The study, involving 500 patients, showed that artificial intelligence (AI) assistance enhanced fracture diagnosis on radiographs and reduced reading time for radiologists of varying experience levels.
Researchers have noted that traumatic fractures are among the most commonly missed diagnoses.1,2 However, a new study suggests that artificial intelligence (AI) may have significant benefit in improving the assessment of fractures.3
In the study of 500 patients (268 men and 232 women), researchers compared unassisted assessment of acute fractures versus assessment with the assistance of an FDA-cleared algorithm (Boneview®, Gleamer) and stand-alone use of AI. The authors found that AI assisted assessment had a 20 percent higher sensitivity (86 percent) of diagnosing fractures on radiographs in comparison to unassisted assessment (66 percent).
The use of AI assistance led to a lower number of false negatives (26) in comparison to unassisted radiograph assessment (64), according to the study. The researchers also noted that AI assistance reduced image reading time by an average of 12.43 seconds.
“Our study confirms that AI significantly increases radiologists’ overall performance and productivity in fracture diagnosis in a representative sample of daily activity in a trauma emergency department,” wrote Sebastien Aubry, MD, PhD, the chief of radiology at CHU de Besancon in Besancon, France, and colleagues.
The three radiologist reviewers in the study included a radiologist with 15 years of experience in musculoskeletal imaging, a fellow with two years of experience in musculoskeletal imaging and a third-year resident. Artificial intelligence assistance resulted in 17, 21 and 19 percent improvements in sensitivity, respectively, for the senior radiologist, fellow and third-year resident, according to the study.
While patients with head radiographs were excluded from the study, Aubry and colleagues noted significant differences in AI-enabled assessment for all body parts imaged in the study. They found the most significant differences in sensitivity between AI-assisted assessment and unassisted assessment were in the spine (77.8 percent versus 47.2 percent) and pelvis (83.3 percent versus 55.6 percent).
The mean age of the study patients was 37. Aubry and colleagues said the use of AI led to a reduction of misdiagnosed fractures in every age group with the exception of patients between the ages of 20-30.
In regard to study limitations, the authors acknowledged the possibilities of order, recall and context biases due to the lack of clinical information supplied to the three reviewing radiologists and the fact that those radiologists reviewed the same radiographs with and without AI assistance. Aubry and colleagues also noted a lower number of reviewing radiologists in comparison to other studies that have examined the use of AI for fracture detection.
References
1. Hussain F, Cooper A, Carson-Stevens A, et al. Diagnostic error in the emergency department: learning from national patient safety incident report analysis. BMC Emerg Med. 2019;19(1):77.
2. Whang JS, Baker SR, Patel R, Luk L, Castro A. The causes of medical malpractice suits against radiologists in the United States. Radiology. 2013;266(2): 548-554.
3. Canoni-Meynet L, Verdot P, Danner A, Calame P, Aubry S. Added value of an artificial intelligence solution for fracture detection in the radiologist’s daily trauma emergencies workflow. Diagn Interv Imaging. 2022 Jun 29;S2211-5684(22)00115-2. doi: 10.1016/j.diii.2022.06.004. Online ahead of print.
AI Algorithm Comparable to Radiologists in Differentiating Small Renal Masses on CT
May 14th 2024An emerging deep learning algorithm had a lower AUC and sensitivity than urological radiologists for differentiating between small renal masses on computed tomography (CT) scans but had a 21 percent higher sensitivity rate than non-urological radiologists, according to new research.
MRI-Based Deep Learning Algorithm Shows Comparable Detection of csPCa to Radiologists
May 8th 2024In a study involving over 1,000 visible prostate lesions on biparametric MRI, a deep learning algorithm detected 96 percent of clinically significant prostate cancer (csPCa) in comparison to a 98 percent detection rate for an expert genitourinary radiologist.
FDA Clears AI-Powered Qualitative Perfusion Mapping for Cone-Beam CT
May 6th 2024Reportedly validated in more than 10 clinical trials, the AngioFlow perfusion imaging software enables timely identification of brain regions with cerebral blood flow reduction and those with significant hypoperfusion.