Objective: To develop a non-invasive cognitive impairment detection system using speech data analysis, addressing the growing global dementia crisis and enabling accessible early screening through daily health monitoring. Methods: Speech data from 223 Korean patients were collected across eight tasks. Patients were classified based on Korean Mini-Mental State Examination scores. Four machine learning models were tested for three binary classification tasks. Voice acoustic features were extracted and analyzed. Results: The Deep Neural Network model performed best in two classification tasks, with Precision-Recall Area Under the Curve scores of 0.737 for severe vs. no impairment and 0.726 for mild vs. no impairment, while Random Forest achieved 0.715 for severe + mild vs. no impairment. Several acoustic features emerged as potentially important indicators, with DDA shimmer from the /i/ task and stdevF0 from the /puh-tuh-kuh/ task showing consistent patterns across classification tasks. Conclusions: This preliminary study suggests that certain acoustic features may be associated with cognitive status, though demographic factors significantly influence these relationships. Further research with demographically matched populations is needed to validate these findings.
Introduction In patients with idiopathic normal pressure hydrocephalus (iNPH), the characteristics of static steady-state balance disturbance are not as well understood as those related to gait. This study aimed to investigate differences in center of pressure (COP) parameters between patients with iNPH and age-matched healthy controls. Methods A total of 56 patients diagnosed with probable iNPH, who responded positively to cerebrospinal fluid tap test (CSFTT), along with 25 age-matched healthy controls, were enrolled. Prior to the CSFTT, COP measurements were taken using a force plate while participants stood quietly with eyes open (EO) and eyes closed (EC). We calculated the values of mediolateral (ML) and anteroposterior (AP) directions and ML/AP ratio in terms of the velocity of COP (vCOP), the standardized COP path length (Std_vCOP), and the peak power spectral density (PSD) at 0–0.5 Hz and 0.5–1.0 Hz. Additionally, we calculated the average distance from the COP center (Dist_COP) and the base of support (BOS). A Mann–Whitney U test with Bonferroni correction was used to compare differences between healthy controls and patients with iNPH under each visual condition (EO vs. EC). Results There were no significant differences in gender, age, height, and weight between the healthy controls and the patients with iNPH. Patients with iNPH demonstrated greater spontaneous sway in the ML direction of vCOP (p < 0.001), increased variability in both ML and AP directions of Std_vCOP (p < 0.001 for both), a larger Dist_COP (p < 0.001) and BOS (p < 0.001) in the EC condition. Under the EO condition, patients with iNPH exhibited a significantly higher ML/AP ratio (p < 0.001) and larger BOS (p < 0.001). Furthermore, peak PSD at 0–0.5 Hz, reflecting low-frequency oscillations, was more pronounced in the ML direction and in the ML/AP ratio (p < 0.001 for both under EO), as well as in the AP direction (p < 0.001 under EC) in patients with iNPH. However, no significant differences were observed in the high-frequency oscillations (0.5–1.0 Hz). Conclusions These findings suggest that patients with iNPH exhibit distinct COP patterns, particularly greater variability in the ML direction during quiet standing. In contrast, the AP direction appears to be more influenced by visual conditions when compared to healthy controls. This result highlights the potential need for tailored balance training in patients with iNPH.