Multimodal MRI Analysis of Brain Metabolism in Maintenance Hemodialysis Patients Based on Cognitive Computing

Abstract

This paper investigates cognitive computation of brain metabolism in maintenance hemodialysis patients with multimodal MRI therapy assessment. This paper constructs a cross-individual emotion recognition method using dynamic sample entropy pattern learning. The cross-individual emotion recognition was carried out on subjects using the EEG emotion dataset SEED. The experimental results show that the proposed dynamic sample entropy-based pattern learning has better performance in cross-individual emotion recognition and exhibits better generalization and generalization ability when compared with the results of existing related studies. The constructed cognitive computing method for cross-individual emotion state recognition achieves optimization and innovation of EEG emotion pattern recognition, which can effectively predict people’s mental emotion state from EEG signals. We also explore the value of diffusion-weighted magnetic resonance imaging and dynamic enhanced magnetic resonance imaging-based volumetric measurements in assessing the efficacy of neoadjuvant therapy in maintenance hemodialysis patients. We analyze and compare the results of different studies to find the best multimodal MRI to assess the efficacy of neoadjuvant therapy in maintenance hemodialysis patients. The use of ADC value growth rates to assess neoadjuvant efficacy provides the best diagnostic efficacy and allows the screening of patients who respond well to neoadjuvant therapy while avoiding the impact of two different b-value combinations commonly used to assess neoadjuvant efficacy.