Performance of artificial intelligence in detecting bone fractures in radiographic results: A systematic literature review

Background: Artificial intelligence (AI), is currently very widely used in various lines of human life, including in the health sector. In radiology, the role of AI has begun to be involved in the interpretation of radiological imaging results with the aim of making radiological diagnosis more efficient. Purpose: To assess the performance of artificial intelligence in detecting human bone fractures on radiological imaging results. Method: A systematic literature review method using the PUBMED search application with a publication period of the last 5 years using search queries ("artificial intelligence" OR "machine learning" OR "deep learning") AND ("bone fracture detection" OR "fracture detection" OR "bone injury detection") AND ("X-ray" OR "radiograph") AND ("sensitivity" OR "diagnosis speed "OR" cost efficiency). Results: There are 27 articles showing that the use of AI in the field of radiology has been widely used in various countries, not only for X-ray radiology imaging, but also for CT Scan and MRI imaging applied to various fractures and its speed and effectiveness have also been compared for diagnosis using human radiology personnel. Conclusion: Artificial intelligence methods have a fairly good ability to identify human fractures in radiological images, which can help doctors avoid misdiagnosis and speed up the time to establish a diagnosis.