Suonita Khung, Abstract Co-Author: Nothing to Disclose

Pauline Masset, Abstract Co-Author: Nothing to Disclose

Jean-Baptiste Faivre MD, Abstract Co-Author: Nothing to Disclose

Nunzia Tacelli MD, Abstract Co-Author: Nothing to Disclose

Jacques Remy MD, Abstract Co-Author: Research Consultant, Siemens AG

Martine J. Remy-Jardin MD, PhD, Presenter: Research Grant, Siemens AG

To evaluate the diagnostic performance of virtually-rendered unfolded views of the ribs.

110 consecutive adult patients referred for polytrauma underwent a chest CT examination, retrospectively reviewed for specific detection of rib fractures according to two independent approaches: (a) analysis of transverse CT sections, completed with multiplanar reformats whenever deemed necessary by the reader (Group 1) ; (b) analysis of unfolded ribs as proposed by the software « CT Bone Reading » that generated a virtually-rendered unfolded view of the ribs and spine, with the possibility of rib analysis along their long axis and creation of standard orthogonal views in different orientations of any area suspected of fracture (Group 2). The gold standard for the diagnosis of rib fractures was established by the combined analysis of Group 1 and Group 2 images. Image analysis was obtained as follows: (a) separate reading of Group 1 and Group 2 images by two independent readers (a junior reader and a senior reader); (b) consensus analysis of Group 1 and Group 2 images by the two readers to establish the final diagnosis of rib fractures.

 From the gold standard analysis, 44 patients had rib fractures (mean number of fractures per patient: 2.85) with a total of 266 ribs fractured and a total of 314 fractures (222 undisplaced; 92 displaced). The “CT bone reading” provided a complete reconstruction of the whole ribcage in 94 patients (85.5%) and partially incomplete reconstructions (1- 5 ribs inadequately reconstructed) in 16 patients (14 .5%). The software performance was established as follows: (a) diagnosis of rib fracture (sensitivity: 0.84; specificity: 1); (b) number of ribs fractured (sensitivity: 0.77; specificity: 0.99); (c) number of displaced fractures (sensitivity: 0.92; specificity: 1). Group 2 analysis allowed detection of 38.6% of rib fractures missed in Group 1 and significantly reduced the junior reader’s reading time (p<0.0001). 

This software has the potential to help detect rib fractures in polytrauma patients.

The detection of rib fractures in a polytrauma patient, often difficult and time consuming, can be helped by the evaluated system.

Khung, S, Masset, P, Faivre, J, Tacelli, N, Remy, J, Remy-Jardin, M, Performance of Automated 3D-rendering of Ribs in Polytrauma Patients: Clinical Experience in 110 Patients.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14017237.html