Predictive FIM Assessment Processing for Timely Rehabilitation Plans
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Therapists in rehabilitation facilities often lack time for frequent assessments of the Functional Independence Measure (FIM), leading to outdated assessments and potentially inappropriate rehabilitation plans due to the time-consuming nature of the 18-item evaluation, which may not capture the patient's latest condition.
Innovation Solution
An information processing system that calculates and displays predicted FIM assessment values on a device for therapists, allowing for correction and updating, using machine learning to refine the model based on actual assessments and therapist attributes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a therapist performs frequent FIM assessments to capture the latest patient condition, then the assessment accuracy and rehabilitation plan appropriateness improve, but the time required for assessment increases significantly
Solution Approach 1:
The system performs preliminary assessment by automatically calculating predicted FIM scores based on historical data and patient information before the therapist conducts the actual assessment. This preliminary calculation reduces the time required for full assessments while maintaining accuracy, as therapists only need to review and confirm predictions rather than perform complete evaluations from scratch
Solution Approach 2:
The system creates a copy of the assessment process by using machine learning models to generate predicted FIM scores that mirror what a therapist would assess. These predicted values serve as proxies for actual assessments, allowing frequent monitoring without requiring therapist time for each individual assessment, thus resolving the contradiction between assessment frequency and time consumption
2Reliability
If the FIM assessment is performed every day to track patient progress, then the timeliness of rehabilitation planning improves, but the FIM values often show no change making the assessments redundant
Solution Approach 1:
The system performs partial assessment by only requiring therapist review of predicted values that are displayed on their devices. Instead of conducting full assessments every day, the system presents pre-calculated predicted FIM scores for confirmation or correction, reducing the assessment burden to minimal verification actions while maintaining timely updates to rehabilitation plans
Solution Approach 2:
The system serves itself by automatically generating predicted FIM assessments using machine learning models without requiring therapist intervention for each assessment. The predicted values are automatically calculated, stored, and made available for therapist review, enabling the system to maintain current assessment data without consuming therapist time, thus achieving both timeliness and efficiency
3Measurement precision
If the therapist reviews and corrects predicted FIM values, then the assessment accuracy improves, but the therapist workload increases
Solution Approach 1:
The system performs the labor-intensive work of calculating FIM predictions in advance using machine learning models, presenting only the results for therapist review. This preliminary calculation eliminates the need for therapists to perform time-consuming manual assessments, allowing them to focus only on reviewing and correcting predictions when necessary, thus improving accuracy without significantly increasing workload
Solution Approach 2:
The system implements a feedback mechanism where predicted FIM values are presented to therapists for verification. Therapists review the predictions and provide corrections when needed, creating a feedback loop that continuously improves the accuracy of the machine learning model while keeping therapist involvement minimal. This feedback-based approach ensures high assessment accuracy without imposing excessive workload on therapists
Data Source
AI summary
An information processing apparatus according to the present invention includes: a calculating unit configured to calculate, based on subject information including a first assessment value representing an assessment of a subject at a predetermined moment for each of a plurality of items set in FIM (Functional Independence Measure), a prediction value representing an assessment of the subject predicted after the predetermined moment; and a control unit configured to set the prediction value as a provisional assessment value representing a provisional assessment of the subject for a predetermined item of the FIM, and control to output so as to display in a correctable manner on an information processing device operated by an assessor.