Explainable deep learning for disease activity prediction in chronic inflammatory joint diseases

Abstract

AbstractAnalysing complex diseases such as chronic inflammatory joint diseases (CIJDs), where many factors influence the disease evolution over time, is a challenging task. CIJDs are rheumatic diseases that cause the immune system to attack healthy organs, mainly the joints. Different environmental, genetic and demographic factors affect disease development and progression. The Swiss Clinical Quality Management in Rheumatic Diseases (SCQM) Foundation maintains a national database of CIJDs documenting the disease management over time for 19’267 patients.We propose the Disease Activity Score Network (DAS-Net), an explainable multi-task learning model trained on patients’ data with different arthritis subtypes, transforming longitudinal patient journeys into comparable representations and predicting multiple disease activity scores.First, we built a modular model composed of feed-forward neural networks, long short-term memory networks and attention layers to process the heterogeneous patient histories and predict future disease activity.Second, we investigated the utility of the model’s computed patient representations (latent embeddings) to identify patients with similar disease progression.Third, we enhanced the explainability of our model by analysing the impact of different patient characteristics on disease progression and contrasted our model outcomes with medical expert knowledge. To this end, we explored multiple feature attribution methods including SHAP, attention attribution and feature weighting using case-based similarity.Our model outperforms non-temporal neural network, tree-based, and naive static baselines in predicting future disease activity scores. To identify similar patients, ak-nearest neighbours regression algorithm applied to the model’s computed latent representations outperforms baseline strategies that use raw input features representation.Author summaryChronic inflammatory joint diseases affect about 200′000 patients in Switzerland alone. These conditions lead to immune system dysfunction resulting in inflammation that targets the joint tissues. Understanding which aspects of patients’ characteristics and disease management history are predictive of future disease activity is crucial to improving patients’ quality of life.A significant obstacle to the widespread adoption of deep learning (DL) methods in healthcare is the challenge of understanding their “black-box” nature (i.e. the underlying decision process for outcome generation). Therefore, the development of “explainable” deep learning methods has become an active area of research. These approaches aim to provide insights into the inner workings of deep learning models, enabling physicians to understand and assess the output of DL models more effectively.We propose DAS-Net: an explainable deep learning model that finds similar patients and predicts future disease activity based on past patient history. In our analysis, we contrast different explainability approaches highlighting which aspects of the patient history impact model predictions the most. Furthermore, we show how computed patient similarities allow us to rank different patient characteristics in terms of influence on disease progression and discuss how case-based explanations can enhance the transparency of deep learning solutions.