PERFORMANCE OF SUSTAINED ATTENTION TEST (SAT) AND PSYCHOMOTOR VIGILANCE TASK (PVT) FOR VIGILANCE EVALUATION METHOD: STUDY IN A TRAIN DRIVING SIMULATOR

• Main Author: Siswanto, Daniel
• Format: Dissertations
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/62332
• Institution: Institut Teknologi Bandung

The monotonous but high-risk characteristics of a train driving demand high vigilance. A decrease in vigilance or sustained attention while driving a train is said to be a contributing factor to accidents. This decrement can occur due to drowsiness or fatigue from sleep deprivation. This sleep deprivation often occurs in train transportation which generally uses a shift system that tends to restrict the duration of sleep. To minimize the risk of train accidents due to vigilance decrement, it is necessary to evaluate train drivers' vigilance before driving. So far, the evaluation of driving fitness has been carried out with medical examinations and interviews related to driver fatigue. However, this is not sufficient as it cannot detect the state of vigilance. Therefore, a valid, accurate, and sensitive vigilance test method is needed to complement the medical examinations that have been carried out. So far, the Psychomotor Vigilance Task (PVT) which measures vigilance with a reaction time indicator is considered the gold standard for vigilance measurement. Many studies had found that PVT was valid and sensitive to measure changes in vigilance due to sleep deprivation. But so far there were no studies that use it in the context of monotonous train driving activities. On the other hand, the Sustained Attention Test (SAT) offers a vigilance measurement not only with a reaction time indicator but also working memory which plays an important role in vigilance function. The SAT has been used to evaluate train drivers' vigilance in Belgium and was able to distinguish between a fit and unfit drivers due to sleep problems. Although new and not widely known, the SAT has the potential to be a method for evaluating train drivers’ vigilance. PVT or SAT can be a potential alternative method of vigilance measurement for train drivers before driving. However, the performance of both measurements still needs to be investigated further in conditions that are often experienced by train drivers, i.e sleep deprivation, monotony, and differences in work shifts (time of day). Performance is evaluated from three aspects, i.e validity, accuracy, and sensitivity. In addition, if it is used to evaluate vigilance, it is necessary to determine the outcome metrics that can be used as a reference and cut-off points that can distinguish the level of vigilance. To the best of my knowledge, there have been no studies that assessed the performance of PVT and SAT concerning the train driving activities and determining the outcome metrics with its cut-off point that can be used as a reference for evaluating vigilance. The objectives of this study were to assess the performance of PVT and SAT, and determine the outcome metrics with its cut-off values that can be used as a reference for evaluating vigilance. To achieve those objectives, two studies involve train driving activity simulation was performed. The safety factor and insignificant performance differences with real conditions were the reasons for using the simulator in this study. Study I aimed to examine the validity of PVT and SAT under restricted sleep duration (4, 6, and 8 hours). Validity was assessed by analysis of changes in vigilance patterns, Two-way repeated-measures ANOVA, and correlation. Study II aimed to assess the performance of PVT and SAT, based on their validity, accuracy, and sensitivity. In addition, the outcome metrics and cutoff values that can be used as a reference for evaluating vigilance were also determined. Analysis of the pattern of change in vigilance, Three-way Mixed ANOVA, correlation, logistic regression, and Receiver Operating Characteristics (ROC) were used in Study II. The results of this study indicate that the SAT was sensitive to sleep deprivation and monotony in the train driving simulation. However, there were no differences in the vigilance from the SAT outcome metrics between morning and afternoon shifts. The SAT showed a strong correlation with PVT, KSS, and relative band powers of the EEG which indicating high criterion validity. The SAT also had high accuracy and sensitivity to distinguish low and high vigilance. Four SAT outcome metrics that can be used as a reference for vigilance evaluation were number of missed targets (%), number of delayed responses (%), mean 1/RT (1/second), and mean RT (milliseconds). PVT was also sensitive to sleep deprivation and monotony in the train driving simulation. However, there were no differences in vigilance from the PVT outcome metrics between morning and afternoon shifts. PVT showed a strong correlation with KSS and relative band powers EEG which showed high criterion validity. PVT also had high accuracy and sensitivity to distinguish low and high vigilance. Three PVT outcome metrics that can be used as a reference for vigilance evaluation were minor lapses (%), mean 1/RT (1/second), and mean RT (milliseconds). It can be concluded that SAT and PVT have good performance to measure vigilance on the effect of sleep deprivation and monotony. However, the SAT has the advantage of involving working memory for vigilance measurement with the number of missed targets as the outcome metric.