Assisted Living Applications Addressing Sleep Dysfunction in Parkinson's Patients
Annotatsiya
The field of assisted living technologies is strongly influenced by neurological disorder management requirements. The aim of the present study was to compare the predictive performance of sleep dysfunction indicators, specifically the patterns of nocturnal disturbance, in Parkinson's disease patients and in non-neurodegenerative control groups due to progressive motor and non-motor symptoms using regression and AHP techniques, and to propose a model for the measurement of sleep quality decline using the SUM (Summation Utility Modelling) method. In this study, we established a new method called Priority-Weighted Regression Assessment (PWRA) to detect variability trends of sleep-related disruptions more sensitively and accurately. Based on multi-model statistical evaluation, we find that movement rigidity, REM behavior disorder, and circadian rhythm irregularity have positive influences on sleep dysfunction severity. Several factors were identified, including frequency of nocturnal awakenings, duration of sleep cycles, motor symptom onset time, dopaminergic medication timing, psychological stress levels, and daily physical activity levels of elderly patients. Calibration and validation of the AHP-regression hybrid model using empirical inputs and ranking outputs (SUM method) to prevent diagnostic uncertainty were implemented. The results show that under the integrated multi-criteria decision-making structure, the changes in feature correlation strength and predictive accuracy index of the proposed model with different data dimensions are all statistically significant. The research results provide important clinical and technological guidance for improving patient-centered care. The modular evaluation protocol applied in this study may be useful for further research studies regarding digital health interventions and other neurological monitoring tools. While sensor-based telemonitoring is rapidly progressing in the area of biometric data acquisition and real-time condition assessment, rising cost barriers and lack of availability of the standardized evaluation frameworks open a demand for affordable and adaptable forms of assisted monitoring systems in resource-constrained care environments. The modular evaluation protocol applied in this study may be useful for further research studies regarding digital health interventions and other neurological monitoring tools.