The ribs unfolded - a CT visualization algorithm for fast detection of rib fractures: effect on sensitivity and specificity in trauma patients.

OBJECTIVE

To assess a radiologist's detection rate of rib fractures in trauma CT when reading curved planar reformats (CPRs) of the ribs compared to reading standard MPRs.

METHODS

Two hundred and twenty trauma CTs (146 males, 74 females) were retrospectively subjected to a software algorithm to generate CPRs of the ribs. Patients were split into two equal groups. Sixteen patients were excluded due to insufficient segmentation, leaving 107 patients in group A and 97 patients in group B. Two radiologists independently evaluated group A using CPRs and group B using standard MPRs. Two different radiologists reviewed both groups with the inverse methods setting. Results were compared to a standard of reference created by two senior radiologists.

RESULTS

The reference standard identified 361 rib fractures in 61 patients. Reading CPRs showed a significantly higher overall sensitivity (P < 0.001) for fracture detection than reading standard MPRs, with 80.9% (584/722) and 71.5% (516/722), respectively. Mean reading time was significantly shorter for CPRs (31.3 s) compared to standard MPRs (60.7 s; P < 0.001).

CONCLUSION

Using CPRs for the detection of rib fractures accelerates the reading of trauma patient chest CTs, while offering an increased overall sensitivity compared to conventional standard MPRs.

KEY POINTS

• In major blunt trauma, rib fractures are diagnosed with Computed Tomography. • Image processing can unfold all ribs into a single plane. • Unfolded ribs can be read twice as fast as axial images. • Unfolding the ribs allows a more accurate diagnosis of rib fractures.