Yahan Zhang

Purpose: To determine the response time and protective mechanism of the cyclic guanosine monophosphate (cGMP) channel in 661w cells. Methods: 661w cells were exposed to 4500Lux visible light for three and four days at the following exposure time periods per day: 20, 30, 60, 90, 120, and 180. Cells were incubated for the rest of the time without any other treatment. Cell activity and cell death rates were measured with Hoechst/PI (diphenylmethane/propidium iodide) staining. Western Blot was used to detect the levels of guanylate cyclase-activating proteins 1 (GCAP1), cGMP, and phosphodiesterase (PDE)6 in the cGMP-gated channel. Results: 661w cells showed low mortality within three days. The mortality rate increased from the fourth day, especially during the longer times (120 and 180 min) of light exposure. After three-day illumination, the level of cGMP increased after 20 and 90 min and the level of GCAP1 increased after 60 and 90 min. After four days of illumination, the level of GCAP1 upregulated after a time of 20 and 60 min, while the cGMP level decreased from 30 min. The expression of PDE6 upregulated at each light period. Conclusion: The survival rate of 661w cells was relevant to the time of light exposure. The changes in GCAP1, cGMP, and PDE6 levels over time were possibly related to cell metabolism and restoration after light-induced damage.

Purpose: To construct a hidden Markov model (HMM) for vigilance assessment to improve the real-time performance and accuracy of current vigilance measurement. Methods: ECG signal was collected by sensors, while the noise and baseline drift was eliminated from the original ECG signal. 10 volunteers were randomly selected. Their heart rate variability (HRV) were measured and trained parameters of the modified Hidden Markov model for vigilance assessment. Then, these data were collected to optimize using the Baum-Welch algorithm and obtained the state transition probability matrix A ̂ and the observation probability matrix B ̂. Finally, the data of three volunteers with different transition patterns of mental state were selected randomly and used the Viterbi algorithm to find the optimal state, which compared with the actual state. Results: The constructed vigilance assessment model had a high accuracy rate the accuracy rate of data prediction for these three volunteers exceeded 80%. Conclusion: The Hidden Markov model for vigilance assessment can accurately predict the vigilance level and indicate broad application prospects.