Artificial intelligence applied to cardiovascular imaging, a critical focus on echocardiography: The point‐of‐view from “the other side of the coin”

Abstract

Cardiovascular imaging has achieved a crucial role in the management of cardiovascular diseases. In this field, echocardiography advantages include wide availability, portability, and affordability, at a relatively low cost. However, echocardiographic assessment requires highly trained operators, and implies high observer variability, as compared with the other cardiac imaging modalities. Hence, artificial intelligence might be extremely helpful. From the point‐of‐view of the peripheral “Spoke” Hospital potential user (“the other side of the coin”), artificial intelligence development appears very slow in the clinical arena. Many limitations are still present, and require full involvement, cooperation, and coordination of professional operators into Hub‐and‐Spoke network.

Artificial Intelligence in Clinical Echocardiography: Many Expectations, but Deep Uncertainties for Defining Strategies to Overcome Difficulties and Obstacles

“…We read with great interest the recently published editorial ''Future Guidelines for Artificial Intelligence in Echocardiography'' by Tseng et al 1 We found many interesting concepts, mostly addressed in our recently published report 2 as well. As we read the editorial, we wondered about the practicalities and challenges of implementing many aspects of artificial intelligence (AI) in an echocardiography laboratory.…”

“…Finally, funding these efforts is another important concern, especially for smaller peripheral institutions. 2 These institutions provide an important source of heterogeneous clinical cases and thus provide extremely important data sets to ensure that the AI algorithms that are developed are widely generalizable. 2 Cooperation and integration between central hub hospitals and smaller peripheral spoke institutions seem to be some of the key points 2 in obtaining a more accepted, standardized, flexible, and generalizable trustworthy AI product.…”

“…2 These institutions provide an important source of heterogeneous clinical cases and thus provide extremely important data sets to ensure that the AI algorithms that are developed are widely generalizable. 2 Cooperation and integration between central hub hospitals and smaller peripheral spoke institutions seem to be some of the key points 2 in obtaining a more accepted, standardized, flexible, and generalizable trustworthy AI product. 3,4 With regard to this goal, planning to concretely incorporate AI into routine clinical practice currently still seems a remote option, but even just sharing ideas, difficulties, and projects is an important beginning.…”

The Role of Artificial Intelligence in Echocardiography

“…AI “black box” models are created directly from raw data by algorithms, meaning that humans, even those who design them, cannot understand how variables are being combined to make predictions. This fact implies that results of AI models are sometimes impossible to interpret and verify from a clinical point of view [ 61 ].…”

“…Lastly, clinical applications of AI in echocardiography also face difficulties from a technical point of view. In addition to the vendor-dependent setup differences, AI clinical applications can also be affected by the frequent inability to obtain optimal image quality or accurate views [ 61 ]. In those cases, the accuracy of the models could be affected, or nonstructural/suboptimal echocardiographic data would need careful preprocessing by operators.…”