-
Scientific Reports Dec 2021Rib fracture detection is time-consuming and demanding work for radiologists. This study aimed to introduce a novel rib fracture detection system based on deep learning...
Rib fracture detection is time-consuming and demanding work for radiologists. This study aimed to introduce a novel rib fracture detection system based on deep learning which can help radiologists to diagnose rib fractures in chest computer tomography (CT) images conveniently and accurately. A total of 1707 patients were included in this study from a single center. We developed a novel rib fracture detection system on chest CT using a three-step algorithm. According to the examination time, 1507, 100 and 100 patients were allocated to the training set, the validation set and the testing set, respectively. Free Response ROC analysis was performed to evaluate the sensitivity and false positivity of the deep learning algorithm. Precision, recall, F1-score, negative predictive value (NPV) and detection and diagnosis were selected as evaluation metrics to compare the diagnostic efficiency of this system with radiologists. The radiologist-only study was used as a benchmark and the radiologist-model collaboration study was evaluated to assess the model's clinical applicability. A total of 50,170,399 blocks (fracture blocks, 91,574; normal blocks, 50,078,825) were labelled for training. The F1-score of the Rib Fracture Detection System was 0.890 and the precision, recall and NPV values were 0.869, 0.913 and 0.969, respectively. By interacting with this detection system, the F1-score of the junior and the experienced radiologists had improved from 0.796 to 0.925 and 0.889 to 0.970, respectively; the recall scores had increased from 0.693 to 0.920 and 0.853 to 0.972, respectively. On average, the diagnosis time of radiologist assisted with this detection system was reduced by 65.3 s. The constructed Rib Fracture Detection System has a comparable performance with the experienced radiologist and is readily available to automatically detect rib fracture in the clinical setting with high efficacy, which could reduce diagnosis time and radiologists' workload in the clinical practice.
Topics: Adult; Aged; Aged, 80 and over; Algorithms; Deep Learning; Female; Humans; Male; Middle Aged; Neural Networks, Computer; Predictive Value of Tests; ROC Curve; Radiologists; Retrospective Studies; Rib Fractures; Sensitivity and Specificity; Tomography, X-Ray Computed; Young Adult
PubMed: 34873241
DOI: 10.1038/s41598-021-03002-7 -
European Journal of Trauma and... Feb 2022Rib fractures following thoracic trauma are frequently encountered injuries and associated with a significant morbidity and mortality. The aim of this study was to...
PURPOSE
Rib fractures following thoracic trauma are frequently encountered injuries and associated with a significant morbidity and mortality. The aim of this study was to provide current data on the epidemiology, in-hospital outcomes and 30-day mortality of rib fractures, and to evaluate these results for different subgroups.
METHODS
A nationwide retrospective cohort study was performed with the use of the Dutch Trauma Registry which covers 99% of the acutely admitted Dutch trauma population. All patients aged 18 years and older admitted to the hospital between January 2015 and December 2017 with one or more rib fractures were included. Incidence rates were calculated using demographic data from the Dutch Population Register. Subgroup analyses were performed for flail chest, polytrauma, primary thoracic trauma, and elderly patients.
RESULTS
A total of 14,850 patients were admitted between 2015 and 2017 with one or more rib fractures, which was 6.0% of all trauma patients. Of these, 573 (3.9%) patients had a flail chest, 4438 (29.9%) were polytrauma patients, 9273 (63.4%) were patients with primary thoracic trauma, and 6663 (44.9%) were elderly patients. The incidence rate of patients with rib fractures for the entire cohort was 29 per 100.000 person-years. The overall 30-day mortality was 6.9% (n = 1208) with higher rates observed in flail chest (11.9%), polytrauma (14.8%), and elderly patients (11.7%). The median hospital length of stay was 6 days (IQR, 3-11) and 37.3% were admitted to the intensive care unit (ICU).
CONCLUSIONS
Rib fractures are a relevant and frequently occurring problem among the trauma population. Subgroup analyses showed that there is a substantial heterogeneity among patients with rib fractures with considerable differences regarding the epidemiology, in-hospital outcomes, and 30-day mortality.
Topics: Adolescent; Aged; Flail Chest; Humans; Length of Stay; Netherlands; Retrospective Studies; Rib Fractures; Thoracic Injuries
PubMed: 32506373
DOI: 10.1007/s00068-020-01412-2 -
The Journal of Trauma and Acute Care... Dec 2022Pulmonary contusion has been considered a contraindication to surgical stabilization of rib fractures (SSRFs). This study aimed to evaluate the association between...
BACKGROUND
Pulmonary contusion has been considered a contraindication to surgical stabilization of rib fractures (SSRFs). This study aimed to evaluate the association between pulmonary contusion severity and outcomes after SSRF. We hypothesized that outcomes would be worse in patients who undergo SSRF compared with nonoperative management, in presence of varying severity of pulmonary contusions.
METHODS
This retrospective cohort study included adults with three or more displaced rib fractures or flail segment. Patients were divided into those who underwent SSRF versus those managed nonoperatively. Severity of pulmonary contusions was assessed using the Blunt Pulmonary Contusion 18 (BPC18) score. Outcomes (pneumonia, tracheostomy, mechanical ventilation days, intensive care unit (ICU) length of stay, hospital length of stay, mortality) were retrieved from patients' medical records. Comparisons were made using Fisher's exact and Kruskal-Wallis tests, and correction for potential confounding was done with regression analyses.
RESULTS
A total of 221 patients were included; SSRF was performed in 148 (67%). Demographics and chest injury patterns were similar in SSRF and nonoperatively managed patients. Surgical stabilization of rib fracture patients had less frequent head and abdominal/pelvic injuries ( p = 0.017 and p = 0.003). Higher BPC18 score was associated with worse outcomes in both groups. When adjusted for ISS, the ICU stay was shorter (adjusted β , -2.511 [95% confidence interval, -4.87 to -0.16]) in patients with mild contusions who underwent SSRF versus nonoperative patients. In patients with moderate contusions, those who underwent SSRF had fewer ventilator days (adjusted β , -5.19 [95% confidence interval, -10.2 to -0.17]). For severe pulmonary contusions, outcomes did not differ between SSRF and nonoperative management.
CONCLUSION
In patients with severe rib fracture patterns, higher BPC18 score is associated with worse respiratory outcomes and longer ICU and hospital admission duration. The presence of pulmonary contusions is not associated with worse SSRF outcomes, and SSRF is associated with better outcomes for patients with mild to moderate pulmonary contusions.
LEVEL OF EVIDENCE
Therapeutic/Care Management; Level IV.
Topics: Adult; Humans; Rib Fractures; Retrospective Studies; Flail Chest; Lung Injury; Contusions; Ribs; Length of Stay
PubMed: 36121283
DOI: 10.1097/TA.0000000000003790 -
Clinical Biomechanics (Bristol, Avon) Feb 2023The primary aim of this study was to determine and compare the biomechanical properties of a fractured or intact rib after implant fixation on an embalmed thorax.
BACKGROUND
The primary aim of this study was to determine and compare the biomechanical properties of a fractured or intact rib after implant fixation on an embalmed thorax.
METHODS
Five systems were fixated on the bilateral fractured or intact (randomly allocated) 6th to 10th rib of five post-mortem embalmed human specimens. Each rib underwent a four-point bending test to determine the bending structural stiffness (Newton per m), load to failure (Newton), failure mode, and the relative difference in bending structural stiffness and load to failure as compared to a non-fixated intact rib.
FINDINGS
As compared to a non-fixated intact rib, the relative difference in stiffness of a fixated intact rib ranged from -0.14 (standard deviation [SD], 0.10) to 0.53 (SD 0.35) and for a fixated fractured rib from -0.88 (SD 0.08) to 0.17 (SD 0.50). The most common failure mode was a new fracture at the most anterior drill hole for the plate and screw systems and a new fracture within the anterior portion of the implant for the clamping systems.
INTERPRETATION
The current fixation systems differ in their design, mode of action, and biomechanical properties. Differences in biomechanical properties such as stiffness and load to failure especially apply to fractured ribs. Insight in the differences between the systems might guide more specific implant selection and increase the surgeon's awareness for localizing hardware complaints or failure.
Topics: Humans; Biomechanical Phenomena; Rib Fractures; Bone Screws; Bone Plates; Ribs; Fracture Fixation; Fracture Fixation, Internal
PubMed: 36623327
DOI: 10.1016/j.clinbiomech.2023.105870 -
European Journal of Trauma and... Feb 2022There is missing knowledge about the association of obesity and mortality in patients with rib fractures. Since the global measure of obesity (body mass index [BMI]) is...
BACKGROUND
There is missing knowledge about the association of obesity and mortality in patients with rib fractures. Since the global measure of obesity (body mass index [BMI]) is often unknown in trauma patients, it would be convenient to use local computed tomography (CT)-based measures (e.g., umbilical outer abdominal fat) as a surrogate. The purpose of this study was to assess (1) whether local measures of obesity and rib fractures are associated with mortality and abdominal injuries and to evaluate (2) the correlation between local and global measures of obesity.
MATERIALS AND METHODS
A retrospective cohort study included all inpatients with rib fractures in 2013. The main exposure variable was the rib fracture score (RFS) (number of rib fractures, uni- or bilateral, age). Other exposure variables were CT-based measures of obesity and BMI. The primary outcome (endpoint) was in-hospital mortality. The secondary outcome consisted of abdominal injuries. Sex and comorbidities were adjusted for with logistic regression.
RESULTS
Two hundred and fifty-nine patients (median age 55.0 [IQR 44.0-72.0] years) were analyzed. Mortality was 8.5%. RFS > 4 was associated with 490% increased mortality (OR = 5.9, 95% CI 1.9-16.6, p = 0.002). CT-based measures and BMI were not associated with mortality, rib fractures or injury of the liver. CT-based measures of obesity showed moderate correlations with BMI (e.g., umbilical outer abdominal fat: r = 0.59, p < 0.001).
CONCLUSIONS
RFS > 4 was an independent risk factors for increased mortality. Local and global measures of obesity were not associated with mortality, rib fractures or liver injuries. If the BMI is not available in trauma patients, CT-based measures of obesity may be considered as a surrogate.
Topics: Abdominal Injuries; Humans; Injury Severity Score; Middle Aged; Obesity; Retrospective Studies; Rib Fractures; Tomography, X-Ray Computed
PubMed: 32892237
DOI: 10.1007/s00068-020-01483-1 -
Journal of Digital Imaging Oct 2023To develop a deep learning-based model for detecting rib fractures on chest X-Ray and to evaluate its performance based on a multicenter study. Chest digital radiography...
To develop a deep learning-based model for detecting rib fractures on chest X-Ray and to evaluate its performance based on a multicenter study. Chest digital radiography (DR) images from 18,631 subjects were used for the training, testing, and validation of the deep learning fracture detection model. We first built a pretrained model, a simple framework for contrastive learning of visual representations (simCLR), using contrastive learning with the training set. Then, simCLR was used as the backbone for a fully convolutional one-stage (FCOS) objective detection network to identify rib fractures from chest X-ray images. The detection performance of the network for four different types of rib fractures was evaluated using the testing set. A total of 127 images from Data-CZ and 109 images from Data-CH with the annotations for four types of rib fractures were used for evaluation. The results showed that for Data-CZ, the sensitivities of the detection model with no pretraining, pretrained ImageNet, and pretrained DR were 0.465, 0.735, and 0.822, respectively, and the average number of false positives per scan was five in all cases. For the Data-CH test set, the sensitivities of three different pretraining methods were 0.403, 0.655, and 0.748. In the identification of four fracture types, the detection model achieved the highest performance for displaced fractures, with sensitivities of 0.873 and 0.774 for the Data-CZ and Data-CH test sets, respectively, with 5 false positives per scan, followed by nondisplaced fractures, buckle fractures, and old fractures. A pretrained model can significantly improve the performance of the deep learning-based rib fracture detection based on X-ray images, which can reduce missed diagnoses and improve the diagnostic efficacy.
Topics: Humans; Rib Fractures; Tomography, X-Ray Computed; X-Rays; Radiography; Retrospective Studies
PubMed: 37407842
DOI: 10.1007/s10278-023-00868-z -
International Journal of Legal Medicine Sep 2022The following study was undertaken to determine if any specific occupant characteristics, crash factors, or associated injuries identified at autopsy could predict the...
The following study was undertaken to determine if any specific occupant characteristics, crash factors, or associated injuries identified at autopsy could predict the occurrence or number of fractured ribs in adults. Data were accrued from the Traffic Accident Reporting System (TARS) and coronial autopsy reports from Forensic Science SA, Adelaide, South Australia, from January 2000 to December 2020. A total of 1475 motor vehicle fatalities were recorded in TARS between January 2000 and December 2020, and 1082 coronial autopsy reports were identified that corresponded to TARS fatal crash data. After applying exclusion criteria involving missing data, 874 cases were included in the analysis. Of the 874 cases, 685 cases had one or more rib fractures. The leading cause of death for those with rib fractures was multiple trauma (54%), followed by head injury (17%) and chest injuries (10%). The strongest predictor of one or more rib fractures was increasing age (p < 0.001). Other factors found in the regression to be predictive of the number of rib fractures were the presence of a variety of other injuries including thoracic spinal fracture, lower right extremity fracture, splenic injury, liver injury, pelvic fracture, aortic injury, lung laceration, and hemothorax. Age is most likely associated with increasing rib fractures due to reduced tolerance to chest deflection with greater injuries occurring at lower magnitudes of impact. The association of other injuries with rib fractures may be a marker of higher impact severity crashes.
Topics: Accidents, Traffic; Adult; Humans; Multiple Trauma; Rib Fractures; Thoracic Injuries; Wounds, Nonpenetrating
PubMed: 35831760
DOI: 10.1007/s00414-022-02866-2 -
Clinical Biomechanics (Bristol, Avon) Dec 2020High rates of morbidity and mortality following flail chest rib fractures are well publicized. Standard of care has been supportive mechanical ventilation, but serious...
INTRODUCTION
High rates of morbidity and mortality following flail chest rib fractures are well publicized. Standard of care has been supportive mechanical ventilation, but serious complications have been reported. Internal rib fixation has shown improvements in pulmonary function, clinical outcomes, and decreased mortality. The goal of this study was to provide a model defining the biomechanical benefits of internal rib fixation.
METHODS
One human cadaver was prepared with an actuator providing anteroposterior forces to the thorax and rib motion sensors to define interfragmentary motion. Cadaveric model was validated using a prior study which defined costovertebral motion to create a protocol using similar technology and procedure. Ribs 4-6 were fixed with motion sensors anteriorly, laterally and posteriorly. Motion was recorded with ribs intact before osteotomizing each rib anteriorly and laterally. Flail chest motion was record with fractures subsequently plated and analyzed. Motion was recorded in the sagittal, coronal and transverse axes.
FINDINGS
Compared to the intact rib model, the flail chest model demonstrated an 11.3 times increase in sagittal plane motion, which was reduced to 2.1 times the intact model with rib plating. Coronal and sagittal plane models also saw increases of 9.7 and 5.1 times, respectively, with regards to flail chest motion. Both were reduced to 1.2 times the intact model after rib plating.
INTERPRETATION
This study allows quantification of altered ribcage biomechanics after flail chest injuries and suggests rib plating is useful in restoring biomechanics as well as contributing to improving pulmonary function and clinical outcomes.
Topics: Biomechanical Phenomena; Bone Plates; Cadaver; Flail Chest; Fracture Fixation, Internal; Humans; Mechanical Phenomena; Rib Fractures; Thoracic Injuries
PubMed: 33045492
DOI: 10.1016/j.clinbiomech.2020.105191 -
Medicine May 2021To evaluate the rib fracture detection performance in computed tomography (CT) images using a software based on a deep convolutional neural network (DCNN) and compare it...
To evaluate the rib fracture detection performance in computed tomography (CT) images using a software based on a deep convolutional neural network (DCNN) and compare it with the rib fracture diagnostic performance of doctors.We included CT images from 39 patients with thoracic injuries who underwent CT scans. In these images, 256 rib fractures were detected by two radiologists. This result was defined as the gold standard. The performances of rib fracture detection by the software and two interns were compared via the McNemar test and the jackknife alternative free-response receiver operating characteristic (JAFROC) analysis.The sensitivity of the DCNN software was significantly higher than those of both Intern A (0.645 vs 0.313; P < .001) and Intern B (0.645 vs 0.258; P < .001). Based on the JAFROC analysis, the differences in the figure-of-merits between the results obtained via the DCNN software and those by Interns A and B were 0.057 (95% confidence interval: -0.081, 0.195) and 0.071 (-0.082, 0.224), respectively. As the non-inferiority margin was set to -0.10, the DCNN software is non-inferior to the rib fracture detection performed by both interns.In the detection of rib fractures, detection by the DCNN software could be an alternative to the interpretation performed by doctors who do not have intensive training experience in image interpretation.
Topics: Adult; Aged; Aged, 80 and over; Deep Learning; Female; Humans; Male; Middle Aged; Predictive Value of Tests; ROC Curve; Radiographic Image Interpretation, Computer-Assisted; Retrospective Studies; Rib Fractures; Software; Tomography, X-Ray Computed; Young Adult
PubMed: 34011107
DOI: 10.1097/MD.0000000000026024 -
Pros and cons of rib unfolding software: a reliability and reproducibility study on trauma patients.Ulusal Travma Ve Acil Cerrahi Dergisi =... Jun 2023Examination of all 24 ribs on axial computed tomography (CT) slices might become a leeway and rib fractures (RF) may easily overlook in daily practice. Rib unfolding...
BACKGROUND
Examination of all 24 ribs on axial computed tomography (CT) slices might become a leeway and rib fractures (RF) may easily overlook in daily practice. Rib unfolding (RU), a computer-assisted software, that promises rapid assessment of the ribs in a two-dimensional plan, was developed to facilitate rib evaluation. We aimed to evaluate the reliability and reproducibility of RU software for RF detection on CT and to determine the accelerating effect to determine any drawback of RU application.
METHODS
Fifty-one patients with thoracic trauma formed the sample to be assessed by the observers. The characterization and distribution of RFs on CT images in this sample were recorded independently by the non-observers. Regarding the presence or ab-sence of RF, CT images were assessed blindedly by two radiologists with 5 years (observer-A) and 18 years (observer-B) of experience in thoracic radiology. Each observer assessed the axial CT and RU images on different days under non-observer supervision.
RESULTS
A total of 113 RFs were detected in 22 patients. The mean evaluation time for the axial CT images was 146.64 s for ob-server-A and 119.29 s for observer-B. The mean evaluation time for RU images was 66.44 s for observer-A and 32.66 s for observer-B. A statistically significant decrease was observed between the evaluation periods of observer-A and observer-B with RU software compared to the axial CT image assessment (p<0.001). The inter-observer κ value was 0.638, while the intra-observer results showed moderate (κ: 0.441) and good (κ: 0.752) reproducibility comparing the RU and axial CT assessments. Observer-A detected 47.05% non-displaced fractures, 48.93% minimally displaced (≤2 mm) fractures, and 38.77% displaced fractures on RU images (p=0.009). Ob-server-B detected 23.52% non-displaced fractures, 57.44% minimally displaced (≤2 mm) fractures, and 48.97% displaced fractures on RU images (p=0.045).
CONCLUSION
RU software accelerates fracture evaluation, while it has drawbacks including low sensitivity in fracture detection, false negativity, and underestimation of displacement.
Topics: Humans; Reproducibility of Results; Observer Variation; Ribs; Rib Fractures; Software
PubMed: 37278081
DOI: 10.14744/tjtes.2023.64359