Deep Learning in Cardiology

TL;DR

This review surveys the application of deep learning across structured cardiology data, signals, and imaging, detailing architectures, data sources, and performance trends. It highlights common DL pipelines and the predominance of CNNs and RNNs in ECG, PCG, MRI, and CT tasks, while noting limitations such as dataset size, generalizability, and interpretability. The authors discuss challenges in data availability, annotation quality, and the need for end-to-end, multimodal models that are clinically interpretable and externally validated. They propose future directions including attention mechanisms, capsule networks, GANs for data augmentation, CRFs for context, and collaborative data sharing to accelerate clinical deployment.

Abstract

The medical field is creating large amount of data that physicians are unable to decipher and use efficiently. Moreover, rule-based expert systems are inefficient in solving complicated medical tasks or for creating insights using big data. Deep learning has emerged as a more accurate and effective technology in a wide range of medical problems such as diagnosis, prediction and intervention. Deep learning is a representation learning method that consists of layers that transform the data non-linearly, thus, revealing hierarchical relationships and structures. In this review we survey deep learning application papers that use structured data, signal and imaging modalities from cardiology. We discuss the advantages and limitations of applying deep learning in cardiology that also apply in medicine in general, while proposing certain directions as the most viable for clinical use.

Figures (2)

• Figure 1: Perceptron. It consists of a set of connections $x_{1, \ldots, n}$ as its input, the weights $w_{1, \ldots, n}$, the bias $b$, the activation function $\phi$ and the output $\alpha$.
• Figure 2: A Convolutional Neural Network that calculates the LV area ($\hat{y}$) from an MRI image ($x$). The pyramidoid structure on top denotes the flow of the feed-forward calculations starting from input image $x$ through the set of feature maps depicted as 3D rectangulars to the output $\hat{y}$. The height and width of the set of feature maps is proportional to the height and width of the feature maps while the depth is proportional to the number of the feature maps. The arrows at the bottom denote the flow of the backpropagation starting after the calculation of the loss using the cost function $J$, the original output $y$ and the predicted output $\hat{y}$. This loss is backpropagated through the filters of the network adjusting their weight. Dashed lines denote a 2D convolutional layer with ReLU and Max-Pooling (which also reduces the height and width of the feature maps), the dotted line denotes the fully connected layer and the dash dotted lines at the end denote the sigmoid layer. For visualization purposes only a few of the feature maps and filters are shown, and they are also not in scale.