Navraj Dhaliwal1, Navdeep Sahota1, Prosanta Mondal2, Ian Lutz3, Jordan Buchko3, Haron Obaid4. 1. Department of Medical Imaging, College of Medicine, Royal University Hospital, University of Saskatchewan, 103 Hospital Drive, SK, S7N 0W8, Saskatoon, Canada. 2. Clinical Research Support Unit (CRSU), College of Medicine, University of Saskatchewan, Saskatoon, Canada. 3. Division of Orthopedic Surgery, Department of Surgery, College of Medicine, University of Saskatchewan, Saskatoon, Canada. 4. Department of Medical Imaging, College of Medicine, Royal University Hospital, University of Saskatchewan, 103 Hospital Drive, SK, S7N 0W8, Saskatoon, Canada. Haron.Obaid@usask.ca.
Abstract
OBJECTIVE: To validate MRI for the quantification of the femoral neck version (FNV) using posterior lesser trochanteric Line (PLTL) and to compare reliability of the PLTL to the epicondylar and retrocondylar measurements. MATERIALS AND METHODS: A retrospective review of 3 T MRI scans performed for femoroacetabular impingement (FAI). Two musculoskeletal radiologists performed the measurements. MRI measurements of the PLTL were compared to CT using Bland Altman, Lin's concordance, and Lin's correlation coefficients. Interobserver reliability was determined using Bland Altman, Lin's concordance, and Lin's correlation coefficients. Intraobserver reliability was determined using Lin's concordance and Lin's correlation coefficients. RESULTS: Forty-five patients (90 lower extremities) met the inclusion criteria. Ages ranged from 20 to 41 years, with a mean of 31.5 years. There were 22 females and 23 males. Lin's concordance correlation coefficient for MRI and CT measurements of PLTL was substantial: 0.96 (95% CI: 0.94-0.98). PLTL Lin's correlation coefficient was 0.825 (95% CI 0.732-0.918) and indicated good interobserver reliability. Epicondylar and retrocondylar methods Lin's correlation coefficients demonstrated moderate interobserver reliability at 0.601 (0.415-0.786) and 0.632 (0.456-0.807), respectively. There was moderate 95% confidence interval overlap between the PLTL and the other measurements. Bland-Altman plots for each measurement were similar and demonstrated no bias. There was excellent intraobserver reliability (> 0.900) with significant 95% confidence interval overlap. CONCLUSION: MRI measurements of the PLTL are comparable to CT. The PLTL has good reliability between readers for the quantification of FNV using MRI, which could help avoid unnecessary radiation exposure using CT and reduce MRI scan time.
OBJECTIVE: To validate MRI for the quantification of the femoral neck version (FNV) using posterior lesser trochanteric Line (PLTL) and to compare reliability of the PLTL to the epicondylar and retrocondylar measurements. MATERIALS AND METHODS: A retrospective review of 3 T MRI scans performed for femoroacetabular impingement (FAI). Two musculoskeletal radiologists performed the measurements. MRI measurements of the PLTL were compared to CT using Bland Altman, Lin's concordance, and Lin's correlation coefficients. Interobserver reliability was determined using Bland Altman, Lin's concordance, and Lin's correlation coefficients. Intraobserver reliability was determined using Lin's concordance and Lin's correlation coefficients. RESULTS: Forty-five patients (90 lower extremities) met the inclusion criteria. Ages ranged from 20 to 41 years, with a mean of 31.5 years. There were 22 females and 23 males. Lin's concordance correlation coefficient for MRI and CT measurements of PLTL was substantial: 0.96 (95% CI: 0.94-0.98). PLTL Lin's correlation coefficient was 0.825 (95% CI 0.732-0.918) and indicated good interobserver reliability. Epicondylar and retrocondylar methods Lin's correlation coefficients demonstrated moderate interobserver reliability at 0.601 (0.415-0.786) and 0.632 (0.456-0.807), respectively. There was moderate 95% confidence interval overlap between the PLTL and the other measurements. Bland-Altman plots for each measurement were similar and demonstrated no bias. There was excellent intraobserver reliability (> 0.900) with significant 95% confidence interval overlap. CONCLUSION: MRI measurements of the PLTL are comparable to CT. The PLTL has good reliability between readers for the quantification of FNV using MRI, which could help avoid unnecessary radiation exposure using CT and reduce MRI scan time.