Clinical Pathway Graph Construction via Hybrid Knowledge and Data Mining
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Developing clinical pathways is a complex and cumbersome process that requires integration of knowledge from multiple stakeholders, with manual methods being ineffective due to the vast amount of data in patient health records and automatic data mining resulting in large, difficult-to-navigate graphs that do not align with healthcare providers' mental models.
Innovation Solution
A computer-implemented method and system that combines knowledge-driven manual user input with automated data-driven mining within a graphical user interface (GUI) to create a cohort clinical pathway graph, reducing complexity by computing data-driven nodes relative to manually defined nodes and presenting a structured GUI that aligns with medical guidelines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If automatic data mining is used to analyze patient health records, then comprehensive clinical pathways can be discovered, but the resulting graphs become large and difficult to navigate
Solution Approach 1:
The patent segments the clinical pathway graph into hierarchical levels: high-level clinical pathways provide overall structure and navigability, while detailed event sequences are contained within collapsible sub-graphs. This segmentation allows users to navigate the high-level structure without being overwhelmed by detailed data, while still accessing comprehensive information when needed.
Solution Approach 2:
The patent introduces temporal dimension to the graph visualization by ordering events chronologically along pathways and allowing collapse/expand operations along the time axis. This dimensional organization transforms the complex multi-dimensional data into a temporally-ordered sequence that is easier to navigate and understand clinically.
2Reliability
If manual methods are used to construct clinical pathways, then the process requires integration of knowledge from multiple stakeholders, but manual methods are ineffective due to the vast amount of data
Solution Approach 1:
The patent introduces an intermediary system that translates clinical knowledge from multiple stakeholders into structured pathway definitions, which then guide automated data mining. This intermediary layer preserves the reliability of expert knowledge integration while enabling automated processing of vast data volumes, bridging the gap between manual expertise and automated efficiency.
Solution Approach 2:
The patent performs preliminary action by having stakeholders define clinical pathways and select relevant variables before data mining begins. This preliminary structuring of clinical knowledge allows the subsequent automated data mining to be focused and efficient, rather than attempting to discover everything from raw data without guidance.
3Productivity
If fully automated data mining is performed, then large amounts of data can be processed, but the results do not align with healthcare providers' mental models
Solution Approach 1:
The patent inverts the traditional approach by having clinical pathways defined from the top down based on healthcare providers' mental models, then using these predefined pathways to guide data mining. Instead of letting data mining generate pathways bottom-up from raw data, the system constrains the mining process to discover events within clinically-defined pathway frameworks, ensuring alignment with clinical mental models.
4Loss of information
If detailed data-driven nodes are included in the graph, then comprehensive event discovery is achieved, but processor utilization and processing time increase
Solution Approach 1:
The patent makes the graph structure dynamic by implementing collapsible and expandable nodes. Detailed data-driven events are contained within collapsible sub-graphs that can be expanded when needed and collapsed for overview. This dynamic structure allows the system to maintain complete event discovery capability while reducing processing and display overhead by only rendering necessary levels of detail at any given time.
Data Source
AI summary
There is provided a method of creating a cohort clinical pathway graph based on knowledge-driven manual user input and automated data-driven mining comprising: receiving via a graphical user interface (GUI), manual selections including: knowledge-driven variable(s) denoting clinically significant values representing elements of a clinical decision making process, and an anchoring location of each knowledge-driven node denoting a respective knowledge-drive variable within a directed acyclic graph (DAG), computing individual clinical pathways for each of the sampled population of patients by automatically computing data-driven nodes denoting the data-driven discovery of event types relative to the manual selections, and aggregating the individual clinical pathways to compute a cohort clinical pathway DAG, wherein the cohort clinical pathway DAG includes nodes comprising the knowledge-driven nodes, the data-driven nodes, and links connecting the nodes, each link denoting an automatically discovered sequence between two respective nodes, and presenting the cohort clinical pathway DAG within the GUI.