How AI in cardiovascular ultrasound can reduce staff burnout

Burnout, characterized by “feelings of exhaustion…and a reduction in the employee’s productivity” [1], is a side effect of many healthcare occupations. A review of articles about occupational burnout syndrome among diagnostic imaging professionals identified increasing workloads and increasingly demanding work hours as major contributors to work stress [2]. Burnout may be implicated in missed pathology while scanning, misinterpretation of results and reduced patient care [3].

Astute healthcare organizations recognize that reducing burnout can help them meet clinical, workflow and employee satisfaction goals. As these organizations seek ways to decrease the stress levels and associated burnout in their cardiovascular imaging departments, they are finding AI to be a valuable ally.

Philips has 26 cardiovascular ultrasound patents that leverage AI to assist with a variety of tasks. Every one of these technologies reflect Philips commitment to responsible development of AI tools. Philips core AI principles are:

User-centered: AI must complement and augment user skills so clinicians can perform more, high-quality exams.

Data excellence: High-integrity, high-quality training data sets fuel advanced pattern recognition, machine and deep learning algorithms.

Empowered clinicians: Clinicians have ultimate decision responsibility and can accept or modify results based on their clinical assessment.

The specific functions of Philips cardiovascular ultrasound AI tools vary, but all address burnout in either one or both of two ways: alleviating cognitive burden [4] and easing workflow.

The sources of information available to clinicians continue to expand at a dizzying speed, while development of effective tools to catalog and efficiently access those sources has lagged. Cardiologists and sonographers are adept at making up for this lack of integration, but it comes at the cost of a great cognitive load.

Our cardiovascular ultrasound AI solutions alleviate cognitive burden with decision support. These AI-powered tools can efficiently integrate data from multiple sources or time periods, giving healthcare providers a longitudinal view of patient health data that informs diagnosis and helps assess treatment. In addition, condition and anatomy-specific applications provide data needed for treatment decisions. For example, 3D Auto Tricuspid Valve Quantification (3D Auto TV) uses AI to help evaluate annulus size to guide device selection.

Other examples of using AI to decrease cognitive burden include:

These AI applications build clinical confidence while also helping clinicians make the best use of their time.

Doing more with less has become a common directive in diagnostic imaging, as the demand for imaging studies swells and the shortage of cardiovascular ultrasound technologists and cardiologists grows. Using AI tools to ease workflow can help cardiovascular imaging departments meet demand while also decreasing the potential for staff burnout.

Philips AI-powered tools are proven to enhance workflow, saving time and helping fast-paced cardiovascular ultrasound departments keep up with demand. When using AI features for end-to-end experiences, users saved 82% total time for acquisition and analysis combined [5]. AI also enabled a 20% reduction in 2D diagnostic echo exam time [6] and a 51% reduction in quantification time [7].

Automated image formation, acquisition, measurement and reporting shorten the time to perform and interpret exams while also limiting time-consuming, manual and repetitive tasks. For example, Philips AutoMeasure provides fully automated 2D length and EF and Doppler measurements. Philips 3D Auto RV offers full 3D quantification for RV volumes and functional assessment.

Additional examples of how AI tools ease workflow include:

Automatically detecting borders and calculating a score based on left ventricular (LV) segments to aid assessment of LV impairment with Segmental Wall Motion Scoring

Reducing clicks by 89% [8] using the Mitral Valve Navigator to take a Live 3D volume of the mitral valve and turn it into an easy-to-interpret model in eight guided steps, providing access to a comprehensive list of MV measurements and calculations

Automating border detection, multi-beat selection and results average using Dynamic HeartModel to yield a more reliable heart function evaluation than single beat in arrhythmia patients

While AI will continue to change how healthcare organizations operate, one thing remains constant: your employees are your most valuable resource. Tools that reduce burnout help you retain experienced professionals, recruit new sonographers and enhance job satisfaction.