First-in-human results of an autonomous bed-integrated robotic transthoracic echocardiography system to support heart failure care
L Kieviet, M Guglielmo, A R Van Rosendael, F Donato, R Bertozzi, S Bonhof, R K Riezebos, M I F J Oerlemans, N T B Scholte, P Van Der HarstAbstract
Background
Access to timely, high-quality transthoracic echocardiography (TTE) is fundamental across cardiovascular care, yet service capacity is constrained by shortages of experienced sonographers and the physical burden of manual probe handling. These limitations are particularly impactful in heart failure (HF) care pathways, as TTE is central to evaluating ventricular function and valvular disease. A robotic and artificial intelligence (AI)-enabled ultrasound system may help standardize image acquisition and reduce operator dependency, potentially improving scalability and access to TTE. We evaluated a bed-integrated robotic platform (BBE) designed to perform largely autonomous TTE.
Purpose
Here, we present the first preliminary results of the AI enhanced semi-autonomous robotic TTE system, focusing on safety and examination characteristics, and its ability to obtain echocardiographic studies in routine practice.
Methods
In this single-centre, prospective study, adult patients referred for clinically indicated TTE were included. Each patient underwent a paired TTE protocol on the BBE system, consisting of a standard manually acquired examination and an autonomous examination. The bed-mounted robotic arm, positioned beneath a thoracic window, uses a standard phased-array TTE probe with multi-axis force/torque sensing. A predefined acquisition protocol of 41 clips covering parasternal, subcostal, and apical views was used, with no measurements or reporting performed. The endpoints of this first evaluation focus on device performance, including the protocol completion rate, examination duration, maximal applied force, and participant usability assessments.
Results
We analysed the first 25 consecutive patients (68% male, mean age 59 ± 15 years, mean body mass index (BMI) 25.2 ± 4.2 kg/m2) examined using the most recent BBE software (version 4.2.0). In the autonomous examinations, nearly all required clips were obtained (99.8%), with a mean duration of 27.4 ± 2.4 minutes. Scan time did not differ materially across BMI categories (<25, 25-30, and > 30kg/m2) and sex. Mean maximal applied force was 13.8 ± 2.5 N, remaining within safety limits in all examinations, without relevant differences between imaging windows. Maximal forces increased modestly with higher BMI. No serious device-related adverse events occurred. Patient reported no relevant discomfort.
Conclusion
First-in-human experience shows that an autonomous, bed-integrated robotic system can safely acquire standard TTE views. These findings support further evaluation of this technology in HF pathways to expand imaging capacity, reduce sonographer workload and enable more scalable HF diagnostics and follow-up.For image description, please refer to the figure legend and surrounding text.For image description, please refer to the figure legend and surrounding text.