AI Cognitive Model for Rapid Diagnosis via Game Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional cognitive state evaluation methods are time-consuming and fatiguing, often requiring multiple hours of medical examinations and tasks, which can deter subjects like children and the elderly from completing diagnoses effectively.
Innovation Solution
A cognitive state evaluation system utilizing a learning-based user-customized cognitive model that extracts game data from a user's input information for a short cognitive game, creating a customized cognitive task performance model through artificial intelligence to automatically process tasks necessary for diagnosing cognitive states, significantly reducing the time and fatigue required for diagnosis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional medical examinations and cognitive tasks are performed to diagnose cognitive states, then diagnostic accuracy is improved, but diagnosis time and subject fatigue increase significantly
Solution Approach 1:
The patent creates a virtual copy of the cognitive evaluation process by training an AI model to simulate human cognitive task performance. The trained model replicates the diagnostic evaluation function, allowing the system to obtain cognitive state assessment results without requiring the subject to actually perform time-consuming medical examinations and cognitive tasks, thus resolving the contradiction between diagnostic accuracy and diagnosis time
Solution Approach 2:
The patent performs preliminary training of the AI model using game data before actual cognitive evaluation. The model is pre-trained and validated to ensure it can accurately diagnose cognitive states. This preliminary action allows the model to be ready for deployment, enabling rapid evaluation without requiring extensive actual testing time while maintaining diagnostic accuracy
2Measurement precision
If conventional medical examinations and cognitive tasks are performed to diagnose cognitive states, then diagnostic accuracy is improved, but subject fatigue and refusal to participate increase
Solution Approach 1:
The patent replaces the need for subjects to perform actual cognitive tasks with an AI model that copies the evaluation function. The model processes game data to generate cognitive state assessments, eliminating the physical and mental burden on subjects while preserving diagnostic accuracy through the model's learned evaluation capabilities
Solution Approach 2:
The system enables self-service cognitive evaluation by using the trained AI model to automatically assess cognitive states from game data without requiring subject participation in traditional medical examinations. The model serves itself by processing available game data to generate diagnostic results, making the evaluation process convenient and comfortable for subjects
3Adaptability or versatility
If multiple cognitive tasks are administered to evaluate different cognitive domains, then comprehensive evaluation is improved, but evaluation time and complexity increase
Solution Approach 1:
The patent creates a universal AI model that can perform multiple cognitive evaluation functions through a single system. The trained model processes various types of game data to assess different cognitive domains (attention, memory, executive function, etc.), enabling comprehensive evaluation without requiring separate specialized tasks for each cognitive domain, thus reducing overall system complexity
Solution Approach 2:
The patent merges multiple cognitive evaluation tasks into a unified AI model framework. Instead of administering separate tests for different cognitive domains, the model integrates processing of game data to simultaneously evaluate multiple cognitive functions, combining what would otherwise be separate evaluation processes into one cohesive system that reduces complexity while maintaining comprehensiveness
Data Source
AI summary
A method of operating a cognitive state evaluation apparatus according to an embodiment of the present invention comprises the steps of extracting first game data for cognitive evaluation from a user's response input information to a cognitive game application; creating a customized cognitive task performance model corresponding to the user by applying the extracted first game data to a user-customized cognitive model based on artificial intelligence learning that a cognitive architecture-based cognitive model has been pre-associatively trained in response to game data for each cognitive task; and obtaining a substitution performance result of the cognitive task selected for each cognitive evaluation item using the customized cognitive task performance model and evaluating the cognitive ability of the user for each cognitive item based on the substitution performance result.