Kelly Spuur1, Jodi Webb2, Ann Poulos3, Sharon Nielsen4, Wayne Robinson5. 1. School of Dentistry and Health Sciences, Faculty of Science, Charles Sturt University, Locked Bag 588, Wagga Wagga, NSW 2678, Australia. Electronic address: kspuur@csu.edu.au. 2. School of Dentistry and Health Sciences, Faculty of Science, Charles Sturt University, Locked Bag 588, Wagga Wagga, NSW 2678, Australia. Electronic address: jodi.webb@health.nsw.gov.au. 3. Discipline of Medical Radiation Sciences, Faculty of Health Sciences, The University of Sydney Cumberland Campus, 75 East St, Lidcombe, NSW 2141, Australia. Electronic address: ann.poulos@sydney.edu.au. 4. Training & Research Office, Charles Sturt University and Sharon Nielsen Statistical Consulting, Boorooma Street, Locked Bag 588, Wagga Wagga, NSW 2678, Australia. Electronic address: sharon@snstats.com.au. 5. Research Office, Charles Sturt University, Boorooma Street, Locked Bag 588, Wagga Wagga, NSW 2678, Australia. Electronic address: wrobinson@csu.edu.
Abstract
AIM: The aim of this study is to determine the clinical rates of the demonstration of the inframammary angle (IMA) on the mediolateral oblique (MLO) view of the breast on digital mammograms and to compare the outcomes with current accreditation standards for compliance. Relationships between the IMA, age, the posterior nipple line (PNL) and compressed breast thickness will be identified and the study outcomes validated using appropriate analyses of inter-reader and inter-rater reliability and variability. Differences in left versus right data were also investigated. METHOD: A quantitative retrospective study of 2270 randomly selected paired digital mammograms performed by BreastScreen NSW was undertaken. Data was collected by direct measurement and visual analysis. Intra-class correlation analyses were used to evaluate inter- and intra-rater reliability. RESULTS: The IMA was demonstrated on 52.4% of individual and 42.6% of paired mammograms. A linear relationship was found between the posterior nipple line (PNL) and age (p-value <0.001). The PNL was predicted to increase by 0.48 mm for every one year increment in age. The odds of demonstrating the IMA reduced by 2% for every one year increase in age (p-value = 0.001); are 0.4% higher for every 1 mm increase in PNL (p-value = 0.001) and 1.6% lower for every 1 mm increase in compressed breast thickness, (p-value<0.001). There was high inter- and intra-rater reliability for the PNL while there was 100% agreement for the demonstration of the IMA. CONCLUSION: Analysis of the demonstration of the IMA indicates clinically achievable rates (42.6%) well below that required for compliance (50%-75%) to known worldwide accreditation standards for screening mammography. These standards should be aligned to the reported evidence base. Visualisation of the IMA is impacted negatively by increasing age and compressed breast thickness but positively by breast size (PNL).
AIM: The aim of this study is to determine the clinical rates of the demonstration of the inframammary angle (IMA) on the mediolateral oblique (MLO) view of the breast on digital mammograms and to compare the outcomes with current accreditation standards for compliance. Relationships between the IMA, age, the posterior nipple line (PNL) and compressed breast thickness will be identified and the study outcomes validated using appropriate analyses of inter-reader and inter-rater reliability and variability. Differences in left versus right data were also investigated. METHOD: A quantitative retrospective study of 2270 randomly selected paired digital mammograms performed by BreastScreen NSW was undertaken. Data was collected by direct measurement and visual analysis. Intra-class correlation analyses were used to evaluate inter- and intra-rater reliability. RESULTS: The IMA was demonstrated on 52.4% of individual and 42.6% of paired mammograms. A linear relationship was found between the posterior nipple line (PNL) and age (p-value <0.001). The PNL was predicted to increase by 0.48 mm for every one year increment in age. The odds of demonstrating the IMA reduced by 2% for every one year increase in age (p-value = 0.001); are 0.4% higher for every 1 mm increase in PNL (p-value = 0.001) and 1.6% lower for every 1 mm increase in compressed breast thickness, (p-value<0.001). There was high inter- and intra-rater reliability for the PNL while there was 100% agreement for the demonstration of the IMA. CONCLUSION: Analysis of the demonstration of the IMA indicates clinically achievable rates (42.6%) well below that required for compliance (50%-75%) to known worldwide accreditation standards for screening mammography. These standards should be aligned to the reported evidence base. Visualisation of the IMA is impacted negatively by increasing age and compressed breast thickness but positively by breast size (PNL).
Keywords:
Breast cancer; Breast imaging; Evidence based practice; Full field digital mammography FFDM; IMA; Image evaluation systems; Image quality; Inframammary angle; Inframammary fold; Mammography; Mediolateral oblique; PGMI