Abstract
Left ventricular ejection fraction (LVEF) is the most important parameter in the assessment of cardiac function. A machine-learning algorithm was trained to guide ultrasound-novices to acquire diagnostic echocardiography images. The artificial intelligence (AI) algorithm then estimates LVEF from the captured apical-4-chamber (AP4), apical-2-chamber (AP2), and parasternal-long-axis (PLAX) loops. We sought to test this algorithm by having first-year medical students without previous ultrasound knowledge scan real patients. Nineteen echo-naïve first-year medical students were trained in the basics of echocardiography by a 2.5 h online video tutorial. Each student then scanned three patients with the help of the AI. Image quality was graded according to the American College of Emergency Physicians scale. If rated as diagnostic quality, the AI calculated LVEF from the acquired loops (monoplane and also a “best-LVEF” considering all views acquired in the particular patient). These LVEF calculations were compared to images of the same patients captured and read by three experts (ground-truth LVEF [GT-EF]). The novices acquired diagnostic-quality images in 33/57 (58%), 49/57 (86%), and 39/57 (68%) patients in the PLAX, AP4, and AP2, respectively. At least one of the three views was obtained in 91% of the attempts. We found an excellent agreement between the machine’s LVEF calculations from images acquired by the novices with the GT-EF (bias of 3.5% ± 5.6 and r = 0.92, p < 0.001 in the “best-LVEF” algorithm). This pilot study shows first evidence that a machine-learning algorithm can guide ultrasound-novices to acquire diagnostic echo loops and provide an automated LVEF calculation that is in agreement with a human expert.
Highlights
Left ventricular ejection fraction (LVEF) is by far the most important parameter in the assessment of cardiac function
The design of this study consisted of four steps: (1) nineteen echo-naïve first-year medical students were briefly trained in the basics of echocardiography, (2) the trained students scanned three patients each with the help of the machinelearning algorithm (Caption Health Inc., Brisbane, CA, USA), for each patient they attempted to acquire three views, (3) the same patients were scanned by one expert cardiologist (MS), the acquired loops were read by three blinded expert cardiologists (MS, TB, PB), the average calculated LVEF was taken as the ground-truth LVEF (GT-EF), (4) the 171 loops which were scanned by the students were evaluated by the machine-learning algorithm
All patients were scanned with a GE S70 (General Electric Healthcare, Chicago, IL, USA) machine to acquire reference images (PLAX, AP4, AP2). These images were read by three expert cardiologists (MS, TB, PB) who were blinded to the results provided by the artificial intelligence
Summary
Left ventricular ejection fraction (LVEF) is by far the most important parameter in the assessment of cardiac function. Echocardiographic calculations of LVEF play an important role in diagnosis and clinical decision-making This is especially true in cardiac device therapy, management of valvular heart disease, oncology patients, and in the diagnosis and treatment of heart failure [1,2,3]. A new software was developed, implementing a fully machine-learning algorithm mimicking the human eye by estimating LVEF from the degree of ventricular expansion and contraction, myocardial thickening, and motion of the mitral annular plane (Fig. 1) [6]. This algorithm has been developed further, integrating the parasternal long axis view into global LVEF assessment
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
