) photosplitting over SnS₂hexagonal plates in the photocatalytic liquid system was 0.016 mL h⁻¹ g⁻¹, and the evolutions were enhanced in Sn_1−xTi_xS₂. In particular, 0.049 mL h⁻¹ g⁻¹of H₂gas was produced in Sn_0.7Ti_0.3S₂without electrolytes and it increased significantly to more than 90.6% (0.47 mL h⁻¹ g⁻¹evolutions) at higher pH using 0.1 M of KOH. Based on the UV-visible absorption spectra, the high photocatalytic activity of Sn_1−xTi_xS₂was attributed to the existence of an appropriate band-gap state that retarded recombination between the electrons and holes.

Jeon KH, et al. Korean Circ J. 2023 Aug;53:e63. https://doi.org/10.4070/kcj.2023.0009

Background: Left ventricular filling pressure is associated with heart failure symptoms and a key prognostic marker and therapeutic target, but a scalable, accessible, and affordable tool for its noninvasive, serial estimation remains lacking. We developed an artificial intelligence (AI) model using a standard 12‐lead ECG to detect increased E/e', a general surrogate of elevated left ventricular filling pressure on echocardiography. Methods: The AI model was built upon a foundation model trained with >1 million multiethnic ECGs and fine‐tuned through a development cohort of 225 737 ECGs and 115 982 echocardiogram records from 92 775 unique patients across 2 tertiary hospitals. The model performance was assessed in a separate internal test set (n=9278) and an independent external cohort from another tertiary hospital (n=17 926). Prognostic significance of the AI‐ECG output was evaluated in these hospital cohorts, as well as the UK Biobank (n=43 347). Finally, we validated the model output against invasively measured left ventricular filling pressure through cardiac catheterization (n=60). Results: The AI‐ECG model detected increased E/e' with an area under the curve of 0.868 (95% CI, 0.859–0.877) and 0.850 (95% CI, 0.841–0.858) in the internal and external test cohorts, respectively. The AI‐ECG output value demonstrated a strong correlation with invasively measured left ventricular end‐diastolic pressure (Pearson's r=0.655) and was significantly associated with incident heart failure and mortality. Conclusions: The AI‐ECG may enable identification of patients with increased left ventricular filling pressure and provide powerful prognostic information. Further prospective studies are warranted to evaluate its clinical utility.

UNSTRUCTURED AI-ECG models trained on 12-lead–derived Lead I demonstrated comparable endpoint discrimination on smartwatch-derived Lead I but substantially divergent training trajectories, revealing cross-modality generalization gaps not detected by endpoint metrics alone.