HERA W9, HERA W10 Diagnostic Ultrasound System

The HERA W9 and HERA W10 are versatile diagnostic ultrasound systems utilizing multi-modal imaging technologies for comprehensive anatomical and fluid analysis across numerous clinical applications. These systems incorporate deep learning-based artificial intelligence for features like HeartAssist, ViewAssist, and enhanced BiometryAssist, aiming to improve diagnostic precision and workflow efficiency through automated image interpretation and measurement support.

The HERA W9 and HERA W10 are general-purpose, mobile, software-controlled diagnostic ultrasound systems designed for acquiring and displaying ultrasound data for a wide range of clinical applications. These applications include fetal/obstetrics, abdominal, gynecological, pediatric, small organ, cephalic (neonatal and adult), trans-rectal, trans-vaginal, musculoskeletal, urological, cardiac (adult and pediatric), and peripheral vessel imaging, as well as fluid analysis. The systems operate using various imaging modes such as B-mode, M-mode, Doppler (Pulsed Wave, Continuous Wave, Color, Power, Tissue), 3D/4D imaging, and ElastoScan+™, and are intended for use by trained healthcare professionals to measure anatomical structures and provide analysis packages aiding in diagnosis. This iteration of the HERA W9/W10 systems introduces several key enhancements, notably the integration of artificial intelligence through new software functions like HeartAssist (HERA W10 only) and ViewAssist. HeartAssist utilizes deep learning to classify fetal heart ultrasound images and propose measurement views, while ViewAssist employs deep learning to classify views and suggest annotations across various applications, both aiming to streamline examinations. The BiometryAssist feature has also been upgraded to an AI version for improved automatic fetal biometric measurements. Additionally, the systems incorporate new transducers (LA2-14A, PA1-5A) and a new software feature, Slice A. The HERA W10 model offers exclusive features like HeartAssist, CrystalVue Flow, and MPI+, alongside some hardware distinctions from the HERA W9. Non-clinical testing confirmed the safety and performance of these new components and software functionalities, including image quality with new transducers and accuracy of AI-based features.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the performance (accuracy, time-saving) of the AI-driven HeartAssist and ViewAssist features compare to experienced sonographers performing the same tasks manually in diverse patient populations?; What specific deep learning architectures are employed in BiometryAssist, HeartAssist, and ViewAssist, and what were the key datasets used for their training and validation to ensure generalizability across different patient demographics and image qualities?; Beyond fetal measurements, what is the potential for expanding the AI-based ViewAssist and annotation capabilities to other complex anatomical regions or diagnostic procedures covered by the system's indications for use?; What are the quantitative improvements in measurement accuracy and inter-operator variability for the AI-enhanced BiometryAssist compared to the previous non-AI version and manual measurements?; How do the new transducers (LA2-14A, PA1-5A) specifically enhance image quality or enable new diagnostic capabilities within the supported clinical applications, beyond just being 'migrated' from another system?