E2123. Detection of Acute Traumatic Rib Fractures Using Maximal-Intensity: Projection (MIP) Images
Authors
Hamza Shogan;
MUN Faculty of Medicine
Claire Woodworth;
MUN Faculty of Medicine
Chigozie Chibuzo Nwosu;
MUN Faculty of Medicine
Lisa Smyth;
MUN Faculty of Medicine
Wesley Chan;
MUN Faculty of Medicine
Objective:
Rib fractures are the most common sequelae of blunt thoracic injury. The oblique orientation of most ribs often precludes visualization of whole ribs on a single thin-collimation axial slice on CT. As a result, fractures are potentially missed during interpretation. Maximal intensity projections (MIPs) may improve rib fracture detection because they represent thickened axial reconstructed sections with an emphasis on high-attenuation voxel structures. This study aimed to assess the diagnostic reliability and efficiency of MIP images compared to conventional axial CT images in detecting rib fractures following traumatic injury.
Materials and Methods:
This was a retrospective study. Two board-certified radiologists, one radiology resident, and one medical student independently analyzed axial and MIP images from 50 deidentified patients ordered by the emergency department following traumatic injury. Readers were timed and instructed to identify the location of fractures, type of fracture, and confidence of fracture detection. Reader data were compared against gold standard consensus established by agreement from two board-certified radiologists (one independent from the study) and a medical student after a 2-month period from reading the images.
Results:
There were a total of 226 fractures established by consensus. Analysis of MIPs show a trend of improved sensitivity (80.7% versus 74%) with similar specificity (97.1% versus 97.8%), positive predictive value (84.7% versus 87.1%), and negative predictive value (96.2% versus 94.9%) when compared to axial images respectively. Interobserver agreement was similar between MIP and axial images (k= 0.677 vs. 0.678, respectively). There was a trend that MIPs offer a 10% faster reading time on average among all readers and a 20% faster reading time in trainees (resident and medical student) compared to axial images.
Conclusion:
Our study demonstrates a trend of improved fracture detection and faster reading times when readers viewed MIPs compared to standard axial images. Our results also suggest that MIPs offer a more efficient way for trainees to detect rib fractures. Further research is needed to determine the influence of fracture location, type of fracture, and patient sex on detection. Clinically, using MIP images in blunt thoracic chest trauma may improve rib fracture detection and turnaround time.
