Rui Zheng1, Doug Hill2, Douglas Hedden1, Marc Moreau1, Sarah Southon1, Edmond Lou3,4. 1. Department of Surgery, University of Alberta, Edmonton, Canada. 2. Alberta Health Services, Edmonton, Canada. 3. Department of Surgery, University of Alberta, Edmonton, Canada. elou@ualberta.ca. 4. Department of Electrical and Computer Engineering, University of Alberta, 9211-116 Street, Edmonton, AB, T6G 1H9, Canada. elou@ualberta.ca.
Abstract
PURPOSE: To investigate the threshold of the curve difference on ultrasound measurement relative to the previous radiographic measurements to detect curves progression in children who have idiopathic scoliosis (IS). METHODS: Two hundred children with IS (F:170, M:30; mean age: 14.6 ± 1.9) were recruited from a single center. A retrospective study on comparing the current ultrasound measurements with the previous radiographic measurements with threshold values from 3° to 8° to detect curve progression was conducted. The receiver operating characteristic (ROC) analysis, accuracy (ACC), and odd ratio (OR) were calculated to determine the optimal threshold value of the curve differences between ultrasound and previous radiographic measurement. RESULTS: Both thresholds of 4° and 5° for curve difference from ultrasound scans presented the sensitivities ≥ 0.90 and specificities ≥ 0.85, and can reduce by 73 and 79% of radiographs on the studied subjects, respectively. Especially, for 4° threshold, the negative likelihood ratio (LR-) was only 0.08, which indicated that there is only 8% probability that the subject has progressed if the US measurement detected non-progression. CONCLUSIONS: The ultrasound imaging method can be applied to identify curve progression in children with IS. Four degree is the preferred threshold value to detect the curve which had progressed, since it also had the lower rate of undetected progressed cases (false negatives).
PURPOSE: To investigate the threshold of the curve difference on ultrasound measurement relative to the previous radiographic measurements to detect curves progression in children who have idiopathic scoliosis (IS). METHODS: Two hundred children with IS (F:170, M:30; mean age: 14.6 ± 1.9) were recruited from a single center. A retrospective study on comparing the current ultrasound measurements with the previous radiographic measurements with threshold values from 3° to 8° to detect curve progression was conducted. The receiver operating characteristic (ROC) analysis, accuracy (ACC), and odd ratio (OR) were calculated to determine the optimal threshold value of the curve differences between ultrasound and previous radiographic measurement. RESULTS: Both thresholds of 4° and 5° for curve difference from ultrasound scans presented the sensitivities ≥ 0.90 and specificities ≥ 0.85, and can reduce by 73 and 79% of radiographs on the studied subjects, respectively. Especially, for 4° threshold, the negative likelihood ratio (LR-) was only 0.08, which indicated that there is only 8% probability that the subject has progressed if the US measurement detected non-progression. CONCLUSIONS: The ultrasound imaging method can be applied to identify curve progression in children with IS. Four degree is the preferred threshold value to detect the curve which had progressed, since it also had the lower rate of undetected progressed cases (false negatives).
Entities:
Keywords:
Curve progression; Idiopathic scoliosis; Sensitivity and specificity; Threshold of curve progression; Ultrasonic imaging method
Authors: Rui Zheng; Amanda C Y Chan; Wei Chen; Douglas L Hill; Lawrence H Le; Douglas Hedden; Marc Moreau; James Mahood; Sarah Southon; Edmond Lou Journal: Spine Deform Date: 2015-03-04
Authors: Cécile M Ronckers; Michele M Doody; John E Lonstein; Marilyn Stovall; Charles E Land Journal: Cancer Epidemiol Biomarkers Prev Date: 2008-03 Impact factor: 4.254
Authors: Rui Zheng; Doug Hill; Douglas Hedden; James Mahood; Marc Moreau; Sarah Southon; Edmond Lou Journal: PLoS One Date: 2018-06-18 Impact factor: 3.240