A Two-Step Approach to Overcoming Data Imbalance in the Development of an Electrocardiography Data Quality Assessment Algorithm: A Real-World Data Challenge.

Continuously acquired biosignals from patient monitors contain significant amounts of unusable data. During the development of a decision support system based on continuously acquired biosignals, we developed machine and deep learning algorithms to automatically classify the quality of ECG data. A total of 31,127 twenty-s ECG segments of 250 Hz were used as the training/validation dataset. Data quality was categorized into three classes: acceptable, unacceptable, and uncertain. In the training/validation dataset, 29,606 segments (95%) were in the acceptable class. Two one-step, three-class approaches and two two-step binary sequential approaches were developed using random forest (RF) and two-dimensional convolutional neural network (2D CNN) classifiers. Four approaches were tested on 9779 test samples from another hospital. On the test dataset, the two-step 2D CNN approach showed the best overall accuracy (0.85), and the one-step, three-class 2D CNN approach showed the worst overall accuracy (0.54). The most important parameter, precision in the acceptable class, was greater than 0.9 for all approaches, but recall in the acceptable class was better for the two-step approaches: one-step (0.77) vs. two-step RF (0.89) and one-step (0.51) vs. two-step 2D CNN (0.94) (p < 0.001 for both comparisons). For the ECG quality classification, where substantial data imbalance exists, the 2-step approaches showed more robust performance than the one-step approach. This algorithm can be used as a preprocessing step in artificial intelligence research using continuously acquired biosignals.

High Circulating Sphingosine 1-Phosphate is a Risk Factor for Osteoporotic Fracture Independent of Fracture Risk Assessment Tool.

Circulating sphingosine 1-phosphate (S1P) levels may be a biomarker for osteoporotic fracture (OF). This study assessed whether the addition of S1P levels to the fracture risk assessment tool (FRAX) could improve predictability of OF risk. Plasma S1P concentrations and FRAX variables were measured in 81 subjects with and 341 subjects without OF. S1P levels were higher in subjects with than those without OF (3.11 ± 0.13 µmol/L vs. 2.65 ± 0.61 µmol/L, P = 0.001). Higher S1P levels were associated with a higher likelihood of OF (odds ratio [OR] = 1.33, 95% confidence interval [CI] = 1.05-1.68), even after adjusting for FRAX probabilities. Compared with the lowest S1P tertile, subjects in the middle (OR = 3.37, 95% CI = 1.58-7.22) and highest (OR = 3.65, 95% CI = 1.66-8.03) S1P tertiles had higher rates of OF after adjustment. The addition of S1P levels to FRAX probabilities improved the area under the receiver-operating characteristics curve (AUC) for OF, from 0.708 to 0.769 (P = 0.013), as well as enhancing category-free net reclassification improvement (NRI = 0.504, 95% CI = 0.271-0.737, P < 0.001) and integrated discrimination improvement (IDI = 0.044, 95% CI = 0.022-0.065, P < 0.001). Adding S1P levels to FRAX probabilities especially in 222 subjects with osteopenia having a FRAX probability of 3.66-20.0% markedly improved the AUC for OF from 0.630 to 0.741 (P = 0.012), as well as significantly enhancing category-free NRI (0.571, 95% CI = 0.221-0.922, P = 0.001) and IDI (0.060, 95% CI = 0.023-0.097, P = 0.002). S1P is a consistent and significant risk factor of OF independent of FRAX, especially in subjects with osteopenia and low FRAX probability.