Predicting Outcomes in Long COVID Patients with Spatiotemporal Attention

TL;DR

This paper introduces a novel spatiotemporal attention mechanism combined with Local-LSTM to improve outcome prediction in long COVID patients, addressing heterogeneity in longitudinal clinical data.

Contribution

It proposes a new spatiotemporal attention model with Local-LSTM to better predict long COVID outcomes from complex longitudinal data.

Findings

Outperformed state-of-the-art methods in outcome prediction

Provided a clinical tool for severity assessment of long COVID

Validated on a challenging clinical dataset

Abstract

Long COVID is a general term of post-acute sequelae of COVID-19. Patients with long COVID can endure long-lasting symptoms including fatigue, headache, dyspnea and anosmia, etc. Identifying the cohorts with severe long-term complications in COVID-19 could benefit the treatment planning and resource arrangement. However, due to the heterogeneous phenotype presented in long COVID patients, it is difficult to predict their outcomes from their longitudinal data. In this study, we proposed a spatiotemporal attention mechanism to weigh feature importance jointly from the temporal dimension and feature space. Considering that medical examinations can have interchangeable orders in adjacent time points, we restricted the learning of short-term dependency with a Local-LSTM and the learning of long-term dependency with the joint spatiotemporal attention. We also compared the proposed method with several state-of-the-art methods and a method in clinical practice. The methods are evaluated on a hard-to-acquire clinical dataset of patients with long COVID. Experimental results show the Local-LSTM with joint spatiotemporal attention outperformed related methods in outcome prediction. The proposed method provides a clinical tool for the severity assessment of long COVID.