The lack of situation awareness (SA) is the main reason for tower controllers' human errors. How to evaluate the SA of tower controllers is an important problem for controlling the risk of air traffic control system. The current method of measuring the SA of air traffic controllers by using eye movement analysis focuses on radar controllers. To verify the applicability of eye movement analysis in measuring the SA of tower controllers, an experiment platform was built by using the eye tracker and tower control simulation software. The comparison of the two group features including SA, task load, and operation performance, were conducted by the nearest neighbor index (NNI). Results show that the subjects whose fixation feature is dispersed distribution have high SA and improve operation performance under a low task load. The Pearson correlation coefficient between the NNI and SA reveals a significant positive moderate association (r=0.602, p<0.01). Eye tracking can measure the SA of tower controllers, and the NNI can be an auxiliary eye index for SA measurement in dynamic, uncontrolled environments. The conclusion of this study can provide a new reference index for the real-time quantitative assessment of tower controllers' situation awareness level.

This paper discusses relationship between Situation Awareness and aviation accident, and proposed SA-oriented Design process for human-machine interface design. After some common designing principles, we discuss emphatically on the 10 levels of automation, and dynamic functional allocation that combines the human and machine generation is preferred.

To explore how pilots' distribution of visual attention affects flight performance, twenty male pilots (novices and experts with 407±11.3 h and 4127±77 h of flight experience, respectively) were enlisted to complete the instrument holding pattern and approach procedure on the DA-42 simulator. The distribution of visual attention was based on eye movement data recorded during the flight to investigate how pilots scan the flight instrument panel, which was divided into six areas of interest (AOIs). To evaluate the pilots' flight performance an expert scoring method was used. During the outbound-leg stage, experts paid significantly more visual attention to the airspeed indicator, altimeter and reference system, whereas for the approach phase, they devoted more attention to the airspeed indicator, altimeter and vertical speed indicator. Results showed that experts' proportions of gaze duration on different AOIs contributed to their better performance. An effective visual attention model can be developed on this study to improve air traffic safety.

PurposeTotal-body positron emission tomography/computed tomography (PET/CT) provides faster scanning speed, higher image quality, and lower injected dose. To compensate for the shortcomings of the maximum standard uptake value (SUVmax), we aimed to normalize the values of PET parameters using liver and blood pool SUV (SUR-L and SUR-BP) to predict programmed cell death-ligand 1 (PD-L1) expression in non-small cell lung cancer (NSCLC) patients.Materials and methodsA total of 138 (104 adenocarcinoma and 34 squamous cell carcinoma) primary diagnosed NSCLC patients who underwent 18F-FDG-PET/CT imaging were analyzed retrospectively. Immunohistochemistry (IHC) analysis was performed for PD-L1 expression on tumor cells and tumor-infiltrating immune cells with 22C3 antibody. Positive PD-L1 expression was defined as tumor cells no less than 50% or tumor-infiltrating immune cells no less than 10%. The relationships between PD-L1 expression and PET parameters (SUVmax, SUR-L, and SUR-BP) and clinical variables were analyzed. Statistical analysis included χ2 test, receiver operating characteristic (ROC), and binary logistic regression.ResultsThere were 36 patients (26%) expressing PD-L1 positively. Gender, smoking history, Ki-67, and histologic subtype were related factors. SUVmax, SUR-L, and SUR-BP were significantly higher in the positive subset than those in the negative subset. Among them, the area under the curve (AUC) of SUR-L on the ROC curve was the biggest one. In NSCLC patients, the best cutoff value of SUR-L for PD-L1-positive expression was 4.84 (AUC = 0.702, P = 0.000, sensitivity = 83.3%, specificity = 54.9%). Multivariate analysis confirmed that age and SUR-L were correlated factors in adenocarcinoma (ADC) patients.ConclusionSUVmax, SUR-L, and SUR-BP had utility in predicting PD-L1 high expression, and SUR-L was the most reliable parameter. PET/CT can offer reference to screen patients for first-line atezolizumab therapy.