In this article we will investigate the different categories of machine learning driven solutions that can provide added value in cathlab procedures. Categorizing machine learning applications in the cathlab, and structurally investigating their respective data needs, aids in developing a systematic approach to the data collection and algorithm development challenges. Machine learning applications in the cathlab can be divided in four data categories, dependent on the type of data they receive as input: 1) image-based, 2) 1D signals such as ECG, respiratory, etc., 3) natural language processing, 4) hybrid or other data sources. Machine learning algorithms, regardless of their input data type, typically address a fine-grained task, such as object detection, signal quality improvement, image registration, event detection, etc. These fine-grained algorithmic blocks then feed into high level applications, such as device navigation, lesion quantification, patient risk stratification, functional parametrization, etc.

The strengths and plasticity of machine learning techniques make them an attractive solution for many tasks that cannot easily be automated otherwise. Particularly, convolutional networks have demonstrated robust performance and versatility in segmentation tasks, and can be easily retrained to handle newly introduced devices. In this article we will investigate the different categories of machine learning driven solutions that can provide added value in cathlab procedures.