Deep learning in fracture detection: a narrative review.
Pishtiwan H S Kalmet1, Sebastian Sanduleanu2, Sergey Primakov2
1Maastricht University Medical Center+, Department of Trauma Surgery, Maastricht.
Acta Orthopaedica
|January 14, 2020
Summary
Deep learning, a type of artificial intelligence (AI), shows promise for detecting fractures on radiographs and CT scans. This review explores AI applications in fracture detection and future directions.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Orthopaedics
Background:
- Artificial intelligence (AI) and deep learning (DL) have advanced machine perception.
- DL, a subset of AI, uses artificial neural networks and has gained significant traction.
- Fracture detection using DL on radiographs is emerging, but applications in CT scans are less explored.
Purpose of the Study:
- To review the application of deep learning in fracture detection on radiographs and CT scans.
- To discuss the value of deep learning in orthopaedics and traumatology.
- To outline future directions for deep learning in fracture detection.
Main Methods:
- Narrative review of existing literature.
- Analysis of deep learning techniques applied to medical imaging for fracture detection.
- Discussion of current and potential future applications.
Main Results:
- Deep learning has demonstrated capability in detecting fractures on radiographs.
- Limited studies currently exist on deep learning for fracture detection and classification in CT scans.
- The review synthesizes current knowledge and identifies research gaps.
Conclusions:
- Deep learning holds significant potential to enhance fracture detection accuracy and efficiency in orthopaedics.
- Further research is needed, particularly for computed tomography (CT) applications.
- Future directions include refining algorithms and expanding clinical validation.
