Xiaoli Chu

The growing prevalence and complexity of multimorbidity pose significant challenges to healthcare systems worldwide. Here, we developed a heterogeneous graph neural network framework for multilabel multimorbidity prediction. The framework jointly encodes symptom and diagnostic information within a heterogeneous graph and incorporates biomedical knowledge subgraphs on both the disease and symptom sides. By integrating macro‐level clinical representations of symptoms and diseases with micro‐level molecular knowledge, the model generates multilabel probabilities across candidate disease categories. We evaluated the framework in two complementary tasks: Patient‐Level Symptom–Disease Association (Task 1) and Visit‐Level Multimorbidity Completion (Task 2). In Task 1, symptoms recorded across all eligible visits within each patient's observation window were aggregated into a patient‐level symptom set to predict the disease categories recorded for that patient within the same observation window. In Task 2, symptoms and a subset of recorded positive diagnoses from the same visit were used to predict the remaining disease categories recorded for that visit. Together, these two tasks provide a unified framework for patient‐level symptom–disease association and visit‐level multimorbidity completion. The framework characterized symptom–disease–biomedical knowledge associations and non‐random patterns of disease co‐occurrence, while supporting both patient‐level symptom–disease association and visit‐level multimorbidity completion.

Chronic non-communicable diseases (CNCDs) are the leading cause of morbidity and mortality worldwide. Currently, diagnostic information for CNCDs remains incomplete and is subject to ongoing change, leading to uncertainty in diagnostic conclusions and severely impacting disease control. Therefore, it is necessary to study decision-making theory and methods for assisting clinical diagnosis. Given the characteristics of clinical diagnostic decision problems, this paper discusses the temporally correlated rough set theory and its three-way clustering model. Firstly, the temporal correlation rough set model based on the gradient and cosine similarity is defined. Then, the clinical diagnostic decision process for CNCDs is transformed into a three-way clustering problem with temporal correlation attributes. A three-way clustering model based on a temporal correlation rough set is constructed to support clinical diagnosis decisions for CNCDs. In this model, the common features of temporal association attributes are identified using a positive domain-based three-way clustering algorithm. Finally, the validity and applicability of the theoretical model are verified using real clinical data from 80,139 visit time points and 83 laboratory examination indicators from 3,094 chronic kidney disease (CKD) patients. Based on the algorithm’s calculations and a retrospective literature review, β2-microglobulin is identified as a candidate marker for CKD that warrants further validation and could provide auxiliary decision support for clinical practice. The main contribution of this paper is twofold. One is to make a new theoretical contribution to the data-driven research paradigm of clinical diagnosis decision-making. Another is to provide quantitative decision support for clinical diagnosis.