METHODS: A dataset containing medical records of 809 patients suspected to suffer from ACS was used. For each subject, 266 clinical factors were collected. At first, a feature selection was performed based on interviews with 20 cardiologists; thereby 40 seminal features for classifying ACS were selected. Next, a feature selection algorithm was also applied to detect a subset of the features with the best classification accuracy. As a result, the feature numbers considerably reduced to only seven. Lastly, based on the seven selected features, eight various common pattern recognition tools for classification of ACS were used.

RESULTS: The performance of the aforementioned classifiers was compared based on their accuracy computed from their confusion matrices. Among these methods, the multi-layer perceptron showed the best performance with the 83.2% accuracy.

METHODS: After 10 min of supine rest, the subject was tilted at a 70-degree angle on a tilt table for approximately a total of 35 min. 400 µg of glyceryl trinitrate (GTN) was administered sublingually after the first 20 min and monitoring continued for another 15 min. Mean imputation and K-nearest neighbors (KNN) imputation approaches to handle missing values. Next, feature selection techniques were implemented, including genetic algorithm, recursive feature elimination, and feature importance, to determine the crucial features. The Mann-Whitney U test was then performed to determine the statistical difference between two groups. Patients with VVS are categorized via machine learning models including Support Vector Machine (SVM), Gaussian Naïve Bayes (GNB), Multinomial Naïve Bayes (MNB), KNN, Logistic Regression (LR), and Random Forest (RF). The developed model is interpreted using an explainable artificial intelligence (XAI) model known as partial dependence plot.

RESULTS: A total of 137 subjects aged between 9 and 93 years were recruited for this study, 54 experienced clinical symptoms were considered positive tests, while the remaining 83 tested negative. Optimal results were obtained by combining the KNN imputation technique and three tilting features with SVM with 90.5% accuracy, 87.0% sensitivity, 92.7% specificity, 88.6% precision, 87.8% F1 score, and 95.4% ROC (receiver operating characteristics) AUC (area under curve).

DESIGN: Serum intact parathormone (PTH) concentrations were measured on samples taken before and during a variable-rate tri-sodium citrate infusion designed to 'clamp' the whole blood ionised calcium concentration 0.20 mmol L-1 below baseline for 120 min.

SUBJECTS: Six Malaysian patients aged 17-42 years with acute malaria, four of whom were restudied in convalescence, and 12 healthy controls aged 19-36 years.

MAIN OUTCOME MEASURES: Whole-blood ionised calcium and serum intact PTH concentrations.

RESULTS: The mean (SD baseline ionised calcium was lower in the malaria patients than in controls (1.09 +/- 0.06 vs. 1.18 +/- 0.03 mmol L-1, respectively; P = 0.01) but PTH concentrations were similar (3.0 +/- 1.8 vs. 3.3 +/- 1.3 pmol L(-1); P = 0.33). Target whole-blood ionised calcium concentrations were achieved more rapidly in the controls than the patients (within 15 vs. 30 min) despite significantly more citrate being required in the patients (area under the citrate infusion-time curve 0.95 (0.25 vs. 0.57 +/- 0.09 mmol kg-1; P < 0.01). The ratio of the change in serum PTH to that in ionised calcium (delta PTH/ delta Ca2+), calculated to adjust for differences in initial rate of fall of ionised calcium, was similar during the first 5 min of the clamp (132 +/- 75 x 10(-6) vs. 131 +/- 43 x 10(-6) in patients and controls, respectively, P > 0.05), as were steady-state serum PTH levels during the second hour (7.0 +/- 2.2 pmol L-1 in each case). Convalescent patients had normal basal ionised calcium levels but the lowest serum intact PTH levels before and during the clamp, consistent with an increase in skeletal PTH sensitivity after treatment.

OBJECTIVE: The purpose of this study was to identify predictors of A4 amplitude and high AVS.

METHODS: We analyzed 64 patients enrolled in MARVEL 2 who had visible P waves on electrocardiogram for assessing A4 amplitude and 40 patients with third-degree AV block for assessing AVS at rest. High AVS was defined as >90% correct atrial-triggered ventricular pacing. The association between clinical factors and echocardiographic parameters with A4 amplitude was investigated using a multivariable model with lasso variable selection. Variables associated with A4 amplitude together with premature ventricular contraction burden, sinus rate, and sinus rate variability (standard deviation of successive differences of P-P intervals [SDSD]) were assessed for association with AVS.

RESULTS: In univariate analysis, low A4 amplitude was inversely related to atrial function assessed by E/A ratio and e'/a' ratio, and was directly related to atrial contraction excursion (ACE) and atrial strain (Ɛa) on echocardiography (all P ≤.05). The multivariable lasso regression model found coronary artery bypass graft history, E/A ratio, ACE, and Ɛa were associated with low A4 amplitude. E/A ratio and SDSD were multivariable predictors of high AVS, with >90% probability if E/A <0.94 and SDSD <5 bpm.