BACKGROUND: Magnetic resonance imaging has become the method of choice to diagnose meniscal and ligament lesions of the knee. New approaches to the surgical treatment of partial anterior cruciate ligament tears led us to try to distinguish the two bundles of that ligament with MRI and to evaluate inter-observer agreement in different viewing planes. METHODS: Images of 50 right and 50 left ligament-intact knees were examined in the coronal and axial viewing planes. Each sequence was independently read by a radiologist and a medical student to note, in each viewing plane, the number of images in which the two-bundle structure of the ligament was clearly seen. Cohen's Kappa coefficients were used to determine inter-observer agreements. RESULTS: The percentage of sequences in which the two bundles were distinguished by the radiologist and the student on one image at least were 82 versus 73% in the coronal plane and 90 versus 93% in the axial plane, respectively. The average number of successive images with clear bundles was higher in the axial (2.7 vs. 1.7) than in the coronal plane (2.2 vs. 1). There was a poor inter-observer agreement in the coronal plane (k=0.176) but an intermediate agreement in the axial plane (k=0.385 or 0.522 depending on number pooling). CONCLUSIONS: The axial viewing plane seems more favourable to distinguish the two bundles of the anterior cruciate ligament with the best possible reproducibility. An oblique-axial plane is deemed to insure a clearer diagnosis of partial tears.
BACKGROUND: Magnetic resonance imaging has become the method of choice to diagnose meniscal and ligament lesions of the knee. New approaches to the surgical treatment of partial anterior cruciate ligament tears led us to try to distinguish the two bundles of that ligament with MRI and to evaluate inter-observer agreement in different viewing planes. METHODS: Images of 50 right and 50 left ligament-intact knees were examined in the coronal and axial viewing planes. Each sequence was independently read by a radiologist and a medical student to note, in each viewing plane, the number of images in which the two-bundle structure of the ligament was clearly seen. Cohen's Kappa coefficients were used to determine inter-observer agreements. RESULTS: The percentage of sequences in which the two bundles were distinguished by the radiologist and the student on one image at least were 82 versus 73% in the coronal plane and 90 versus 93% in the axial plane, respectively. The average number of successive images with clear bundles was higher in the axial (2.7 vs. 1.7) than in the coronal plane (2.2 vs. 1). There was a poor inter-observer agreement in the coronal plane (k=0.176) but an intermediate agreement in the axial plane (k=0.385 or 0.522 depending on number pooling). CONCLUSIONS: The axial viewing plane seems more favourable to distinguish the two bundles of the anterior cruciate ligament with the best possible reproducibility. An oblique-axial plane is deemed to insure a clearer diagnosis of partial tears.
Authors: James S Starman; Corinne Vanbeek; Derek R Armfield; Amit Sahasrabudhe; Champ L Baker; James J Irrgang; Freddie H Fu Journal: Knee Surg Sports Traumatol Arthrosc Date: 2007-01-16 Impact factor: 4.342
Authors: Carlos Eduardo Sanches Vaz; Olavo Pires de Camargo; Paulo José de Santana; Antonio Carlos Valezi Journal: Clinics (Sao Paulo) Date: 2005-12-12 Impact factor: 2.365
Authors: Fábio Cavalli; Anela Izadi; Ana Paula R B Ferreira; Larissa Braga; Andresa Braga-Baiak; Marco Antonio Schueda; Mihir Gandhi; Ricardo Pietrobon Journal: Adv Orthop Date: 2011-07-07