PET Quantitative Localization System for Epilepsy Focus Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current PET imaging for medial temporal lobe epilepsy relies heavily on subjective human interpretation, leading to inaccurate localization of epileptogenic foci, especially in cases with mild hippocampal sclerosis, resulting in inconclusive results and potential need for invasive procedures.
Innovation Solution
A PET quantitative localization system that includes a processor and memory circuit for acquiring and processing PET and MRI images, performing spatial registration, intensity normalization, and using machine learning models to calculate lateralization indexes, thereby providing objective and accurate determination of target regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If subjective visual comparison of PET images is used for determining epilepsy lesions, then the method is simple and accessible, but the measurement precision and reliability deteriorate when differences between sides are not large
Solution Approach 1:
The patent replaces the manual visual comparison method (mechanical human interpretation) with an automated computer-based quantitative analysis system. The system uses processors to execute algorithms that objectively compare PET images, calculate asymmetry ratios, and generate standardized uptake value maps, thereby eliminating subjective bias and improving measurement precision while maintaining ease of operation through automated processing.
Solution Approach 2:
The patent transforms the qualitative visual assessment into quantitative parameters by calculating standardized uptake values (SUV), asymmetry ratios, and lateralization indexes. These numerical parameters enable precise objective comparison of metabolic activity between left and right temporal lobes, significantly improving measurement precision while the automated calculation maintains operational simplicity.
2Reliability
If invasive intracranial electrode implantation is performed to determine epilepsy lesions, then the diagnostic reliability improves, but the device complexity and patient risk increase
Solution Approach 1:
The patent creates a virtual quantitative model of brain metabolism by generating standardized uptake value maps and asymmetry ratio images that replicate the diagnostic information needed for lesion localization. This virtual copy provides reliable diagnostic data without requiring physical invasive procedures, thereby maintaining diagnostic reliability while eliminating device complexity and patient risk associated with intracranial electrodes.
3Productivity
If conventional PET imaging without quantitative analysis is used, then the workflow is simple, but the loss of information occurs due to inability to detect subtle metabolic differences
Solution Approach 1:
The patent performs preliminary quantitative processing of PET images by calculating standardized uptake values and asymmetry ratios before clinical interpretation. This preliminary action extracts and highlights subtle metabolic differences that would be imperceptible in conventional visual analysis, preserving information while the automated nature of the calculation maintains workflow simplicity.
Data Source
AI summary
The present disclosure provides an operation method of a PET (positron emission tomography) quantitative localization system, which includes steps as follows. The PET image and the MRI (magnetic resonance imaging) of the patient are acquired; the nonlinear deformation is performed on the MRI and the T1 template to generate deformation information parameters; the AAL (automated anatomical labeling) atlas is deformed to an individual brain space of the patient, so as to generate an individual brain space AAL atlas, where the AAL atlas and the T1 template are in a same space; lateralization indexes of the ROIs of the individual brain space AAL atlas corresponding to the PET image normalized through the gray-scale intensity are calculated; the lateralization indexes are inputted into one or more machine learning models to analyze the result of determining a target.