Review of machine learning for optical imaging of burn wound severity assessment.

Significance: Over the past decade, machine learning (ML) algorithms have rapidly become much more widespread for numerous biomedical applications, including the diagnosis and categorization of disease and injury. Aim: Here, we seek to characterize the recent growth of ML techniques that use imaging data to classify burn wound severity and report on the accuracies of different approaches. Approach: To this end, we present a comprehensive literature review of preclinical and clinical studies using ML techniques to classify the severity of burn wounds. Results: The majority of these reports used digital color photographs as input data to the classification algorithms, but recently there has been an increasing prevalence of the use of ML approaches using input data from more advanced optical imaging modalities (e.g., multispectral and hyperspectral imaging, optical coherence tomography), in addition to multimodal techniques. The classification accuracy of the different methods is reported; it typically ranges from ∼70% to 90% relative to the current gold standard of clinical judgment. Conclusions: The field would benefit from systematic analysis of the effects of different input data modalities, training/testing sets, and ML classifiers on the reported accuracy. Despite this current limitation, ML-based algorithms show significant promise for assisting in objectively classifying burn wound severity.

Engaging Underrepresented Youth: Remote Transition of the Health Career Collaborative Program.

INTRODUCTION: Pipeline programs can help increase diversity in health care by engaging underrepresented minority groups to pursue higher education and training in medical fields. Here we describe the implementation of Health Career Collaborative, a pipeline program designed to connect high school students with health care professionals, and the transition to remote delivery of the curriculum. METHODS: This study is a retrospective, descriptive observational study where the baseline characteristics of participating students were evaluated via preparticipation surveys. This study took place in a community with an area deprivation index of 6 at a high school in southern California in conjunction with an academic medical center and level I trauma center. Due to the coronavirus disease 2019 pandemic, the program transitioned to a virtual setting in the second half of the academic year. RESULTS: A total of 37 high school student participants enrolled in the 2019-2020 Health Career Collaborative program, with over 97% identifying as Hispanic, 89% female, and 92% between the ages of 15 and 17. Ninety-five percent of students indicated plans to graduate from high school and attend college, and 89% agreed with having a mentor to help plan for their future. While high school students had exposure to several health topics prior to the program, students reported a preference to learn about health topics from doctors compared to other sources. CONCLUSIONS: An online platform helped facilitate more interaction with health care professionals and could improve feasibility of implementing pipeline programs because physical space and transportation are not required.

Temporal trends of postinjury multiple-organ failure: still resource intensive, morbid, and lethal.

BACKGROUND: While the incidence of postinjury multiple-organ failure (MOF) has declined during the past decade, temporal trends of its morbidity, mortality, presentation patterns, and health care resources use have been inconsistent. The purpose of this study was to describe the evolving epidemiology of postinjury MOF from 2003 to 2010 in multiple trauma centers sharing standard treatment protocols. METHODS: "Inflammation and Host Response to Injury Collaborative Program" institutions that enrolled more than 20 eligible patients per biennial during the 2003 to 2010 study period were included. The patients were aged 16 years to 90 years, sustained blunt torso trauma with hemorrhagic shock (systolic blood pressure < 90 mm Hg, base deficit ≥ 6 mEq/L, blood transfusion within the first 12 hours), but without severe head injury (motor Glasgow Coma Scale [GCS] score < 4). MOF temporal trends (Denver MOF score > 3) were adjusted for admission risk factors (age, sex, body max index, Injury Severity Score [ISS], systolic blood pressure, and base deficit) using survival analysis. RESULTS: A total of 1,643 patients from four institutions were evaluated. MOF incidence decreased over time (from 17% in 2003-2004 to 9.8% in 2009-2010). MOF-related death rate (33% in 2003-2004 to 36% in 2009-2010), intensive care unit stay, and mechanical ventilation duration did not change over the study period. Adjustment for admission risk factors confirmed the crude trends. MOF patients required much longer ventilation and intensive care unit stay, compared with non-MOF patients. Most of the MOF-related deaths occurred within 2 days of the MOF diagnosis. Lung and cardiac dysfunctions became less frequent (57.6% to 50.8%, 20.9% to 12.5%, respectively), but kidney and liver failure rates did not change (10.1% to 12.5%, 15.2% to 14.1%). CONCLUSION: Postinjury MOF remains a resource-intensive, morbid, and lethal condition. Lung injury is an enduring challenge and should be a research priority. The lack of outcome improvements suggests that reversing MOF is difficult and prevention is still the best strategy. LEVEL OF EVIDENCE: Epidemiologic study, level III.