# Edges are all you need: Potential of medical time series analysis on complete blood count data with graph neural networks

## TL;DR

This paper explores using graph neural networks to analyze blood count data for sepsis diagnosis, showing improved performance when incorporating time series information.

## Contribution

The novel contribution is integrating time series information into GNNs via patient-centric graphs, achieving higher classification accuracy for sepsis detection.

## Key findings

- Standard GNNs on similarity graphs achieved an AUROC of up to 0.8747, comparable to other machine learning models.
- Patient-centric graphs with time series information improved AUROC to 0.9565, outperforming other methods.
- Feature importance and slope varied significantly between GNNs and traditional models like XGBoost.

## Abstract

Machine learning is a powerful tool to develop algorithms for clinical diagnosis. However, standard machine learning algorithms are not perfectly suited for clinical data since the data are interconnected and may contain time series. As shown for recommender systems and molecular property predictions, Graph Neural Networks (GNNs) may represent a powerful alternative to exploit the inherently graph-based properties of clinical data. The main goal of this study is to evaluate when GNNs represent a valuable alternative for analyzing large clinical data from the clinical routine on the example of Complete Blood Count Data.

In this study, we evaluated the performance and time consumption of several GNNs (e.g., Graph Attention Networks) on similarity graphs compared to simpler, state-of-the-art machine learning algorithms (e.g., XGBoost) on the classification of sepsis from blood count data as well as the importance and slope of each feature for the final classification. Additionally, we connected complete blood count samples of the same patient based on their measured time (patient-centric graphs) to incorporate time series information in the GNNs. As our main evaluation metric, we used the Area Under Receiver Operating Curve (AUROC) to have a threshold independent metric that can handle class imbalance.

Standard GNNs on evaluated similarity-graphs achieved an Area Under Receiver Operating Curve (AUROC) of up to 0.8747 comparable to the performance of ensemble-based machine learning algorithms and a neural network. However, our integration of time series information using patient-centric graphs with GNNs achieved a superior AUROC of up to 0.9565. Finally, we discovered that feature slope and importance highly differ between trained algorithms (e.g., XGBoost and GNN) on the same data basis.