Using topic modeling to develop multi-level descriptions of naturalistic driving data from drivers with and without sleep apnea

Abstract

One challenge in using naturalistic driving data is producing a holistic analysis of these highly variable datasets. Typical analyses focus on isolated events, such as large g-force accelerations indicating a possible near-crash. Examining isolated events is ill-suited for identifying patterns in continuous activities such as maintaining vehicle control. We present an alternative approach that converts driving data into a text representation and uses topic modeling to identify patterns across the dataset. This approach enables the discovery of non-linear patterns, reduces the dimensionality of the data, and captures subtle variations in driver behavior. In this study topic models are used to concisely described patterns in trips from drivers with and without untreated obstructive sleep apnea (OSA). The analysis included 5000 trips (50 trips from 100 drivers; 66 drivers with OSA; 34 comparison drivers). Trips were treated as documents, and speed and acceleration data from the trips were converted to “driving words.” The identified patterns, called topics, were determined based on regularities in the co-occurrence of the driving words within the trips. This representation was used in random forest models to predict the driver condition (i.e., OSA or comparison) for each trip. Models with 10, 15 and 20 topics had better accuracy in predicting the driver condition, with a maximum AUC of 0.73 for a model with 20 topics. Trips from drivers with OSA were more likely to be defined by topics for smaller lateral accelerations at low speeds. The results demonstrate topic modeling as a useful tool for extracting meaningful information from naturalistic driving datasets.

Driving among individuals with chronic conditions: A systematic review of applied research using kinematic driving sensors

“…The 29 studies of kinematic driving data from drivers with chronic conditions can be broadly categorized as: (1) feasibility testing of kinematic driving data collection in the context of chronic conditions 13,25,26 ; (2) comparing standardized on‐road or simulation assessments with kinematic driving behavior in naturalistic settings 15,27,28 ; (3) characterizing driving behavior effects associated with chronic conditions 2,14,16,18–24,29–35 ; or (4) predicting symptoms or disease classification from kinematic driving data 17,36–40 . The median duration of driving data collection was 1.8 months (mode = 2 weeks).…”

“…10,11 In addition to studying the effects of chronic health conditions on driving behavior, these studies have explored kinematic driving data as a preclinical biomarker of chronic conditions. [12][13][14][15] These studies have considerable variance in their methods, populations, and findings suggesting a need for critical analysis and appraisal for feasibility and methodological guidelines. The goals of this review are to assess kinematic driving studies of adults with chronic conditions for study feasibility, characteristics, and key findings, to generate recommendations for future applications of kinematic driving data.…”

A systematic review on sensor-based driver behaviour studies: coherent taxonomy, motivations, challenges, recommendations, substantial analysis and future directions

“…Multiple sclerosis (MS) is a potential disease that disables the brain and spinal cord, and as such, Krasniuk et al (2019) undertook an on-road assessment that incorporated strategic driving maneuvers to understand the underlying differences between MS drivers who passed the driving maneuvers against those who failed them. Sleep apnea is a serious sleep disorder, and McLaurin et al (2018) used topic modeling to identify the effect of sleep apnea on driving, by comparing trips of drivers who suffer from obstructive sleep apnea against drivers who do not have it.…”

“…Researchers compare MS patients who succeeded versus those who failed on-road assessment in Krasniuk et al (2019). Similarly, sleep apnea, affects patients' alertness and performance, so researchers have used topic modeling to identify patterns from drivers with OSA during driving trips (McLaurin et al, 2018). More studies on how other diseases affect driving behaviour should be investigated as well to provide more insight for medical practitioners on how to find better treatments for their patients.…”

Nighttime features derived from topic models for classification of patients with COPD

“…A model-free unsupervised methodology that is gaining popularity in sensor data processing is the implementation of topic modeling to extract meaningful information from daily life data. Its applicability spans from the discovery of activity routines [17,21], human activity recognition [22,23] to behaviours analysis [24,25]. In this study, multimodal sensor data recorded continuously by an activity monitor during nighttime are initially transformed into artificial "words" used to create a corpus of text documents in which each document represents the night of a subject.…”