Deep Learning - Methods to Amplify Epidemiological Data Collection and Analyses.

Deep learning is a subfield of artificial intelligence and machine learning based mostly on neural networks and often combined with attention algorithms that has been used to detect and identify objects in text, audio, images, and video. Serghiou and Rough (Am J Epidemiol. 0000;000(00):0000-0000) present a primer for epidemiologists on deep learning models. These models provide substantial opportunities for epidemiologists to expand and amplify their research in both data collection and analyses by increasing the geographic reach of studies, including more research subjects, and working with large or high dimensional data. The tools for implementing deep learning methods are not quite yet as straightforward or ubiquitous for epidemiologists as traditional regression methods found in standard statistical software, but there are exciting opportunities for interdisciplinary collaboration with deep learning experts, just as epidemiologists have with statisticians, healthcare providers, urban planners, and other professionals. Despite the novelty of these methods, epidemiological principles of assessing bias, study design, interpretation and others still apply when implementing deep learning methods or assessing the findings of studies that have used them.

Walking for transportation in large Latin American cities: walking-only trips and total walking events and their sociodemographic correlates.

Walking for transportation is a common and accessible means of achieving recommended physical activity levels, while providing important social and environmental co-benefits. Even though walking in rapidly growing urban areas has become especially challenging given the increasing dependence on motorised transportation, walking remains a major mode of transportation in Latin American cities. In this paper we aimed to quantify self-reported walking for transportation in Mexico City, Bogota, Santiago de Chile, Sao Paulo, and Buenos Aires, by identifying both walking trips that are conducted entirely on foot and walking events involved in trips mainly conducted on other means of transportation (e.g. private vehicle, public transit) among individuals ≥5-years old. We show how walking-only trips account for approximately 30% trips in the analysed cities, and we evidence how the pedestrian dimension of mobility is largely underestimated if walking that is incidental to other transportation modes is not accounted for: when considering all walking events, we observed an increase between 73% and 217% in daily walking time. As a result, we estimated that between 19% and 25% of residents in these cities meet the WHO physical activity guidelines solely from walking for transportation. The results of the study also suggest that the promotion of public transportation in large Latin American cities can especially help certain population groups achieve the daily recommended levels of physical activity, while among low-income groups accessibility and safety seem to be the key challenges to be addressed.

Fireworks type, injury pattern, and permanent impairment following severe fireworks-related injuries.

BACKGROUND: There is a paucity of clinical data on severe fireworks-related injuries, and the relationship between firework types, injury patterns, and magnitude of impairment is not well understood. Our objective was to describe the relationship between fireworks type, injury patterns, and impairment. METHODS: Retrospective case series (2005-2015) of patients who sustained consumer fireworks-related injuries requiring hospital admission and/or an operation at a Level 1 Trauma/Burn Center. Fireworks types, injury patterns (body region, injury type), operation, and permanent impairment were examined. RESULTS: Data from 294 patients 1 to 61years of age (mean 24years) were examined. The majority (90%) were male. 119 (40%) patients were admitted who did not undergo surgery, 163 (55%) patients required both admission and surgery, and 12 (5%) patients underwent outpatient surgery. The greatest proportion of injuries was related to shells/mortars (39%). There were proportionally more rocket injuries in children (44%), more homemade firework injuries in teens (34%), and more shell/mortar injuries in adults (86%). Brain, face, and hand injuries were disproportionately represented in the shells/mortars group. Seventy percent of globe-injured patients experienced partial or complete permanent vision loss. Thirty-seven percent of hand-injured patients required at least one partial or whole finger/hand amputation. The greatest proportion of eye and hand injuries resulting in permanent impairment was in the shells/mortars group, followed by homemade fireworks. Two patients died. CONCLUSIONS: Severe fireworks-related injuries from homemade fireworks and shells/mortars have specific injury patterns. Shells/mortars disproportionately cause permanent impairment from eye and hand injury.