FDA Clears RIVANNA's BoneEnhance Machine Learning Module for Accuro XV Ultrasound

The US Food and Drug Administration has cleared BoneEnhance, a machine learning software module for RIVANNA's Accuro XV diagnostic ultrasound system, through the 510(k) pathway, the Charlottesville, Virginia, company announced on October 22.

The FDA's 510(k) database lists submission K262093 for the Accuro XV from Rivanna Medical, Inc., received on June 22, 2026, with a decision of substantially equivalent on October 20, 2026. The device is classified as a pulsed-echo ultrasonic imaging system under regulation 892.1560 and was reviewed by the radiology panel as a traditional 510(k). RIVANNA said K262093 is the clearance that adds BoneEnhance to the system.

What the software does

Accuro XV uses a conformable three-dimensional probe that, according to the company, automates large field-of-view image acquisition after the patient is positioned, producing a volumetric dataset with each scan. RIVANNA said BoneEnhance identifies bone and surrounding tissues within that volume and highlights them with color-coded overlays on the two-dimensional B-mode image, leaving the original image data visible underneath. A three-dimensional indicator beside the image shows the scan orientation and where the displayed slice sits in the volume.

The company said BoneEnhance is the first machine learning module cleared for the platform. The Accuro XV system itself was cleared in April 2026 under K254021 for B-mode imaging in musculoskeletal conventional and superficial applications, according to RIVANNA.

"BoneEnhance is where the stack begins to close that gap for musculoskeletal care: automated volumetric acquisition, now paired with cleared software-assisted review," said Will Mauldin, RIVANNA's co-founder and chief executive.

Validation data

The performance figures come from the company and have not been published in a peer-reviewed journal. RIVANNA said BoneEnhance was trained and validated within its eight-site US clinical study. Three sites supplied training data, and five others provided an independent test set of 233 volumetric datasets, made up of 69,900 images from 173 subjects, none of which were used for training or tuning. Three board-certified radiologists set the reference labels.

The company reported per-volume label prediction accuracy of 95.1% to 100.0%, DICE scores of 0.64 to 0.85 and HD95 scores of 2.26 to 7.48 mm. RIVANNA said these measures evaluate segmentation and localization, not diagnostic accuracy.

Separately, the company cited a 205-patient study in the Journal of Emergency Medicine in which, it said, operators with one hour of hands-on training acquired wrist and ankle images rated adequate for diagnostic interpretation in more than 90% of scans.

Fracture module still investigational

RIVANNA said it is developing a musculoskeletal injury detection module meant to help clinicians evaluate potential extremity fractures on Accuro XV images. The company stated that the module is investigational, not FDA-cleared and not available for sale. Its clinical study has expanded from eight to ten sites to build the training data, the company said.

The work has been supported with federal funds from the Biomedical Advanced Research and Development Authority (BARDA), part of the US Department of Health and Human Services, under contract 75A50121C00035, according to the release. AuntMinnie also reported the clearance and the BARDA support for the fracture-detection work.

RIVANNA said it positions Accuro XV as a point-of-triage tool for low-severity extremity injuries in emergency departments, with the aim of reducing reliance on X-ray referrals. The company said its initial applications are in interventional guidance and trauma triage, beginning with neuraxial procedures.