Automatic classification of distal radius fracture using a two-stage ensemble deep learning framework

Distal radius fractures (DRFs) are one of the most common types of wrist fracture and can be subdivided into intra- and extra-articular fractures. Compared with extra-articular DRFs which spare the joint surface, intra-articular DRFs extend to the articular surface and can be more difficult to treat. Identification of articular involvement can provide valuable information about the characteristics of fracture patterns. In this study, a two-stage ensemble deep learning framework was proposed to differentiate intra- and extra-articular DRFs automatically on posteroanterior (PA) view wrist X-rays. The framework firstly detects the distal radius region of interest (ROI) using an ensemble model of YOLOv5 networks, which imitates the clinicians' search pattern of zooming in on relevant regions to assess abnormalities. Secondly, an ensemble model of EfficientNet-B3 networks classifies the fractures in the detected ROIs into intra- and extra-articular. The framework achieved an area under the receiver operating characteristic curve of 0.82, an accuracy of 0.81, a true positive rate of 0.83 and a false positive rate of 0.27 (specificity of 0.73) for differentiating intra- from extra-articular DRFs. This study has demonstrated the potential in automatic DRF characterization using deep learning on clinically acquired wrist radiographs and can serve as a baseline for further research in incorporating multi-view information for fracture classification.

1061 Problems During Daytime Due To Poor Sleep Are Associated With Indicators Of Poor Dental Health In A Largely Non-smoking Population

Introduction

Growing evidence connects periodontal disease, a major and modifiable cause of local and systemic inflammation, with metabolic and cardiovascular morbidity and mortality, as well as mental illness. Sleep has been previously predictively associated with metabolic and psychiatric morbidity and mortality, and has recently been linked with periodontal disease. We are now evaluating associations between self-reported insomnia measures and surrogate indicators of periodontal disease in a population with a very low prevalence of smoking — a major confounder in previous studies.

Methods

Dental and sleep questionnaires (Pittsburgh Sleep Quality Index) were obtained from 3881 Old Order Amish from Lancaster county. Difficulty falling or staying asleep, sleep quality, problems during the day due to poor sleep, and sleepiness during daytime were related to self-reported loose teeth, partial dentures, full dentures and any dentures using linear models with adjustment for age and sex.

Results

Significant associations emerged between problems falling asleep and loose teeth (p<0.05), problems staying asleep and any dentures (p<0.05), sleep quality with loose teeth and partial, as well as complete dentures (p<0.05 for both). Problems during daytime due to poor sleep were associated with loose teeth (p<0.05), any dentures (p<0.003) and full dentures (p<0.0001 — the only associations resisting Bonferroni correction). Sleepiness during daytime, which is the most important marker associated with sleep apnea was not associated with any dental health measures.

Conclusion

Limitations include not accounting for family aggregation, limited generalizability, not fully differentiating between respiratory versus non-respiratory sleep impairment, and periodontal versus traumatic dental pathology. Yet, the results of our study, which minimizes the strong potential confounding by smoking, confirm associations between sleep-related problems and periodontal disease, and justify future longitudinal and interventional research to dissect causality and identify multi-target treatment modalities.

Support

This work was supported by the Mid-Atlantic Nutrition Obesity Research Center Preliminary Developmental NORC grant (Postolache, PI), a sub-award of the parent grant P30 DK072488 (Mitchell, PI), and intramural funds from the University of Maryland, Baltimore.