Assessment of a Deep Learning Algorithm for the Detection of Rib Fractures on Whole-Body Trauma Computed Tomography.
- Weikert T Clinic of Radiology and Nuclear Medicine, University Hospital Basel, University of Basel, Basel, Switzerland. thomas.weikert@usb.ch.
- Noordtzij LA Clinic of Radiology and Nuclear Medicine, University Hospital Basel, University of Basel, Basel, Switzerland.
- Bremerich J Clinic of Radiology and Nuclear Medicine, University Hospital Basel, University of Basel, Basel, Switzerland.
- Stieltjes B Clinic of Radiology and Nuclear Medicine, University Hospital Basel, University of Basel, Basel, Switzerland.
- Parmar V Clinic of Radiology and Nuclear Medicine, University Hospital Basel, University of Basel, Basel, Switzerland.
- Cyriac J Clinic of Radiology and Nuclear Medicine, University Hospital Basel, University of Basel, Basel, Switzerland.
- Sommer G Clinic of Radiology and Nuclear Medicine, University Hospital Basel, University of Basel, Basel, Switzerland.
- Sauter AW Clinic of Radiology and Nuclear Medicine, University Hospital Basel, University of Basel, Basel, Switzerland.
- 2020-06-12
Published in:
- Korean journal of radiology. - 2020
Computed tomography
Computer-assisted image interpretation
Deep learning
Rib fractures
Trauma
Adult
Aged
Deep Learning
Female
Humans
Image Interpretation, Computer-Assisted
Male
Middle Aged
Retrospective Studies
Rib Fractures
Tomography, X-Ray Computed
Whole Body Imaging
Wounds and Injuries
English
OBJECTIVE
To assess the diagnostic performance of a deep learning-based algorithm for automated detection of acute and chronic rib fractures on whole-body trauma CT.
MATERIALS AND METHODS
We retrospectively identified all whole-body trauma CT scans referred from the emergency department of our hospital from January to December 2018 (n = 511). Scans were categorized as positive (n = 159) or negative (n = 352) for rib fractures according to the clinically approved written CT reports, which served as the index test. The bone kernel series (1.5-mm slice thickness) served as an input for a detection prototype algorithm trained to detect both acute and chronic rib fractures based on a deep convolutional neural network. It had previously been trained on an independent sample from eight other institutions (n = 11455).
RESULTS
All CTs except one were successfully processed (510/511). The algorithm achieved a sensitivity of 87.4% and specificity of 91.5% on a per-examination level [per CT scan: rib fracture(s): yes/no]. There were 0.16 false-positives per examination (= 81/510). On a per-finding level, there were 587 true-positive findings (sensitivity: 65.7%) and 307 false-negatives. Furthermore, 97 true rib fractures were detected that were not mentioned in the written CT reports. A major factor associated with correct detection was displacement.
CONCLUSION
We found good performance of a deep learning-based prototype algorithm detecting rib fractures on trauma CT on a per-examination level at a low rate of false-positives per case. A potential area for clinical application is its use as a screening tool to avoid false-negative radiology reports.
To assess the diagnostic performance of a deep learning-based algorithm for automated detection of acute and chronic rib fractures on whole-body trauma CT.
MATERIALS AND METHODS
We retrospectively identified all whole-body trauma CT scans referred from the emergency department of our hospital from January to December 2018 (n = 511). Scans were categorized as positive (n = 159) or negative (n = 352) for rib fractures according to the clinically approved written CT reports, which served as the index test. The bone kernel series (1.5-mm slice thickness) served as an input for a detection prototype algorithm trained to detect both acute and chronic rib fractures based on a deep convolutional neural network. It had previously been trained on an independent sample from eight other institutions (n = 11455).
RESULTS
All CTs except one were successfully processed (510/511). The algorithm achieved a sensitivity of 87.4% and specificity of 91.5% on a per-examination level [per CT scan: rib fracture(s): yes/no]. There were 0.16 false-positives per examination (= 81/510). On a per-finding level, there were 587 true-positive findings (sensitivity: 65.7%) and 307 false-negatives. Furthermore, 97 true rib fractures were detected that were not mentioned in the written CT reports. A major factor associated with correct detection was displacement.
CONCLUSION
We found good performance of a deep learning-based prototype algorithm detecting rib fractures on trauma CT on a per-examination level at a low rate of false-positives per case. A potential area for clinical application is its use as a screening tool to avoid false-negative radiology reports.
- Language
-
- English
- Open access status
- hybrid
- Identifiers
-
- DOI 10.3348/kjr.2019.0653
- PMID 32524789
- Persistent URL
- https://sonar.ch/global/documents/5856
Statistics
Document views: 46
File downloads:
- Full-text: 0