Precision Imaging Ultrasound Technology - Does It Improve Accuracy And Increase Confidence In Diagnosing Breast Tumours?
Access status:
Open Access
Type
ThesisThesis type
Masters by ResearchAuthor/s
Lau, Siu Fong LouisaAbstract
Objective To determine the effect of Precision Imaging (PI), an innovative speckle reduction algorithm, on the diagnostic efficacy in breast ultrasound Material and methods Patients aged from 20 to 84 years screened by the breast clinic from October 2010 to June 2011 were included ...
See moreObjective To determine the effect of Precision Imaging (PI), an innovative speckle reduction algorithm, on the diagnostic efficacy in breast ultrasound Material and methods Patients aged from 20 to 84 years screened by the breast clinic from October 2010 to June 2011 were included in this research. The commercial ultrasound scanner Toshiba AplioMX, with compact linear transducers 15-7MHz and 12-5 MHz was used for image acquisition. A single projection image that was considered to best represent the lesion was recorded without PI (L0), then with all other 3 levels of PI, namely Precision 1 (L1), Precision 2 (L2) and Precision 3 (L3), with higher numbers signifying greater speckle reduction. Fifty one breast lesions (20 malignant and 31 benign) were selected from over 200 collected lesions, with selection criteria based on the 1- 5 classification system developed by National Breast Cancer Centre in collaboration with the Royal Australian and New Zealand College of Radiologists. These selected images were cropped to remove the technical details, which included patient information as well as PI level. These processed images were then organised into four sets (A,B,C,D) with images in same PI level. These four sets of images were evaluated by six radiologists and six sonographers dedicated to breast imaging, scoring each lesion between 1 and 6.These scores were subjected to Q-Perform software, DBMMRMC, Mann-Whitney U-test, Wilcoxon Signed rank test and IBM SPSS statistics for statistical analyses. Results The overall means ROCAUC for L0 was 0.79, L1 was 0.80, L2 was 0.81, and L3 was 0.81. The overall means sensitivity for L0 was 0.75, L1 was 0.79, L2 was 0.80, and L3 was 0.78.Overall means specificity for L0 was 0.74, L1 was 0.72, L2 was 0.73, and L3 was 0.71. Conclusion The data analysis on ROC, sensitivity, and specificity did not demonstrate any significant improvement in diagnostic efficacy amongst expert observers in this study.
See less
See moreObjective To determine the effect of Precision Imaging (PI), an innovative speckle reduction algorithm, on the diagnostic efficacy in breast ultrasound Material and methods Patients aged from 20 to 84 years screened by the breast clinic from October 2010 to June 2011 were included in this research. The commercial ultrasound scanner Toshiba AplioMX, with compact linear transducers 15-7MHz and 12-5 MHz was used for image acquisition. A single projection image that was considered to best represent the lesion was recorded without PI (L0), then with all other 3 levels of PI, namely Precision 1 (L1), Precision 2 (L2) and Precision 3 (L3), with higher numbers signifying greater speckle reduction. Fifty one breast lesions (20 malignant and 31 benign) were selected from over 200 collected lesions, with selection criteria based on the 1- 5 classification system developed by National Breast Cancer Centre in collaboration with the Royal Australian and New Zealand College of Radiologists. These selected images were cropped to remove the technical details, which included patient information as well as PI level. These processed images were then organised into four sets (A,B,C,D) with images in same PI level. These four sets of images were evaluated by six radiologists and six sonographers dedicated to breast imaging, scoring each lesion between 1 and 6.These scores were subjected to Q-Perform software, DBMMRMC, Mann-Whitney U-test, Wilcoxon Signed rank test and IBM SPSS statistics for statistical analyses. Results The overall means ROCAUC for L0 was 0.79, L1 was 0.80, L2 was 0.81, and L3 was 0.81. The overall means sensitivity for L0 was 0.75, L1 was 0.79, L2 was 0.80, and L3 was 0.78.Overall means specificity for L0 was 0.74, L1 was 0.72, L2 was 0.73, and L3 was 0.71. Conclusion The data analysis on ROC, sensitivity, and specificity did not demonstrate any significant improvement in diagnostic efficacy amongst expert observers in this study.
See less
Date
2015-07-15Faculty/School
Faculty of Health SciencesAwarding institution
The University of SydneyShare