Automated AAA detection

Start date: 08-03-2021
End date: 17-12-2021

Clinical problem

An abdominal aortic aneurysm (AAA) is a local enlargement of the abdominal artery. Patients with a AAA have an increased risk of rupture, often leading to death. It is therefore vital to early detect and closely monitor these patients. Detection and surveillance of AAA are currently performed by a specialist in the hospital, who measures the diameter of the AAA using ultrasound.

Ultrasound devices have recently become cheaper and portable. These portable devices can be connected to laptops, tablets and even smartphones, making them accessible for a wide range of users, including physicians in the first line.

Solution

In this project we will evaluate the use of a standardized sweep protocol using a low-cost ultrasound device that can be connected to a smartphone. The data that will be acquired with this protocol will be used to train a deep learning algorithm for automated measurement of the aorta diameter.

Approach

Development of deep learning algorithms that automatically determines the aortic diameter using a predefined ultrasound acquisition protocol.

Results

When the algorithms have sufficient sensitivity and accuracy, and are able to run on a smartphone, they will be integrated in the current prototype.

People

Douwe van Erp

Douwe van Erp

Master Student

Diagnostic Image Analysis Group

Thomas van den Heuvel

Thomas van den Heuvel

Postdoctoral Researcher

Diagnostic Image Analysis Group

Chris de Korte

Chris de Korte

Professor

Medical UltraSound Imaging Centre

Bram van Ginneken

Bram van Ginneken

Professor

Diagnostic Image Analysis Group

Tom Heskes

Tom Heskes

Professor

Data Science, Radboud University