AISAP
AISAP sells AI software for cardiac ultrasound at the point of care, built at the Heart Center of Sheba Medical Center in Ramat Gan, Israel. Cardio V1.0 analyzes acquired adult cardiac scans and returns a report a qualified physician uses alongside the examination: left ventricular ejection fraction, left ventricular end diastolic diameter, right ventricular fractional area change, inferior vena cava diameter, aortic root and atrial areas, plus grading of mitral, tricuspid and aortic regurgitation and aortic stenosis. FDA cleared those functions in August 2024 under K234141 for images from FDA cleared scanners, excluding congenital disease, intracardiac lesions, prosthetic valves and ventricular assist devices.
The wider POCUS-OS platform organizes scans, prepares reports, sends finalized reports to a medical record and adds an image archive that anonymizes DICOM files and stores them locally or in the cloud. It is marketed for any point of care ultrasound device, with Philips, GE, Siemens, Mindray, Clarius and Butterfly hardware shown. Sheba Medical Center, Stony Brook Medicine, Inova Health System and Amavita Heart and Vascular Health appear among its users, and patient stories describe findings outside the cleared list, a large pericardial effusion among them.
A prospective multicenter validation found an ejection fraction root mean squared error of 6.20 percentage points against cardiologists' readings, and a reader study in the clearance tested physicians with and without AI on four valve findings.
Capability Axes
An AI Health Index grade measures what a buyer can verify from public sources on the date shown. It is not a rating of how good the product is. A vendor can build an excellent system and grade low on an axis because it publishes nothing an outsider can check. How grades read
Cardio V1.0 measures structures and function from acquired cardiac ultrasound images and grades listed valve conditions for physician review, through classification, regression and instance segmentation models that identify structures and turn image patterns into measurements or categories.
POCUS-OS also handles scan workflow, image archiving and report delivery, services that support the clinical work, while the reason to buy Cardio is its automated reading of cardiac point of care ultrasound. The homepage sells the whole platform as an AI operating system for point of care ultrasound.
Cardio V1.0 presents its findings to qualified, privileged physicians as an adjunct to reading an adult cardiac scan, and the clearance summary keeps the physician responsible for patient management and says when suspected significant valve disease calls for a comprehensive examination. Clinicians review and finalize the report before it reaches the medical record, and the indications rule out congenital heart disease, intracardiac lesions, prosthetic valves and ventricular assist devices. The bedside report gives measurements and findings and leaves further testing and treatment to the clinicians responsible for the patient.
The Cardio V1.0 clearance summary describes neural networks for classification, regression and instance segmentation, states that two academic institutions supplied more than 140,000 exams to train them and summarizes their clinical characteristics, and separate tests measured the numeric and valve outputs against cardiologist interpretations. That ties a named software version to a model approach, a training corpus and a set of performance measurements. How customers learn of later model changes, and whether every POCUS-OS deployment runs the same Cardio release, get less detail, and no predetermined change control plan is authorized.
The FDA summary names Cardio V1.0 and describes AISAP's own neural networks for valve assessment and measurements, and the clearance fixes that software release and its intended use. Scanners produce the images, the archive stores them and the reporting workflow carries results onward, each with its own role. The parties that can process an image or model output in the hosted platform are not listed, and no retention or processing terms are given for any outside inference infrastructure.
A prospective multicenter study evaluated Cardio V1.0 measurements against cardiologists' readings: ejection fraction root mean squared error was 6.20 percentage points, against 8.23 between cardiologists, with separate results for chamber areas, inferior vena cava diameter and right ventricular fractional area change. The FDA file adds a randomized crossover reader study on 260 cases for four valve findings.
AISAP supported and helped design the prospective study, and the evaluations measure accuracy for the indicated findings rather than clinical outcome benefit in each setting. Homepage figures, billing raised to 95% and reporting time cut by up to 90%, carry no source.
The POCUS-OS image archive anonymizes DICOM files and stores them in the cloud or on a local installation, and the platform puts each image into a report workflow where the physician reviews findings before finalizing them. How the anonymization removes identifiers is not explained.
How long clinical images or identifying information stay in each part of the platform, whether they are reused to train models and how AISAP investigates a reported safety event involving an output are not stated.
POCUS-OS handles ultrasound studies and prepared reports that can go to the patient's clinical record, and AISAP shows a HIPAA mark and offers an image archive that anonymizes DICOM files. The public privacy notice covers the company's website and social media, and the terms send service use to separate agreements.
No business associate agreement sets out the scope for the archive, model processing or report delivery, so the compliance mark and the anonymization option stand apart from the data terms of a clinical installation.
The image archive offers local and cloud storage within an architecture that connects hospital systems to hosted components, and the website shows a HIPAA mark. The 510(k) summary says only that cybersecurity was evaluated and verified through software verification testing, without naming penetration testing, vulnerability scanning or any outside tester.
No SOC 2 examination, ISO 27001 certificate or security testing schedule covers the service that processes ultrasound images and reports, and the clearance speaks to the device and its indications rather than to how AISAP runs its operations.
Cardio V1.0 holds FDA clearance K234141 for qualified physicians reading adult cardiac point of care ultrasound from FDA cleared scanners, covering seven measurements and four valve findings. The homepage sells the platform as FDA cleared diagnostic algorithms for any point of care ultrasound device and promises bedside diagnoses from FDA cleared models, and a patient story credits AISAP with revealing a large pericardial effusion that led to a tamponade diagnosis. Pericardial effusion is not a cleared output, no other authorized module covers it, and the clearance limits use to FDA cleared scanners rather than any device.
The FDA file describes prespecified analyses of Cardio V1.0 performance by sex, age, body mass index and ejection fraction, and by scanner manufacturer, image quality and operator training, comparing outputs with cardiologist assessments in the measurement and valve studies. The valve classification results exceeded the prespecified AUC criterion of 0.80 across the studied groups, and the aided reader study's AUC rose across the same prespecified subgroups, with more detail in labeling. The analyses span clinical and technical sources of performance difference for the cleared outputs, with nothing equivalent for findings outside the indication or for a changed deployment.
The Cardio V1.0 validation reports numerical error against a cardiologist reference for each measurement, and the FDA summary sets out the reference method and case counts, including a 200 case evaluation of structural and functional measurements, with variation between cardiologists reported for comparison, so an institution can test the stated error against its own cases. Clinicians review images and report content before signing. No compensation provision, correction timeline or patient route to make AISAP remedy an incorrect result appears in the service terms.
POCUS-OS receives ultrasound images, stores DICOM files in its own archive and sends finalized reports to an electronic medical record after physicians review the findings, and the homepage says it works with any ultrasound device, PACS and EMR. No Epic, Oracle Health or other record system is named for the report, and no deployment shows which reads and writes are supported. Scanner compatibility and DICOM handling cover the imaging side of the workflow rather than a named record interface.
The dedicated image archive stores DICOM files in the cloud or on premises, within an architecture that links ultrasound devices and local hospital systems to cloud services, and a local archive lets the hospital choose where that copy of the images rests.
Which cloud regions hold hosted archives and backups, which analyses run entirely in the facility, whether images leave the facility for inference when local storage is chosen, and how tenants and cross border processing are handled are not stated.
AISAP sells Cardio and the connected POCUS-OS workspace through a demonstration and commercial discussion, as a package covering exam workflow, image storage, AI analysis and report delivery, and the homepage pitches it as a way to raise billing by turning images into billable encounters.
No rate per scan, unit or facility is shown, nor whether the archive and software can be bought on different bases, and implementation and connection charges for scanners, archives or record systems are settled in the contract.
Cardio analyzes adult cardiac point of care ultrasound and is used by teams in emergency, inpatient and outpatient care: Sheba Medical Center describes screening and medicine workflows, Amavita primary care and cardiovascular settings, and Stony Brook Medicine, Inova and Soffer Health Institute appear among its partners.
The clearance specifies adult cardiac images, privileged physicians and FDA cleared scanners and excludes congenital disease, intracardiac lesions, prosthetic valves and ventricular assist devices, and the seven measurements and four valve findings have reported validation. The scope stays within one imaging specialty and does not extend to pediatric or general ultrasound.
Alternatives to AISAP
Other products that do the same kind of work as AISAP, and what each one does.
EchoNous makes Kosmos, a handheld ultrasound system positioned against cart based systems rather than pocket scanners.
Vscan Air is GE HealthCare's handheld ultrasound: a wireless dual probe that pairs to a phone or tablet and fits in a coat pocket.
Clarius Mobile Health makes wireless handheld ultrasound scanners and the AI that runs on them.
Exo, pronounced echo, makes the Exo Iris handheld ultrasound probe and describes itself as a medical imaging software and devices company.
Philips Lumify is a handheld ultrasound transducer that plugs into a smart device the customer already owns, over USB on Android or Lightning on Apple hardware, with the app downloaded from the ordinary app stores.
Sonio makes prenatal ultrasound software for sonographers, maternal fetal medicine specialists and the obstetricians who read the scans, built as one platform with AI reading aids and a reporting and image management layer.
Pricing
Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.
| Entry Price | Pricing Basis | BAA Tier | Implementation | Source |
|---|---|---|---|---|
|
Quote based
|
Hospital and practice contracts through sales | No business associate agreement is offered for the platform. | Not published | Vendor Published |
Cardio V1.0 is sold within the POCUS-OS workflow, with license and implementation charges set in the agreement.