Workflow Enhancement in Echo-Labs Using AI-Based Technology in Patient with Atrial Fibrillation

 

Liat Alper-Suissa, David Rosenmann MD, David Marmor MD, Shemy Carasso MD

Shaare Zedek Medical Center, Jerusalem, ISRAEL

 

Introduction

 

Global LV function evaluation is performed in every echo examination, however, quantitative measurements of ejection fraction (EF) and global longitudinal strain (GLS) are performed in selected patients due to time constraints, availability of analysis software tools on the ultrasound devices, and team qualifications. GLS evaluation is mainly recommended for patients receiving chemotherapy to monitor cardiotoxicity (Roberto M. Lang, 2015). According to published literature, GLS evaluation may be beneficial for many other clinical indications and should be done routinely for most echo patients (Nina Ajmone Marsan, 2020). A novel Artificial Intelligence (AI) based vendor-neutral system that automatically performs EF and GLS evaluation to all cardiac ultrasound exams, without user involvement, has recently become available. The system runs on the hospital’s server, processes the echo examination, selects the optimal apical views, generates EF and GLS measurements, and sends the results as secondary capture to the PACS viewer. The system was installed in an echo lab environment. The goal of the project was to evaluate the potential clinical value in echo lab workflow and patient care in the echo lab by providing automated quantitative measurements of EF and GLS to all echo examinations.

 

Methods

 

The AI-based system was implemented in an echo lab and integrated into the routine workflow with no change in the way echo images were scanned or saved. The automated EF was obtained from apical 2 and 4 chamber views and GLS was obtained from apical 2, 3, and 4 chamber views. EF and GLS results become available within minutes and are presented as part of a standard examination review on the PACS. One of the cases reviewed involved an 82-year-old woman with atrial fibrillation who had undergone Transthoracic Echocardiography (TTE) and Transesophageal Echocardiogram (TEE) to evaluate the Left Ventricular (LV) function and to exclude Left Atrial Appendage (LAA) thrombus before cardioversion. The patient complained of difficulty during exercise. The exact time the patient run into atrial fibrillation is unknown. The patient started Apixaban 2.5mg x 2 a day. The Echo examination was sent to the AI-based automated system, the analysis results were sent to the PACS.

 

Results

The results obtained by the AI-based automated system (EF=28%, GLS=-9%), were very low compared to the EF that was initially determined by visual estimation. LV thrombus was ruled out during TTE, therefore the automatically obtained results lead to a re-assessment of the LV contraction of this patient and to immediate subsequent cardioversion to allow improvement in LV systolic function and cardiac output.

 

 

 

Discussion

Incorporation of AI-based software into echocardiography allows easy and seamless quantitation of LV function (volumetric and mechanics), empowering clinicians' workflows and decision-making in the echo lab.

​

​
 

​

 

 

Figure 1

AI-Based EF and GLS Results

 

References

1. Lang, R. M. et al. Recommendations for Cardiac Chamber Quantification by Echocardiography in Adults: An Update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Journal of the American Society of Echocardiography (2015;28:1-39)

2. Nina Ajmone Marsan et al. EACVI survey on standardization of cardiac chambers quantification by transthoracic echocardiography. European Heart Journal-Cardiovascular Imaging (2020;21,119–123)