Automated detection scheme for acute myocardial infarction using convolutional neural network and long short-term memory.

The early detection of acute myocardial infarction, which is caused by lifestyle-related risk factors, is essential because it can lead to chronic heart failure or sudden death. Echocardiography, among the most common methods used to detect acute myocardial infarction, is a noninvasive modality for the early diagnosis and assessment of abnormal wall motion. However, depending on disease range and severity, abnormal wall motion may be difficult to distinguish from normal myocardium. As abnormal wall motion can lead to fatal complications, high accuracy is required in its detection over time on echocardiography. This study aimed to develop an automatic detection method for acute myocardial infarction using convolutional neural networks (CNNs) and long short-term memory (LSTM) in echocardiography. The short-axis view (papillary muscle level) of one cardiac cycle and left ventricular long-axis view were input into VGG16, a CNN model, for feature extraction. Thereafter, LSTM was used to classify the cases as normal myocardium or acute myocardial infarction. The overall classification accuracy reached 85.1% for the left ventricular long-axis view and 83.2% for the short-axis view (papillary muscle level). These results suggest the usefulness of the proposed method for the detection of myocardial infarction using echocardiography.

A Better Predictor of Acute Kidney Injury After Cardiac Surgery: The Largest Area Under the Curve Below the Oxygen Delivery Threshold During Cardiopulmonary Bypass.

Background The aim of this study was to compare the predictive accuracy of acute kidney injury (AKI) after cardiac surgery using cardiopulmonary bypass for the largest area under the curve (AUC) below the oxygen delivery (DO2) threshold and the cumulative AUC below the DO2 threshold. Methods and Results From March 2017 to October 2019, 202 patients who had undergone cardiac surgery with cardiopulmonary bypass were enrolled. The perfusion parameters were recorded every 20 seconds, and the DO2 (10×pump flow index [L/min per m2]×[hemoglobin (g/dL)×1.36×arterial oxygen saturation (%)+partial pressure of arterial oxygen (mm Hg)×0.003]) threshold of 300 mL/min per m2 was considered to define sufficient DO2. The nadir DO2, the cumulative AUC below the [Formula: see text], and the largest AUC below the [Formula: see text] were used to predict the incidence of AKI. Postoperative AKI was observed in 12.4% of patients (25/202). By multivariable analysis, the largest AUC below the [Formula: see text] ≥880 (odds ratio [OR], 4.9; 95% CI, 1.2-21.5 [P=0.022]), preoperative hemoglobin concentration ≤11.6 g/dL (OR, 7.6; 95% CI, 2.0-32.3 [P=0.004]), and red blood cell transfusions during cardiopulmonary bypass ≥2 U (OR, 3.3; 95% CI, 1.0-11.1 [P=0.041]) were detected as independent risk factors for AKI. Receiver operating curve analysis revealed that the largest AUC below the [Formula: see text] was more accurate to predict postoperative AKI compared with the nadir DO2 and the cumulative AUC below the [Formula: see text] (differences between areas, 0.0691 [P=0.006] and 0.0395 [P=0.001]). Conclusions These data suggest that a high AUC below the [Formula: see text] is an important independent risk factor for AKI after cardiopulmonary bypass, which could be considered for risk prediction models of AKI.

Correction to: Cell salvage processing of residual cardiopulmonary bypass volume in minimally invasive cardiac surgery.

In the article "Cell salvage processing of residual cardiopulmonary bypass volume in minimally invasive cardiac surgery" published in Heart and Vessels, there were several errors in numerical values.

Cell salvage processing of residual cardiopulmonary bypass volume in minimally invasive cardiac surgery.

Several reports demonstrated positive effects of processing residual cardiopulmonary bypass volume using a cell salvage device in conventional open heart surgery via sternotomy on hemostasis. The present study aimed to investigate whether cell salvage processing has the same effects on postoperative blood loss and transfusion in minimally invasive cardiac surgery. Between July 2015 and April 2018, 80 consecutive patients undergoing minimally invasive aortic valve replacement via right anterolateral minithoracotomy were enrolled in the present study. Perioperative outcomes and coagulation data of 40 patients who were retransfused with processed cardiopulmonary bypass volumes were compared with those of 40 patients receiving unprocessed residual blood (control group). Postoperative blood loss in patients receiving processed residual blood was significantly less than that in the control group at 6 h (115 ± 50 vs. 73 ± 33 ml, p < 0.001) and 12 h (167 ± 70 vs. 125 ± 67 ml, p = 0.009) after surgery, and the rate of fresh frozen plasma use after surgery was significantly reduced in patients receiving processed residual blood (18 vs. 0%, p = 0.012). In conclusion, processing of residual cardiopulmonary bypass volume reduced postoperative blood loss and postoperative use of fresh frozen plasma and could be useful for hemostasis in minimally invasive cardiac surgery.

Novel method for estimating the total blood volume: the importance of adjustment using the ideal body weight and age for the accurate prediction of haemodilution during cardiopulmonary bypass.

OBJECTIVES: Although total blood volume (TBV) is central to the estimation of the haemodilution rate during cardiopulmonary bypass (CPB), conventional formulas lack sufficient accuracy. The aim of this study was to establish a new formula using ideal body weight (BW) with adjustment for gender or age to estimate TBV for a more accurate prediction of the haemodilution rate during CPB. METHODS: A total of 214 consecutive patients who underwent cardiac surgery with CPB were included in this study. TBV was retrospectively estimated using the following formulae: (1) Conventional TBV = actual BW × fixed 70 ml/kg, (2) gender-based modified TBV = ideal BW × 75 ml/kg (male) or 65 ml/kg (female) and (3) age-based modified TBV = ideal BW × 70 ml/kg (<65 years old) or 60 ml/kg (≥65 years old). The relationship between actual and predicted haemodilution rates calculated by these formulas was examined. RESULTS: The actual haemodilution rate based on the haematocrit value was 24.4 ± 4.4%. There was no significant correlation between the actual and predicted haemodilution rates obtained by the conventional formula, whereas both modified formulae with the ideal BW showed a significant correlation. Furthermore, the age-based modified formula showed the highest correlation level (r = 0.45, P < 0.001) as well as a strong correlation between the actual and predicted postdilution haematocrit values (y = 0.903x + 3.385, R2 = 0.892). CONCLUSIONS: The conventional formula is unable to predict the actual haemodilution rate accurately. Our new formula with a combination of the ideal BW and adjustment for age was shown to be useful for the accurate prediction of the haemodilution rate during CPB.