No abstract

Building a fully autonomous self-driving system has been discussed for more than 20 years yet remains unsolved. Previous systems have limited ability to scale. Their localization subsystem needs labor-intensive map recording for running in a new area, and the accuracy decreases after the changes occur in the environment. In this paper, a new localization method is proposed to solve the scalability problems, with a new method for detecting and making sense of diverse traffic lines. Like the way human drives, a self-driving system should not rely on an exact position to travel in most scenarios. As a result, without HD Maps, GPS or IMU, the proposed localization subsystem relies only on detecting driving-related features around (like lane lines, stop lines, and merging lane lines). For spotting and reasoning all these features, a new line detector is proposed and tested against multiple datasets.

Background:Clostridioides difficile is a leading cause of nosocomial infectious diarrhea in developed countries, and it has a significant economic impact throughout the world. Early detection of the pathogen and its toxins is critical because early treatment significantly reduces infection-related morbidity, mortality, and medical cost. Surveillance of healthcare-associated infections (HAIs) is conducted using the NHSN standardized infection ratio (SIR). This metric allows comparison of a facility’s observed infection rate to a national benchmark. The SIR can be elevated due to both a lack of institutional criteria for stool submission and the use of highly sensitive but poorly specific testing as a standalone test for diagnosis. The SIR can be artificially elevated by inclusion of C difficile carriers rather than infected patients due to inappropriate testing and overly sensitive methods. We aimed to determine the impact of an institutional nursing-driven protocol for stool submission as well as 2-step testing on the SIR. Methods: Starting from the fourth quarter of 2018, we instituted a nursing protocol for initiation of C. difficile testing. If the patient had ≥3 soft, loose, or liquid stools in 24 hours within the first 3 days of admission, they were placed on contact precautions and an unformed stool sample was submitted for C. difficile nucleic acid amplification testing (NAAT). A positive result prompted further evaluation with a stool enzyme immunoassay toxin test for confirmation of active infection. From hospital day 4 onward, stricter criteria were implemented for testing for C. difficile infection. Data were extrapolated for calculation of a quarterly SIR. This value was then compared to retrospective SIR data from the first quarter of 2016 to the third quarter of 2018. Results: The quarterly total of hospital-onset C. difficile infections from the first quarter of 2016 to the third quarter of 2018 ranged from 24 to 39 incidents per quarter. After implementing the nursing-driven protocol and 2-step testing, the quarterly total of hospital onset C. difficile infections decreased to 5–6 per quarter. The SIR prior to initiation ranged from 0.66 to 1.37 and decreased to 0.306–0.386 after the nursing-driven protocol and 2-step testing were implemented. Conclusions: Implementation of both an institutional nursing-driven protocol for stool submission and a 2-step testing protocol reduced the number of quarterly hospital-onset C. difficile events as well as our facility’s quarterly SIR to below the national standard.Funding: NoneDisclosures: Ioana Chirca, University Hospital