SonoMaxx‑Manufactured Handheld Ultrasound Validated in Peer‑Reviewed Study for AI‑Driven Knee Cartilage Assessment
VANCOUVER, Canada – A new study published in Ultrasound in Medicine & Biology demonstrates that combining portable ultrasound with a lightweight artificial intelligence (AI) model can enable scalable, real‑time assessment of knee cartilage – a key step toward early detection of osteoarthritis. The research validated several ultrasound systems, including a Viatom handheld device that is OEM‑manufactured by SonoMaxx(a company of BMV MEDTECH), highlighting the clinical readiness of SonoMaxx’s engineering for AI‑powered musculoskeletal imaging.
Closing the gap between acquisition and interpretation
Early osteoarthritis detection requires two things to become widely accessible: the imaging itself, and the pipeline that converts images into reliable quantitative measurements. Handheld ultrasound solves the first part by bringing imaging to clinics, training rooms, and community settings. However, without AI that runs directly on the probe – processing scans in real time and remaining accurate across different devices and operators – the bottleneck simply shifts from image acquisition to interpretation.
The newly published paper introduces MonoUNet, a neural network that is 10× to 700× smaller than other lightweight segmentation models, while achieving Dice scores up to 94.8% for knee cartilage segmentation. This small footprint allows the AI to run directly on a handheld ultrasound device – no laptop, no cloud, no waiting – enabling real‑time cartilage assessment in small clinics, athletic training rooms, and field‑based research.
Testing across devices – including SonoMaxx‑powered Viatom
The study evaluated MonoUNet on multiple ultrasound platforms:
- Clarius Mobile Health HD3 L15 handheld probe
- GE HealthCare LOGIQ P9 (cart‑based) and LOGIQ e (laptop)
- A Viatom handheld device, manufactured by SonoMaxx
Across all devices, the AI demonstrated excellent agreement with manual cartilage outcomes:
- Intraclass correlation coefficient (ICC) = 0.96 for cartilage thickness
- ICC = 0.99 for echo intensity
These results confirm that the Viatom handheld – built on SonoMaxx’s OEM platform – delivers clinical‑grade image quality suitable for AI‑driven segmentation, matching or exceeding the performance of larger, cart‑based systems for this application.
Quoting SonoMaxx leadership
“This independent validation is a strong endorsement of our engineering approach. We designed the Viatom handheld to be a versatile, high‑performance platform, and seeing it perform on par with established systems in a rigorous AI segmentation study confirms that our OEM solutions are ready for advanced point‑of‑care applications.” – SonoMaxx spokesperson
About the study
Lead author Alvin Kimbowa and senior collaborators Ilker Hacihaliloglu, Arjun Parmar, and others developed MonoUNet to address the need for scalable, operator‑independent cartilage assessment. The full paper is available at:
🔗 https://www.umbjournal.org/article/S0301-5629(26)00157-2/fulltext
About SonoMaxx
SonoMaxx is a leading OEM manufacturer of handheld and portable ultrasound devices, delivering high‑quality imaging solutions for global medical technology brands. From design to production, SonoMaxx enables partners to bring reliable, innovative ultrasound systems to point‑of‑care markets worldwide.