[OES-1-1] Gaze behavior during Trail Making Test contributing to the prediction of cognitive functional independence after stroke
Introduction
Trail Making Test (TMT) is a common neuropsychological test. Recently, novel technologies like eye trackers enabled the assessments of gaze behaviors during the TMT, and a relationship between gaze behavior and cognitive functions has been revealed. Although it is known that the TMT is useful in the prediction of rehabilitation outcomes such as the Functional Independence Measure (FIM) after stroke, how the gaze behaviors during the TMT contribute to the prediction of the outcomes is unclear. This study aimed to evaluate gaze behavior during the TMT and determine their predictive capability for cognitive function after stroke.
Methods
Twelve patients with stroke (11 males and 1 female, 66.5±12.3 years, mean UE-FMA 62.3±4.9/66 points) were recruited. The study was approved by the Ethical Review Board of the hospital, and informed consent was obtained from all patients. A glasses-type eye tracker (Tobii pro Glass 2, Tobii Technology Co., Ltd.) and an electronic TMT (electronic peg board K3000, Work joy Co., Ltd.) were used. The subjects wore the eye tracker and performed the electronic TMT part A. They were asked to touch numbers on the screen as quickly as possible with a touch pen in their non-paretic hand, based on the rule of TMT. In all 25 targets, the number of times that the gaze stayed on the target even after the target was tapped was calculated as “overstay fixation”. The cognitive FIM was measured as a clinical assessment. A stepwise multiple regression analysis was performed with five cognitive FIM items as the dependent variables, and the age, onset time, overstay fixation, and completion time in the electronic TMT. The multiple regression analysis determined the goodness of fit using adjusted R-squared. The statistical software SPSS ver. 23.0 J was used for the analysis. Values of p < 0.05 were considered significant.
Results
Two multiple regression models were identified. First, the FIM Comprehension score was predicted by the electronic TMT (standard β = -1.08, p < 0.01) and the onset time (standard β = -0.37, p = 0.03), and the adjusted R-squared value was 0.85. Second, the FIM Memory score was predicted by the electronic TMT (standard β = -0.75, p < 0.01) and the overstay fixation (standard β = -0.40, p = 0.04), and the adjusted R-squared value was 0.71.
Discussion
The FIM Comprehension and Memory scores were accurately predicted by the electronic TMT and the overstay fixation. This suggests that combining conventional assessment of completion time with information on gaze behavior or techniques not interfering with visual search may contribute to a more accurate prediction of rehabilitation outcomes in specific cognitive domains after stroke.
Trail Making Test (TMT) is a common neuropsychological test. Recently, novel technologies like eye trackers enabled the assessments of gaze behaviors during the TMT, and a relationship between gaze behavior and cognitive functions has been revealed. Although it is known that the TMT is useful in the prediction of rehabilitation outcomes such as the Functional Independence Measure (FIM) after stroke, how the gaze behaviors during the TMT contribute to the prediction of the outcomes is unclear. This study aimed to evaluate gaze behavior during the TMT and determine their predictive capability for cognitive function after stroke.
Methods
Twelve patients with stroke (11 males and 1 female, 66.5±12.3 years, mean UE-FMA 62.3±4.9/66 points) were recruited. The study was approved by the Ethical Review Board of the hospital, and informed consent was obtained from all patients. A glasses-type eye tracker (Tobii pro Glass 2, Tobii Technology Co., Ltd.) and an electronic TMT (electronic peg board K3000, Work joy Co., Ltd.) were used. The subjects wore the eye tracker and performed the electronic TMT part A. They were asked to touch numbers on the screen as quickly as possible with a touch pen in their non-paretic hand, based on the rule of TMT. In all 25 targets, the number of times that the gaze stayed on the target even after the target was tapped was calculated as “overstay fixation”. The cognitive FIM was measured as a clinical assessment. A stepwise multiple regression analysis was performed with five cognitive FIM items as the dependent variables, and the age, onset time, overstay fixation, and completion time in the electronic TMT. The multiple regression analysis determined the goodness of fit using adjusted R-squared. The statistical software SPSS ver. 23.0 J was used for the analysis. Values of p < 0.05 were considered significant.
Results
Two multiple regression models were identified. First, the FIM Comprehension score was predicted by the electronic TMT (standard β = -1.08, p < 0.01) and the onset time (standard β = -0.37, p = 0.03), and the adjusted R-squared value was 0.85. Second, the FIM Memory score was predicted by the electronic TMT (standard β = -0.75, p < 0.01) and the overstay fixation (standard β = -0.40, p = 0.04), and the adjusted R-squared value was 0.71.
Discussion
The FIM Comprehension and Memory scores were accurately predicted by the electronic TMT and the overstay fixation. This suggests that combining conventional assessment of completion time with information on gaze behavior or techniques not interfering with visual search may contribute to a more accurate prediction of rehabilitation outcomes in specific cognitive domains after stroke.