RIVANNA logo dark

The Problem

Decision-critical information is not consistently available at the point of care.

Across care settings, information and action remain separated.

Diagnostic decisions wait on specialized imaging infrastructure and interpretation, acquired apart from the decision point. In interventional procedures, clinicians often rely on surface landmarks or a preprocedural scan, without continuous image guidance during needle advancement. The result: delayed care, deeper dependence on specialized resources, and less consistency from one setting to the next.

The Solution

RIVANNA builds the decision-grade ultrasound platform.

Ultrasound engineered to automate complex anatomical analysis and support the next clinical action at the point of care.

RIVANNA brings decision-critical anatomical information into point-of-care workflows. Clinical judgment remains with the clinician; the platform reduces variability in acquisition and interpretation and provides a shared technical foundation for multiple clinical applications.

Four views explain the platform

01

Workflow architecture

Shows how it operates.

 02

Operating stack

Shows what RIVANNA engineers.

 03

Clinical data asset

Shows how models are developed and evaluated.

 04

Patent estate

Shows how the technology is protected.

Workflow Architecture

Decision support begins at acquisition.

Five configurable functions connect acquisition, verification, interpretation, guidance, and structured records.

Just-in-time training, guided acquisition, and at-capture quality feedback are built into the workflow to help shorten the learning curve associated with conventional ultrasound.

Five-function workflow with adaptive-acquisition feedback Five connected functions in sequence: acquire, verify, interpret, guide, connect. Data flows left to right along the spine. An orange feedback arc loops from interpret back to acquire, labeled adaptive acquisition, showing that accumulated information can reorder remaining views or end acquisition early. Adaptive acquisition 1 2 3 4 5 ACQUIRE VERIFY INTERPRET GUIDE CONNECT

Acquire

Task-defined acquisition

Capture task-defined views or live image data with controlled scan geometry and coverage of the intended anatomy.

Verify

Quality control at capture

Check sensor status, coupling, positioning, and image artifacts; prompt correction during acquisition.

Interpret

Structured anatomical and pathology outputs

Identify anatomy, classify or detect pathology, quantify relevant features, and consolidate information across frames, scans, and anatomical views into scan- and subject-level outputs.

Guide

Task-specific decision support

Use interpreted outputs to support the next action. Support rule-out or confirmation objectives, workflow prompts, measurements, and guidance for target location, depth, trajectory, alignment, avoidance, and confidence.

Connect

Structured outputs and system records

Package images, measurements, model metadata, timestamps, and device records for clinical-system exchange and cloud analytics.

Adaptive acquisition: In applicable workflows, accumulated information can reorder remaining views or end acquisition before the full planned set is complete.

Operating Stack

RIVANNA engineers the stack from transducer to cloud.

Four subsystems connect acoustic hardware, acquisition control, inference, and data infrastructure.

Each subsystem is designed around the inputs, outputs, and quality requirements of the others.

The operating stack Four subsystems in a stack from transducer to cloud: acoustic hardware, acquisition control, inference, and data infrastructure. Grey pulses flow left to right through them. An orange bus runs beneath all four, showing that each subsystem is designed around the shared inputs, outputs, and quality requirements of the others. Acoustic hardware Acquisition control Inference Data infrastructure INPUTS · OUTPUTS · QUALITY

Acoustic hardware and patient interfaces

Purpose-built probe architectures shape the data available for analysis and guidance.

Representative probe architectures include conformable volumetric imaging probes built around custom linear arrays and motorized translation, and paired-array probe geometries with a central access path. The subsystem also includes custom beamformers and front-end electronics, acoustic coupling materials, patient-contact structures, and a magnetic-field tracking system comprising probe-integrated sensors and an instrument-mounted field source.

Acquisition-control and quality software

Control software governs scan motion, planned views, and acquisition quality.

The subsystem includes motion-control software; planned-view protocols; device- and sensor-status checks; coupling-, positioning-, and artifact-detection modules; corrective-prompt logic; view-coverage inputs; and adaptive-sequencing and early stopping logic.

Image-processing and inference software

Algorithms and models convert ultrasound data into anatomical, pathology, and guidance information.

Image formation: multi-channel processing, signal separation, multi-angle reconstruction, fundamental and harmonic frequency methods, image enhancement.

Anatomical and pathology inference: segmentation, classification, model registration, multi-level inference, metadata and view-coverage inputs.

Guidance generation: quantification and overlay-generation modules.

Edge, record, and connectivity infrastructure

Compute and data services support local inference, traceability, and system integration.

The infrastructure includes on-device GPUs; DICOM-based image exchange with PACS; procedure-data integration with EMR; structured output and system records; software and model versioning; device-generated operating records; service monitoring; device and procedure analytics; and cloud data infrastructure.

Clinical Data

Reference-linked data make performance measurable.

Eight-plus clinical sites contribute ultrasound studies and reference data; ground truth is established through annotation and radiology linkage.

RIVANNA maintains a private, multi-site dataset pairing ultrasound studies with reference radiographs and radiology findings. It also includes image-quality metrics, patient metadata and outcomes, site and operator identifiers, and selected longitudinal device-linked performance data. A private annotation pipeline produces ground-truth ultrasound segmentations; radiology findings establish ground-truthed pathology classifications. Together, these assets support model training, validation, and performance analysis across sites, operators, and time. Selected, governed clinical data and device records from deployment can inform future model development, validation, and controlled platform updates.

Reference-linked dataset pipeline Ultrasound studies and reference radiographs are paired, passed through annotation to establish ground truth, and used to train and validate models. A dashed orange loop returns governed deployment data to the pipeline to inform future development. Ultrasound study Reference radiograph Annotation Ground truth Train / validate Governed deployment data

Intellectual Property

The estate covers integrated workflows, not isolated features.

Five cross-cutting claim domains cover both the components and the interfaces competitors would need to reproduce.

0+

active issued patents and pending applications

0+

patent families

Coverage

United States, Europe, Japan, and China

FTO

diligence completed across key feature sets

Portfolio figures as of Q2 2026.

RIVANNA owns or exclusively controls every filing, and foundational filings plus continuation practice extend claim coverage as the architecture develops. Trade secrets and know-how in signal processing, quality control, and acquisition standardization reinforce the estate. The portfolio grows about 30% year over year.

Claim domain
Capability coverage
AI and machine learning
Anatomy segmentation, classification, and target detection; synthetic augmentation of anatomical features missing from, obscured in, or distorted in source images; frame-, scan-, and subject-level pathology inference across anatomical views; image and device quality control.
Cloud analytics and data infrastructure
Device connectivity; structured records; model and software versioning; device-generated operating records; device and procedure analytics; cloud data infrastructure.
Robotics and automation
Motorized transducer translation; controlled scan paths; adaptive scan ordering and early stopping; image-to-model registration; hardware-linked instrument guidance; magnetic-field tracking of instrument position and trajectory; pending application coverage directed to six-degree-of-freedom tracking of needle position and orientation relative to the probe.
Image processing and workflow
Bone and soft-tissue signal separation; multi-angle reconstruction; image formation from fundamental and harmonic frequency components; two- and three-dimensional bone imaging; acquisition standardization; source-and-synthetic-image fusion.
Hardware and interfaces
Beamformers and ultrasound front-end electronics; custom linear arrays and volumetric imaging probe architectures; paired convex arrays with a central access path; acoustic coupling materials; patient-contact structures; sterile consumables.

The 100th filing

The 100th filing describes a unified pathology-detection pipeline that combines custom multi-channel ultrasound inputs and feature enhancement with frame-, scan-, and subject-level inference across anatomical views; clinical metadata and view coverage; adaptive scan ordering and early stopping; and self-supervised pretraining. The filing is structured to support downstream continuation filings.

Commercialization Infrastructure

RIVANNA takes technology from patent filing to controlled production.

An ISO 13485:2016-certified quality management system and 1,300+ systems shipped globally as of Q2 2026 provide operating evidence for diligence and transfer-to-production planning.

ISO 13485:2016

Certified quality management system

0+

systems shipped globally as of Q2 2026

Since 2010, RIVANNA has built in-house ultrasound, electrical, mechanical, software, AI, quality, supply-chain, and manufacturing capabilities in Charlottesville. Its controlled development and production environment includes design controls, risk analysis, software and hardware verification, supplier qualification, production documentation, and change control.

Partner and acquirer diligence can draw on patent schedules, claims, and assignments; freedom-to-operate memoranda; data-use agreements and de-identification procedures; software bills of materials; and controlled engineering and manufacturing records.

Commercial Pathways

One technical foundation. Two commercial entry points.

Initial applications establish the commercial base; adjacent procedures, populations, and care settings build on shared core technology.

One foundation, two entry points, room to expand A shared core technology foundation at the center radiates outward to two solid entry-point nodes, trauma triage and interventional guidance, and to four dashed slots representing adjacent expansion. Trauma triage Interventional guidance FOUNDATION Shared core Entry point Expansion

Trauma triage

Adult extremity injury assessment in settings where radiography access or capacity constrains the next action.

Interventional guidance

Anatomical interpretation and instrument guidance for spinal procedures that rely on landmark techniques or preprocedural ultrasound rather than continuous image guidance during needle advancement.