SoniΦ Assist™

Supporting Standard View Acquisition
Anywhere

SoniΦ Assist™ is an intelligent scanning guide
that analyzes probe positioning and orientation
in real time, empowering examiners to
seamlessly acquire the standard views required
for precise measurement and analysis.

From emergency rooms (ER) and intensive care
units (ICU) to medically underserved regions,
it ensures consistent standard view acquisition
across diverse clinical environments.

Enhancing Consistency
in Standardized Image Acquisition

Real-time AI guidance elevates the consistency
of standard view acquisition, serving as the very foundation
for examination standardization.

• Enhanced Exam Consistency
• Improved Patient Comfort

CORE FUNCTIONS

Real-Time Transducer Orientation Inference
& Automated Standard View Classification

01

AI Image Quality
Assessment Module

Simultaneously with image acquisition,
the AI scores image quality in real time,
assisting clinicians in seamlessly selecting
the optimal frames best suited
for precise measurement and analysis.

Technical References:
Jeong, Sihyeon, et al. "Deep learning-based multi-view
echocardiographic framework for comprehensive diagnosis
of pericardial disease." European Heart Journal-Digital Health 7.3(2026)

Park, Jiesuck, et al. "Single-View Echocardiographic Analysis
for Left Ventricular Outflow Tract Obstruction Prediction
in Hypertrophic Cardiomyopathy: A Deep Learning Approach."
Journal of the American Society of Echocardiography 38.12 (2025):1115-1126.

Park, Jiesuck, et al. "Artificial intelligence-enhanced automation
of left ventricular diastolic assessment: a pilot study
for feasibility, diagnostic validation, and outcome prediction."
Cardiovascular Diagnosis and Therapy 14.3 (2024): 352-366.

02

Real-Time Transducer
Orientation Inference

AI instantly analyzes the live image
to display the transducer movement direction
vector on-screen in real time.

This directly addresses the highly angle-dependent nature
of echocardiography, where even subtle angular variations can
significantly impact image quality.

03

Reduction of Repetitive Scanning

AI guidance helps minimize repetitive scanning
caused by initial failures in standard view acquisition.
By integrating seamlessly with handheld and
compact ultrasound devices, it is highly adaptable
to various clinical environments.

SoniΦ Assist™

4 Steps
of the AI Scan Guide

Step 1: Real-Time Video Input

Analyzes the ultrasound stream frame-by-frame, providing vendor- and model-agnostic compatibility.
Step 2: Image Quality Assessment

Supports standardized examination quality through real-time AI scoring and feedback.
Step 3: Transducer Manipulation Inference

Calculates movement directions and angles to assist probe manipulation via on-screen guidance.
Step 4: Standard View Capture & Transfer

Supports the examination workflow by prompting image freeze/save and enabling seamless SoniX EchoQuant™ integration.

Three Core Values

Workload Reduction via
Workflow Optimization

Lowers clinician workload
by shortening examination times
and maximizing equipment utilization efficiency.
Enhanced Consistency
Through Exam Standardization

Reduces variability across
the examination process
by ensuring consistent image quality.
Optimal Positioning via
Real-Time Navigation

Assists operators in rapidly locating
the precise standard view position.

Advanced Technology Roadmap

Evolving into the Future

Building upon SoniΦAssist™’s current Software AI-based
real-time guidance, our research and development is advancing toward
the next frontier: Physical AI. We are committed to continuously evolving
our solutions to overcome the long-standing challenges of echocardiography.

제품 성능 표
Phase 1 · Present	Phase 2 · Near Future	Phase 3 · Future
2024 – 2026 Intelligent Image Acquisition Support • Real-time transducer orientation guidance • 5-tier image quality grading and feedback 37-view classification + HOLD detection • Ongoing integration of lightweight models into Samsung Medison platforms	2026 – 2027 Autonomous Scan Assistance System • Accounting for multi-initial position variability • IMU sensor fusion-based position tracking • Quantitative transducer movement inference • AI training with Yale & UCSF clinical datasets	2027 ~ Physical AI • Autonomous Robotic Scanning • Automated transducer manipulation via robotic arms • Fully autonomous, end-to-end standard view acquisition • Specialist-independent response for emergency and remote care settings • Realization of a fully autonomous echocardiography platform

Partnership

We are engaged in the joint research
and development of a robotic ultrasound system
with Life Science Robotics (Denmark).

By combining collaborative robot platforms
with SoniΦ Assist™ technology, we are currently advancing
the research phase to realize
a Physical AI echocardiography system capable
of physical transducer control.

Other products

SoniX EchoQuant™

Automatically measures 81 indicators
across multiple cardiac cycles,
reducing measurement variability
and shortening exam time.

More View

SoniX EchoPilot LVH™

Differential Diagnosis Support.
Assists clinicians in differentiating
the underlying causes of myocardial hypertrophy,
including hypertrophic cardiomyopathy (HCM),
cardiac amyloidosis, and hypertensive heart disease.

More View

Supporting Standard View Acquisition Anywhere

Enhancing Consistency in Standardized Image Acquisition

CORE FUNCTIONS

AI Image Quality Assessment Module

Real-Time Transducer Orientation Inference