HOME » PROJECTS » MEDICINE » ECO PROJECT
ECO PROJECT

The echocardiogram is a test widely used in the diagnosis of heart disease. His analysis is strongly related to the physician's experience. Methods that apply artificial intelligence have become a powerful tool to help doctors who are specialists in this exam.

We are a group of researchers who apply Artificial Intelligence, Machine Learning, Computer Vision and technological resources to the development of methods to computer aided diagnostics.

Methodology

Ultrasound equipment generates high-definition images, making it possible to apply Artificial Intelligence models to quantify measurements of cardiac structures, create detection models of the internal and external structures of the heart and create models for classifying heart diseases.

Artificial Intelligence models reduce the time needed to analyze exams, thus simplifying clinical decision-making. In this way, heart disease can be identified early, increasing the chances of successful treatment.

The task of identifying structures of interest, also known as Region of Interest segmentation (ROI) can be performed with extreme ease by current Convolutional Neural Networks.
The initial results of the models developed in the project showed an accuracy of 93.77% in the segmentation task.
With this, specialists can dedicate their attention to more relevant tasks that require their participation in the examination and construction of the diagnosis.

Partners in Eco Project