Skip to main content

Single Heart beat MApping of myocardial Performance, Activation, and Scar by ultrasound

Final Report Summary - HEARTMAPAS (Single Heart beat MApping of myocardial Performance, Activation, and Scar by ultrasound)

The overall goal of HeartMAPAS was to provide the cardiologist with information on local mechanical properties of the heart that are currently not available through non-invasive ultrasound imaging. As such, maps of the performance of the cardiac muscle (i.e. contractility), mechanical activation of the muscle and scar were to be built.

The key idea behind HeartMAPAS was that this mapping is possible through advanced signal/image processing in case the (volumetric) ultrasound data is acquired at sufficiently high temporal resolution. In order to achieve this, HeartMAPAS proposed to scan the anatomically relevant space only, which is in contrast to the current state-of-the-art (volumetric) ultrasound scan sequences that scan a predetermined pyramidal volume of data. This new anatomical imaging paradigm combined with novel fast imaging strategies was shown to increase frame rate by a factor of 15 to 30 leading to an effective (volumetric) frame rates of 500-1000Hz (where the state-of-the-art systems currently provide typically 30-40Hz only).

In the project, the following developments were made in order to make such fast anatomical imaging a reality:

1) An algorithm was developed that enables an ultrasound system to fully automatically recognize the location of the cardiac muscle and thereby define the anatomically relevant space.

2) An algorithm was proposed to automatically determine the optimal scan sequence of the anatomical space. Hereto, it first determines the required image line layout and then defines a scan sequence to reconstruct these lines as time-efficient as possible while retaining image quality.

3) Algorithms were developed to automatically process the recorded high frame rate data in order to extract physiologically relevant parameters of the cardiac muscle such as its local motion and deformation. These data combined with the shape information available from step ‘1’ above, enabled to construct maps of the local performance and activation of the cardiac muscle as well as the local scarring of the muscle.

The above developments were initially made on synthetic data generated by a fast ultrasound simulation methodology (also developed within the project) but were subsequently enabled on an experimental ultrasound scanner. Each of the developments mentioned above were individually validated in experimental setups.

At present, the clinical translation of the developed ultrasound imaging protocol is being made which will lead to a further verification of the technology developed but also to a true clinical validation of the diagnostic value of these technologies.