Voice Spectrogram Dementia Diagnosis via CNN
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for early dementia diagnosis are inconvenient, expensive, and require specialized medical expertise, often resulting in late detection of mild cognitive impairment or Alzheimer's disease, as patients need to visit hospitals for neurocognitive tests and imaging procedures.
Innovation Solution
A method and apparatus using an electronic device that determines dementia severity based on user voice analysis through a series of pre-produced content tasks, generating spectrogram images, and processing them with updated convolutional neural networks and deep neural networks to output dementia degrees.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional neurocognitive tests and imaging procedures are used for dementia diagnosis, then diagnostic accuracy is improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces complex medical imaging systems (MRI, SPECT, PET) and specialized neurocognitive testing equipment with a simple mobile terminal device that uses voice recording and basic image display capabilities. The diagnostic function is transferred from sophisticated medical equipment to a commonplace consumer device, eliminating the need for specialized medical institutions while maintaining diagnostic capability through software-based analysis of voice spectrograms and image responses.
2Reliability
If specialized medical expertise is required for neurocognitive testing, then diagnostic reliability is improved, but ease of operation deteriorates
Solution Approach 1:
The system enables users to perform their own dementia screening without requiring medical professionals. The mobile terminal automatically presents test images, records user responses and voice readings, generates spectrogram analysis, and provides diagnostic results independently. The embedded algorithms automatically evaluate the collected data and determine dementia risk levels, eliminating the need for expert intervention in the testing process while maintaining diagnostic reliability through standardized evaluation criteria.
3Measurement precision
If hospital visits are required for dementia diagnosis, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The system performs preliminary dementia screening actions that can be completed immediately at home using the mobile terminal, eliminating the need for patients to wait for hospital appointments. By conducting the initial assessment remotely through voice and image-based tests, the system provides rapid preliminary results that can guide whether further medical intervention is necessary, significantly reducing the time from symptom recognition to initial diagnosis.
4Measurement precision
If expensive imaging procedures are used, then measurement precision is improved, but cost increases
Solution Approach 1:
The patent replaces expensive, resource-intensive imaging procedures (MRI, SPECT, PET) with a low-cost software-based solution that runs on existing mobile terminals. The diagnostic system uses freely available or pre-installed camera and microphone functions, processes data through lightweight algorithms, and generates results without requiring costly medical equipment. This substitutes expensive consumable resources with inexpensive computational processing, making dementia screening accessible to broader populations.
Data Source
AI summary
In order to determine a degree of dementia of a user, contents are output through a user terminal, a voice of the user for a content acquired by a microphone of the user terminal is received, a spectrogram image is generated by visualizing the voice, and the degree of dementia of the user is determined by means of a convolutional neural network (CNN) and a deep neural network (DNN) based on the spectrogram image.