Validation of Electrocardiogram Based Photoplethysmogram Generated Using U-Net Based Generative Adversarial Networks.

Photoplethysmogram (PPG) performs an important role in alarming atrial fibrillation (AF). While the importance of PPG is emphasized, there is insufficient amount of openly available atrial fibrillation PPG data. We propose a U-net-based generative adversarial network (GAN) which synthesize PPG from paired electrocardiogram (ECG). To measure the performance of the proposed GAN, we compared the generated PPG to reference PPG in terms of morphology similarity and also examined its influence on AF detection classifier performance. First, morphology was compared using two different metrics against the reference signal: percent root mean square difference (PRD) and Pearson correlation coefficient. The mean PRD and Pearson correlation coefficient were 27% and 0.94, respectively. Heart rate variability (HRV) of the reference AF ECG and the generated PPG were compared as well. The p-value of the paired t-test was 0.248, indicating that no significant difference was observed between the two HRV values. Second, to validate the generated AF PPG dataset, four different datasets were prepared combining the generated PPG and real AF PPG. Each dataset was used to optimize a classification model while maintaining the same architecture. A test dataset was prepared to test the performance of each optimized model. Subsequently, these datasets were used to test the hypothesis whether the generated data benefits the training of an AF classifier. Comparing the performance metrics of each optimized model, the training dataset consisting of generated and real AF PPG showed a test accuracy result of 0.962, which was close to that of the dataset consisting only of real AF PPG data at 0.961. Furthermore, both models yielded the same F1 score of 0.969. Lastly, using only the generated AF PPG dataset resulted in test accuracy of 0.945, indicating that the trained model was capable of generating valuable AF PPG. Therefore, it can be concluded that the generated AF PPG can be used to augment insufficient data. To summarize, this study proposes a GAN-based method to generate atrial fibrillation PPG that can be used for training atrial fibrillation PPG classification models.

Suture tie-down forces and cyclic contractile forces after undersized tricuspid annuloplasty using a Tri-Ad Adams tricuspid annuloplasty ring in an ovine model.

OBJECTIVES: This study evaluated suture tie-down forces and cyclic contractile forces (CCFs) after undersized tricuspid annuloplasty using a hybrid band. METHODS: Downsized tricuspid annuloplasty was planned in adult male sheep using 8 force transducers attached from the septal to the anterior annular areas of the ring (segments 1 and 2, flexible septal; segments 3 and 4, semi-rigid posterior; segments 5 and 6, semi-rigid anterior; segments 7 and 8, flexible anterior). CCFs were analysed at 3 different levels of peak right ventricular pressure (RVP): 30, 50 and 70 mmHg. RESULTS: Eight 5-year-old male Corriedale sheep (average body weight = 66.8 kg) were used. The average suture tie-down force was 4.42 [standard deviation (SD): 2.32] N. When the forces were compared, it was lowest in the flexible anterior area and highest in the flexible septal area (P < 0.001). With the RVP of 30 mmHg, the average CCFs was lowest at segment 3 [0.07 (SD: 0.07) N] and highest at segment 7 [0.15 (SD: 0.08) N]. The CCFs were 0.12 (SD: 0.1) N, 0.09 (SD: 0.12) N, 0.14 (SD: 0.1) N and 0.13 (SD: 0.09) N in the flexible septal, semi-rigid posterior, semi-rigid anterior and flexible anterior parts, respectively (P = 0.208). As the peak RVP increased to 50 and 70 mmHg, the CCFs of each area increased significantly (P < 0.001). Despite this increase, the CCFs remained low (0.1 and 0.3 N), and differences in CCFs between segments and between annular areas showed similar patterns. CONCLUSIONS: The flexible end of the hybrid band reduces the CCFs and might prevent annular tears after ring tricuspid annuloplasty, and the risk of tear would be low even in the septal area.

Ultrasonic blood flowmeter with a novel Xero algorithm for a mechanical circulatory support system.

Mechanical circulatory support systems (MCSSs) are crucial devices for transplants in patients with heart failure. The blood flowing through the MCSS can be recirculated or even stagnated in the event of critical blood flow issues. To avoid emergencies due to abnormal changes in the flow, continuous changes of the flowrate should be measured with high accuracy and robustness. For better flowrate measurements, a more advanced ultrasonic blood flowmeter (UFM), which is a noninvasive measurement tool, is needed. In this paper, we propose a novel UFM sensor module using a novel algorithm (Xero) that can exploit the advantages of both conventional cross-correlation (Xcorr) and zero-crossing (Zero) algorithms, using only the zero-crossing-based algorithm. To ensure the capability of our own developed and optimized ultrasonic sensor module for MCSSs, the accuracy, robustness, and continuous monitoring performance of the proposed algorithm were compared to those of conventional algorithms after application to the developed sensor module. The results show that Xero is superior to other algorithms for flowrate measurements under different environments and offers an error rate of at least 0.92%, higher robustness for changing fluid temperatures than conventional algorithms, and sensitive responses to sudden changes in flowrates. Thus, the proposed UFM system with Xero has a great potential for flowrate measurements in MCSSs.

Suture tie-down forces and cyclic contractile forces after an undersized tricuspid annuloplasty using a 3-dimensional rigid ring in an ovine model.

OBJECTIVES: This study was conducted to measure suture tie-down forces and evaluate cyclic contractile forces (CCFs) in beating hearts after undersized 3-dimensional (3D) rigid-ring tricuspid valve annuloplasty (TAP). METHODS: Eight force transducers were attached to the 3D rigid TAP ring. Segments 1 to 8 were attached from the mid-septal to anterior-septal commissural area in a counterclockwise order. Two-sizes-down ring TAPs were performed in 6 sheep. Tie-down forces and CCF were recorded and analysed at the 8 annular segments and at 3 levels of peak right ventricular pressure (RVP: 30, 50 and 70 mmHg). RESULTS: The overall average tie-down forces and CCF were 4.34 ± 2.26 newtons (N) and 0.23 ± 0.09 N, respectively. The CCF at an RVP of 30 mmHg were higher at 3 commissural areas (segments 3, 5 and 8) than at the other segments. The increases in the CCF following changes in the RVP were statistically significant only at the 3 commissural areas (P = 0.012). However, mean CCFs remained low at all annular positions (ranges of average CCF = 0.06-0.46 N). CONCLUSIONS: The risk of suture dehiscence after down-sized 3D rigid-ring TAP might be minimal because the absolute forces remained low in all annular positions even in the condition of high RVP. However, careful suturing in the septal annular area and commissures is necessary to prevent an annular tear during a down-sized 3D rigid-ring TAP.

Point-of-care testing of plasma free hemoglobin and hematocrit for mechanical circulatory support.

Hematological analysis is essential for patients who are supported by a mechanical circulatory support (MCS). The laboratory methods used to analyze blood components are conventional and accurate, but they require a mandatory turn-around-time for laboratory results, and because of toxic substances, can also be hazardous to analysis workers. Here, a simple and rapid point-of-care device is developed for the measurement of plasma free hemoglobin (PFHb) and hematocrit (Hct), based on colorimetry. The device consists of camera module, minimized centrifuge system, and the custom software that includes the motor control algorithm for the centrifuge system, and the image processing algorithm for measuring the color components of blood from the images. We show that our device measured PFHb with a detection limit of 0.75 mg/dL in the range of (0-100) mg/dL, and Hct with a detection limit of 2.14% in the range of (20-50)%. Our device had a high correlation with the measurement method generally used in clinical laboratories (PFHb R = 0.999, Hct R = 0.739), and the quantitative analysis resulted in precision of 1.44 mg/dL for PFHb value of 14.5 mg/dL, 1.36 mg/dL for PFHb value of 53 mg/dL, and 1.24% for Hct 30%. Also, the device can be measured without any pre-processing when compared to the clinical laboratory method, so results can be obtained within 5 min (about an 1 h for the clinical laboratory method). Therefore, we conclude that the device can be used for point-of-care measurement of PFHb and Hct for MCS.