Automated fracture screening using an object detection algorithm on whole-body trauma computed tomography

Abstract

The emergency department is an environment with a potential risk for diagnostic errors during trauma care, particularly for fractures. Convolutional neural network (CNN) deep learning methods are now widely used in medicine because they improve diagnostic accuracy, decrease misinterpretation, and improve efficiency. In this study, we investigated whether automatic localization and classification using CNN could be applied to pelvic, rib, and spine fractures. We also examined whether this fracture detection algorithm could help physicians in fracture diagnosis. A total of 7664 whole-body CT axial slices (chest, abdomen, pelvis) from 200 patients were used. Sensitivity, precision, and F1-score were calculated to evaluate the performance of the CNN model. For the grouped mean values for pelvic, spine, or rib fractures, the sensitivity was 0.786, precision was 0.648, and F1-score was 0.711. Moreover, with CNN model assistance, surgeons showed improved sensitivity for detecting fractures and the time of reading and interpreting CT scans was reduced, especially for less experienced orthopedic surgeons. Application of the CNN model may lead to reductions in missed fractures from whole-body CT images and to faster workflows and improved patient care through efficient diagnosis in polytrauma patients.

Artificial Intelligence in Emergency Trauma Care: A Preliminary Scoping Review

“… Salient features contributing to prediction include vital sign trends, PPG perfusion index, and ECG waveforms, highlighting the potential for continuous application of this approach to improve triage and predict clinical deterioration in apparently stable patients. Treatment [ 21 ] Takaki, Inoue, Maki, Furuya, Mikami, Mizutani, Takada, Okimatsu, Yunde, Miura, Shiratani, Nagashima, Maruyama, Shiga, Inage, Orita, Eguchi, Ohtori Automated fracture screening using an object detection algorithm on whole-body trauma computed tomography. Retrospective Observational Japan Convolutional neural network (CNN) deep learning methods were investigated for automatic localization and classification of pelvic, rib, and spine fractures in trauma care.…”

“…Algorithmic fracture nominations focus radiologists’ attention to suspicious areas, improving detection rates by 65.7% and reducing reading times. 21 Minimizing false positives and negatives remains an open challenge. 19 , 32 …”

Identification of Location and Geometry of Invisible Internal Defects in Structures using Deep Learning and Surface Deformation Field

“…Although several researchers have explored crack detection in structures such as roads, [ 40 ] aeroengine blades, [ 41 ] pipelines, [ 42 ] bridges, [ 43 ] tunnels, [ 44 ] and human limbs, [ 45 ] these studies have been limited to surface cracks. The proposed method is an advancement over existing methods in that it can identify invisible subsurface defects and reveal the geometric information at a low cost for real‐time monitoring of structures in their natural states.…”

Cervical spine fracture detection in computed tomography using convolutional neural networks

“…Salehinejad et al use a ResNet-50 network (Salehinejad et al 2021) for an axial slice wise classification (fracture/no fracture). Also using a 2D approach, Inoue et al (2022) applied in a recent paper a R-CNN network (Ren et al 2015) to a general automated fracture screening in CT images.…”

“…Case-wise cervical detection results for the VerSe2019/2020 data. Cases are classifiedInoue et al (2022) presents also an evaluation at fracture level, reporting for spine fractures a sensitivity of 78% at an average rate of 4.9 false positives per image.The two publications on the solitary commercially available AI solution(Small et al 2021, Voter et al 2021) present only results for a per image classification task into fracture/non-fracture cases. In Small et al (2021) a set of 665 cases was examined.…”