Temporal similarity measures for querying clinical workflows.

Clinical workflows involve sequences of activities carried out by healthcare professionals. These workflows are often stored in databases for analysis and reuse. Prior research has shown that workflow modeling can support decision-making and quality improvement. However, comparing workflows across different patient cases remains challenging. Existing methods focus on structural similarity but often ignore temporal aspects. Temporal constraints are crucial for evaluating how clinical activities unfold over time. No prior work had resolved how to systematically compare workflows based on both structure and timing. This gap motivated the development of a new approach to assess temporal similarity between clinical workflow cases. The need for such a method arises from the desire to improve information retrieval and quality evaluation in clinical settings.

The study aims to develop a formal method for evaluating temporal similarity between clinical workflow cases. Clinical workflows involve activities performed by specific actors in a defined order. These workflows can vary in structure and timing depending on patient conditions. The authors propose a way to model clinical processes using temporal constraint networks. This approach allows for comparing workflows based on both activity order and duration. The goal is to enable querying of clinical databases by temporal similarity. The method also supports quality assessment by comparing real cases to ideal ones. The study focuses on stroke prevention and management workflows in Italy. The proposed approach addresses limitations in existing workflow comparison techniques.

The researchers first describe a conceptual model for clinical processes using temporally extended workflow schemata. Clinical activities are represented with temporal constraints to capture timing relationships. Workflow cases are modeled as sets of activities with associated temporal constraints. The authors define temporal constraint networks to represent these cases formally. A similarity function is introduced to compare workflow cases based on activity order and duration. The function also considers the presence or absence of specific activities. The approach is applied to clinical workflows related to stroke prevention and management. The method is tested on clinical databases storing workflow cases from different patient scenarios.

The proposed approach successfully models clinical workflows using temporal constraint networks. The similarity function evaluates differences in activity order and duration between workflow cases. The method can identify cases with similar temporal patterns despite structural variations. The approach supports querying clinical databases based on temporal similarity. The function can compare real clinical cases to ideal or synthetic reference cases. The method is particularly useful for stroke prevention workflows in Italy. The similarity measure accounts for both structural and temporal aspects of workflows. The results demonstrate the feasibility of using temporal similarity for clinical workflow analysis.

The authors propose a method to evaluate temporal similarity between clinical workflow cases. The approach uses temporal constraint networks to model workflows formally. The similarity function considers both activity order and duration. The method supports querying clinical databases for similar cases. The approach can be used to assess service quality by comparing cases to ideal references. The method is applicable to stroke prevention workflows in Italy. The results suggest that temporal similarity can improve information retrieval in clinical settings. The study demonstrates the feasibility of using temporal similarity for workflow analysis.