BACKGROUND:Fatigue measurement is conducted while the eyes are gazing at something to accomplish the determination and monitoring of a real-time and non-contact fatigue.OBJECTIVE:To figure out the percentage of eyelid closure (PERCLOS)value for fatigue and deduct the subject's fatigue state.DESIGN:To detect the subject's pupils with a reconstructed eye tracking system ,the video frequency of the pupils is collected, from which the PERCLOS value is worked out by image processing. The PERCLOS value is used to reflect the subject's fatigue state.SETTING:School of Mechanical and Electronic Engineering, Beijing Jiaotong University.MATERIALS :The experiment was conducted in the Ergonomics Lab from September 2003 to November 2004. The equipments used were video transmission cable, frame grab card and computer.METHODS:The video frequency images of pupil module sent from eye tracking system were processed to deduct the closure state of the pupil and figure out the PERCLOS value for the determination of fatigue.MAIN OUTCOME MEASURES :The images collected by image collecting card reflected the state of opening and closure of the pupil.RESULTS:A fatigue determination system which was able to determine effectively the examinee's fatigue state was established.CONCLUSION:This new method expands the application of eye tracking system which can be safely used off-line or on-line.

Objective:To investigate the incidence and risk factors of polymyxin B-associated acute kidney injury (AKI) in patients with severe infections caused by extensive drug resistance Gram negative bacteria (XDR-GNB)in intensive care unit (ICU).Methods:A retrospective study of adult patients with severe infection who received polymyxin B for more than 3 days in the department of critical care medicine of Nanjing Drum Tower Hospital Affiliated to Nanjing University Medical School from April 1st 2018 to January 31st 2020 were performed. AKI was diagnosed by Kidney Disease: Improving Global Outcomes (KDIGO) criteria. The baseline data, indicators during treatment period and prognostic factors were compared between AKI group and non-AKI group. Factors with statistically significant difference in univariate analysis and important clinical factors were included in the Logistic regression model to analyze the risk factors of AKI.Results:Seventy-two patients were treated with polymyxin B for more than 3 days. Forty-nine patients were finally enrolled, with 32 patients developing polymyxin B-associated AKI, and the incidence was 44.4%. The baseline data was balanced in AKI group and non-AKI group, and there was no significant difference in the prognosis [death or discharge without medial order (cases): 14 vs. 6, discharged for improvement (cases): 18 vs. 11, χ 2 = 0.329, P = 0.566]. Polymyxin B-associated AKI occurred from 1 day to 14 days after treatment, with an average of (6.8±3.8) days. Among the 32 AKI patients, 2 cases were lost to follow up after discharge, while renal function recovered in 18 cases and unrecovered in 12 cases. The prognosis of patients without recovery of renal function was significantly worse than that of patients with renal function recovery [death or discharge without medial order (cases): 12 vs. 2, discharged for improvement (cases): 0 vs. 16, P = 0.000]. Single factor analysis showed that daily dosage of polymyxin B in AKI group was higher than that in non-AKI group (mg: 151.6±23.7 vs. 132.4±30.3), numbers of patients with daily polymyxin B dose ≥ 150 mg, using vasoactive drugs, or severe hypoalbuminemia (albumin≤25 g/L) were higher than those in non-AKI group (cases: 29 vs. 10, 18 vs. 4, 9 vs. 0), with statistically significant differences (all P < 0.05). Multivariate Logistic regression analysis showed that daily dosage of polymyxin B ≥ 150 mg and use of vasoactive drugs were independent risk factors for polymyxin B-associated AKI [odds ratio ( OR) = 37.466, 95% confidence interval (95% CI) was 2.676-524.586, P = 0.007; OR = 22.960, 95% CI was 1.710-308.235, P = 0.018]. Conclusions:Comparing with non-AKI patients, more patients with polymyxin B-associated AKI had severe hypoalbuminemia, and the probability of using vasoactive drugs and the daily dose of polymyxin B were higher than non-AKI patients. Daily dose of polymyxin B ≥ 150 mg and using vasoactive drugs were independent risk factors for polymyxin B-associated AKI.