Vietnam's first AI echocardiography deployment
The Cardiovascular Imaging Unit at the University of Medicine and Pharmacy Hospital, Ho Chi Minh City, became the first in Vietnam to apply AI to echocardiography. Over 120 studies were analyzed in the initial deployment, with the unit reporting a substantial reduction in the time taken to reach a diagnosis.
At a glance
Reading time in a high-volume imaging unit
A transthoracic echocardiogram is one of the most information-dense studies in cardiology and one of the most labor-intensive to report. The acquisition is only the first half of the work. A complete study requires chamber quantification, valve assessment, Doppler indices, and strain, each measured by hand on the workstation before the reader can compose a structured report and reach an interpretation.
In a university teaching hospital running a high daily study volume, that measurement burden sits directly on the critical path between a patient being scanned and a clinical decision being made. It also competes with teaching, with complex-case review, and with the subspecialty work that an academic cardiovascular imaging unit exists to do. The constraint is rarely imaging capability. It is the reader hours available to convert acquired images into a quantified, reportable study.
A first-in-country deployment at a university hospital
The University of Medicine and Pharmacy Hospital in Ho Chi Minh City deployed Us2.ai in its Cardiovascular Imaging Unit, becoming the first hospital in Vietnam to apply AI to echocardiography. The deployment was reported in January 2025 by Sức khỏe & Đời sống, the newspaper of Vietnam's Ministry of Health.
At the time of reporting, more than 120 echocardiograms at the hospital had been analyzed using the software. The hospital described strong diagnostic results and positive feedback from staff on both patient care and efficiency. The Cardiovascular Imaging Unit is headed by Associate Professor Dr Le Minh Khoi, who was quoted on the effect the tooling had on turnaround.
Professor Dr Truong Quang Binh, Chairman of the hospital's Scientific Council, framed the adoption as part of a wider institutional programme rather than an isolated software purchase, describing it as a breakthrough in the hospital's digital healthcare transformation and in its management of heart failure and cardiovascular disease.
Automated measurement, clinician interpretation
Us2.ai takes standard DICOM 2D echocardiograms and performs the measurement work end to end, returning a structured report with chamber quantification, valve assessment, Doppler indices, strain, and guideline-referenced diagnostic flags. No manual tracing or click-through approval of individual measurements is required to produce the report.
The clinical interpretation stays with the clinician. The reporting cardiologist reviews the automated measurements, edits anything the study warrants, and signs the report. What changes is not who is responsible for the diagnosis but how much of the reader's time is spent producing measurements before that judgement can be applied.
That distinction is the reason a deployment like this one can move quickly. The software slots into an existing DICOM workflow and an existing reporting responsibility, rather than asking the unit to reorganise around it.
What the imaging unit reported
Associate Professor Dr Le Minh Khoi, speaking to Sức khỏe & Đời sống, described the effect on diagnostic turnaround. His statement was given in Vietnamese; translated faithfully, it reads:
"AI can significantly shorten diagnosis time from 35 to 40 minutes down to about 10 minutes. This innovation not only eases the workload for medical staff but also minimizes errors and reduces treatment costs."
Associate Professor Dr Le Minh Khoi, Head of the Cardiovascular Imaging Unit, University of Medicine and Pharmacy Hospital, Ho Chi Minh City. Translated from Vietnamese.
This is a clinician's reported observation at a single site during an initial deployment, and it is worth reading as exactly that rather than as a specification. The figure describes the time to reach a diagnosis in that unit's workflow, it is given as an approximation, and it has not been published as a measured endpoint in a peer-reviewed study.
An independent figure from a second health system
A separately measured figure from a different health system points the same direction. At Endeavor Health in the United States, Dr Ilya Karagodin, an imaging cardiologist and the system's Director of AI and Advanced Analytics, reported that the AI workflow saves on average 10 to 12 minutes per study, a total saving of about 33% per study.
The two figures are not comparable and should not be averaged. They come from different countries, different baseline workflows, and different definitions of what is being timed. Read together, what they support is a directional claim about where automated measurement removes time from an echo service, not a single number that generalises across sites.
A template for high-volume public hospitals
Academic and public hospitals across Southeast Asia face a version of the same constraint: real imaging capability, high study volumes, and a reporting bottleneck that limits how many patients can move from scan to decision in a day. A first-in-country deployment at a university teaching hospital is a useful reference point because it is the setting where that pressure is most visible and where the clinical governance around AI-assisted reporting is most carefully examined.
For health systems evaluating a similar step, the operational shape of this deployment is the transferable part: a standard DICOM intake, automated measurement, an unchanged clinician sign-off, and a unit able to report a turnaround improvement inside the first cohort of studies.
Publications & announcements
AI Echo in Vietnam: Revolutionizing Cardiac Care and Heart Failure Diagnosis
Us2.ai news
Media coverage · Jan 2025Bệnh viện đầu tiên ở Việt Nam ứng dụng AI trong siêu âm tim
Sức khỏe & Đời sống, the Ministry of Health of Vietnam
Interview · Jul 2026Inside Endeavor Health's AI Echo Adoption: A Conversation with Dr. Ilya Karagodin
Us2.ai news, independent figures from a United States site