Echocardiographic Detection of Regional Wall Motion Abnormalities Using Artificial Intelligence Compared to Human Readers

BACKGROUND

Although regional wall motion abnormality (RWMA) detection is foundational to transthoracic echocardiography, current methods are prone to interobserver variability. We aimed to develop a deep learning (DL) model for RWMA assessment and compare it to expert and novice readers.

METHODS

We used 15,746 transthoracic echocardiography studies-including 25,529 apical videos-which were split into training, validation, and test datasets. A convolutional neural network was trained and validated using apical 2-, 3-, and 4-chamber videos to predict the presence of RWMA in 7 regions defined by coronary perfusion territories, using the ground truth derived from clinical transthoracic echocardiography reports. Within the test cohort, DL model accuracy was compared to 6 expert and 3 novice readers using F1 score evaluation, with the ground truth of RWMA defined by expert readers. Significance between the DL model and novices was assessed using the permutation test.

RESULTS

Within the test cohort, the DL model accurately identified any RWMA with an area under the curve of 0.96 (0.92-0.98). The mean F1 scores of the experts and the DL model were numerically similar for 6 of 7 regions: anterior (86 vs 84), anterolateral (80 vs 74), inferolateral (83 vs 87), inferoseptal (86 vs 86), apical (88 vs 87), inferior (79 vs 81), and any RWMA (90 vs 94), respectively, while in the anteroseptal region, the F1 score of the DL model was lower than the experts (75 vs 89). Using F1 scores, the DL model outperformed both novices 1 (P = .002) and 2 (P = .02) for the detection of any RWMA.

CONCLUSIONS

Deep learning provides accurate detection of RWMA, which was comparable to experts and outperformed a majority of novices. Deep learning may improve the efficiency of RWMA assessment and serve as a teaching tool for novices.

A review of standardized high-throughput cardiovascular phenotyping with a link to metabolism in mice

Cardiovascular diseases cause a high mortality rate worldwide and represent a major burden for health care systems. Experimental rodent models play a central role in cardiovascular disease research by effectively simulating human cardiovascular diseases. Using mice, the International Mouse Phenotyping Consortium (IMPC) aims to target each protein-coding gene and phenotype multiple organ systems in single-gene knockout models by a global network of mouse clinics. In this review, we summarize the current advances of the IMPC in cardiac research and describe in detail the diagnostic requirements of high-throughput electrocardiography and transthoracic echocardiography capable of detecting cardiac arrhythmias and cardiomyopathies in mice. Beyond that, we are linking metabolism to the heart and describing phenotypes that emerge in a set of known genes, when knocked out in mice, such as the leptin receptor (Lepr), leptin (Lep), and Bardet-Biedl syndrome 5 (Bbs5). Furthermore, we are presenting not yet associated loss-of-function genes affecting both, metabolism and the cardiovascular system, such as the RING finger protein 10 (Rfn10), F-box protein 38 (Fbxo38), and Dipeptidyl peptidase 8 (Dpp8). These extensive high-throughput data from IMPC mice provide a promising opportunity to explore genetics causing metabolic heart disease with an important translational approach.

Self-reported skills and self-confidence in point-of-care ultrasound: a cross-sectional nationwide survey amongst Finnish emergency physicians

BACKGROUND

The use of point-of-care ultrasound (POCUS) is increasing. Numerous investigators have evaluated the learning curves in POCUS, but there are no published studies on how emergency physicians perceive their own competence level with this skill.

METHODS

A nationwide survey amongst Finnish emergency physicians was conducted. The respondents reported their use of POCUS and how it has affected their clinical decision-making. The number of POCUS examinations performed was compared to the self-assessed skill level with different applications. Cut-off values were determined for the number of examinations required to acquire a good self-assessed skill level in each POCUS application. The correlation between self-confidence and the self-estimated skill level was analyzed. Several different statistical methods were used, such as Student's t-test, Pearson's correlation test, Loess method and ROC curve analysis.

RESULTS

A total of 134 out of 253 Finnish emergency medicine specialists and residents (52%) responded to the survey. The most commonly used POCUS applications were POCUS-assisted procedures, pleural effusion and pneumothorax, inferior vena cava and lower extremity deep venous thrombosis. The initial rate of perceived skill acquisition was very steep with the curve flattening with greater skill and more experience. The number of examinations performed to reach a self-assessed good competence varied from seven to 75 with different applications. The lowest cut-off point for self-assessed good competence was obtained for rapid ultrasound for the shock and hypotension-protocol and the highest for focused cardiac examinations. There was an excellent correlation between self-confidence and the self-assessed skill level.

CONCLUSIONS

The Finnish emergency practitioners' self-assessed development of POCUS skills parallels the previously published learning curves of POCUS. The correlation of self-confidence and the self-assessed skill level was found to be excellent. These findings add information on the development of perceived POCUS skills amongst emergency physicians and could complement a formal performance assessment.

Con: The Length of Adult Cardiothoracic Anesthesiology Fellowship Training Should Not Be Extended Beyond One Year

Fellowship training in adult cardiothoracic anesthesiology (ACTA) is a one-year postgraduate experience with formal accreditation by the Accreditation Council for Graduate Medical Education. ACTA is a competitive and evolving subspeciality. With expanding knowledge, clinical roles and technical skills required of the modern cardiothoracic anesthesiologists, the optimal structure and duration of the fellowship training are worth considering. This manuscript provides supporting rationale for fellowship training in ACTA to remain one year in duration. The expanding responsibilities of the cardiothoracic anesthesiologist and strategies to best train the future of the subspecialty within the current training structure are discussed. It also briefly examines the history and current status of the fellowship training, reviews considerations for increasing fellowship duration, and highlights personal and financial considerations during the training.