Leveraging driver vehicle and environment interaction: Machine learning using driver monitoring cameras to detect drunk driving
Authors:
Kevin Koch,
Martin Maritsch,
Eva van Weenen,
Stefan Feuerriegel,
Matthias Pfäffli,
Elgar Fleisch,
Wolfgang Weinmann,
Felix Wortmann
Abstract:
Excessive alcohol consumption causes disability and death. Digital interventions are promising means to promote behavioral change and thus prevent alcohol-related harm, especially in critical moments such as driving. This requires real-time information on a person's blood alcohol concentration (BAC). Here, we develop an in-vehicle machine learning system to predict critical BAC levels. Our system…
▽ More
Excessive alcohol consumption causes disability and death. Digital interventions are promising means to promote behavioral change and thus prevent alcohol-related harm, especially in critical moments such as driving. This requires real-time information on a person's blood alcohol concentration (BAC). Here, we develop an in-vehicle machine learning system to predict critical BAC levels. Our system leverages driver monitoring cameras mandated in numerous countries worldwide. We evaluate our system with n=30 participants in an interventional simulator study. Our system reliably detects driving under any alcohol influence (area under the receiver operating characteristic curve [AUROC] 0.88) and driving above the WHO recommended limit of 0.05g/dL BAC (AUROC 0.79). Model inspection reveals reliance on pathophysiological effects associated with alcohol consumption. To our knowledge, we are the first to rigorously evaluate the use of driver monitoring cameras for detecting drunk driving. Our results highlight the potential of driver monitoring cameras and enable next-generation drunk driver interaction preventing alcohol-related harm.
△ Less
Submitted 22 March, 2023; v1 submitted 21 January, 2023;
originally announced January 2023.
Estimating Risk-Adjusted Hospital Performance
Authors:
Eva van Weenen,
Stefan Feuerriegel
Abstract:
The quality of healthcare provided by hospitals is subject to considerable variability. Consequently, accurate measurements of hospital performance are essential for various decision-makers, including patients, hospital managers and health insurers. Hospital performance is assessed via the health outcomes of their patients. However, as the risk profiles of patients between hospitals vary, measurin…
▽ More
The quality of healthcare provided by hospitals is subject to considerable variability. Consequently, accurate measurements of hospital performance are essential for various decision-makers, including patients, hospital managers and health insurers. Hospital performance is assessed via the health outcomes of their patients. However, as the risk profiles of patients between hospitals vary, measuring hospital performance requires adjustment for patient risk. This task is formalized in the state-of-the-art procedure through a hierarchical generalized linear model, that isolates hospital fixed-effects from the effect of patient risk on health outcomes. Due to the linear nature of this approach, any non-linear relations or interaction terms between risk variables are neglected.
In this work, we propose a novel method for measuring hospital performance adjusted for patient risk. This method captures non-linear relationships as well as interactions among patient risk variables, specifically the effect of co-occurring health conditions on health outcomes. For this purpose, we develop a tailored neural network architecture that is partially interpretable: a non-linear part is used to encode risk factors, while a linear structure models hospital fixed-effects, such that the risk-adjusted hospital performance can be estimated. We base our evaluation on more than 13 million patient admissions across almost 1,900 US hospitals as provided by the Nationwide Readmissions Database. Our model improves the ROC-AUC over the state-of-the-art by 4.1 percent. These findings demonstrate that a large portion of the variance in health outcomes can be attributed to non-linear relationships between patient risk variables and implicate that the current approach of measuring hospital performance should be expanded.
△ Less
Submitted 13 November, 2020; v1 submitted 10 November, 2020;
originally announced November 2020.
