O

Solutions

Using machine learning, we build tools for everything echo.
Take a tour of our solutions below.

iCardio.ai Ejection Fraction Simpson's Method

Ejection Fraction via Simpson's Method and other Left Ventricular Calcuations

Calculate left ventricular area and volume, ejection fraction, axis of symmetry, and diameters, all at systolic, diastolic and mid points in the cardiac cycle in the Apical, Two, Three, and Four Chamber perspectives.

EF: 33%
LVSV: 170ML
LVDV: 245ML
LVSD: 6.31CM

iCardio.ai Ejection Fraction Teicholz's Method

Ejection Fraction via Teicholz's Method and Left Ventricular Calcuations

Ejection Fraction via a modified Teichholz's Method and Left Ventricular Calcuations at Mitral, Papillary and Apex levels in the Parasternal Short Axis perspective.

EF: 41%
LVSV: 90ML
LVDV: 152ML

Left Atrium in Apical Views

Identification, segmentation and quantification of the left atrium.

Automatically determine left atrial function and dimensions.

LA Dimension: 3.3CM

Right Atrium in Apical Views

Identification, segmentation and quantification of the right atrium.

Automatically determine right atrial function and dimensions.

RA Dimension: 2.9CM

Aortic Annulus and Sinus of Valsalva in Parasternal Views

This model can automatically determine the diameter of at the aorta annulus. Our AI by default measures the area between the aorta annullus and the sinotubular junction, to extract additional area information and to reinforce conclusions behind the diamteric measurement of the sinotubular junction.

Sinus of Valsalva: 2.8CM
Aortic Annulus: 1.8CM

Sinotubular Junction in Parasternal Views

Automatically determine the diameter of the sinotubular junction across the cardiac and valve cycle and instantly flag abnormalities.

Sinotubular Junction: 2.4CM

Left Ventricular Outflow Tract in the A5C Perspective

Quantify the area of the left ventricular outflow tract to produce a first-pass assessment.

LVOT Area: Healthy

Inferior Vena Cava

Quantification of the inferior vena cava from the right atrium to the hepatic vain.

Navigation and Guidance

This model extracts features in real-time using a mobilenet model architecture. By providing labeled features of the heart, ultrasound novices can orient themselves around the heart.

High fidelity in sub-par and moving conditions

View Classification and Grading

This model extracts features in real-time using a mobilenet model architecture. By providing labeled features of the heart, ultrasound novices can orient themselves around the heart.

High fidelity in sub-par and moving conditions

Performance and Data Science

All models undergo rigorous performance testing. In this example, we can visualize performance of the model's prediction over each frame of the cardiac cycle.