Presentation is loading. Please wait.

Presentation is loading. Please wait.

Rachel Embree and Christina Sillery David Channin, MD, Advisor

Similar presentations


Presentation on theme: "Rachel Embree and Christina Sillery David Channin, MD, Advisor"— Presentation transcript:

1 Automatic Volumetric Breast Density Assessment in the eXtensible Imaging Platform
Rachel Embree and Christina Sillery David Channin, MD, Advisor In Cooperation with Alex Shnayder, Pat Mongkolwat, Ray Wu

2 Overview Introduction Material and Methods Results Discussion
Conclusion

3 Introduction

4 What is Breast Density? The breast is composed of fibroglandular tissue embedded in a background of fatty tissue. The amount of fibroglandular tissue and fat varies among women. Breast cancer arises in this fibroglandular tissue. Mammography is projection radiography of the breast. In mammograms, dense fibroglandular tissue attenuates more radiation than does fat. Breast density refers to the appearance of this fibroglandular tissue in a mammogram [5] - Different b/c of differences in the relative amounts and X-ray attenuation characteristics of fat and fibroglandular tissue - Fat - radiolucent - Fibroglandular - radiodense

5 Breast Density Positively correlated with breast cancer
Four to six times greater risk of breast cancer for women with greater than 60% dense tissue [1] Breast density as a risk factor accounts for as many as 30% of breast cancer cases [1] Can be changed by hormonal, dietary and other interventions

6 Assessment of Breast Density in Mammograms
Subjective Visual Assessment by Radiologists (Mandatory) Score the breast on a 4-point scale in the Breast Imaging-Reporting and Data System (BIRADS), based on patterns developed by Wolfe [5] Manual Image Processing Techniques (for research) Planimetry Interactive Thresholding

7 Previous Approaches Interactive Thresholding

8 Goal Create a breast density assessment program that is… Automated
Volumetric Accurate In the eXtensible Imaging Platform (XIP) The goal of this project is to create an accurate, automated breast density assessment program that can be easily implemented into the current radiology work platform. Create using RADBuilder so it can be immediately plugged into the GE PACS Workstation and used by radiologists Take the subjectivity out of breast density assessment

9 Material and Methods

10 What is XIP? eXtensible Imaging Platform
Funded by NIH, NCI Cancer Bioinformatics Grid (caBIG) Program A program to share data, informatics tools, technologies and infrastructure between the funded Comprehensive Cancer Centers across the nation Open Source, Open Standards, Open Architecture Platform for development of medical image processing and analysis applications for research and clinical purposes. Will support the DICOM Application Hosting standard to facilitate transportability of XIP applications

11 XIPBuilder Visual Programming Environment for developing imaging applications; based on OpenInventor. Contains modules (lots!) that can be assembled to perform sophisticated tasks. Includes vTK – The Visualization Toolkit (NIH) iTK – The Insight Toolkit for Registration and Segmentation (NIH) Extensible by creating new plug-n-play modules IDE – software development program

12 Mammography

13 Algorithmic Foundations
Highnam et al 1997 Standard Mammogram Form (SMF) Volumetric approach Mammograms on X-ray film Van Engeland et al 2006 Improved upon SMF Full Field Digital Mammography (FFDM) Utilized DICOM headers

14 The Model [2]

15 Volumetric Based on Van Engeland’s algorithm [4]
Tissue composition is computed at each pixel and represents a rectangular cylinder of tissue Area of the pixel is computed from detector characteristics and the geometry of acquisition. Height of the rectangular cylinder is computed from the compression thickness. Volumetric Approach to Breast Density Assessment Assumption that in the breast, there are two types of tissues: fatty tissue and fibro glandular tissue At a given pixel, there is a percent composition of each The tissue composition at a given pixel will be computed, using a reference value for fatty tissue determined by the maximum pixel value in the breast tissue projection.

16 RADBuilder Process

17 RADBuilder Process

18 SoXIPLoadDICOM Loads … DICOM Image Pixel Data
Necessary DICOM Header Information

19 RADBuilder Process

20 SoItkBinaryThresholdImageFilter
Each pixel compared to threshold value Output image of 0 and 1 values Input Output

21 RADBuilder Process

22 SoItkMultiplyImageFilter
Pixel-wise multiplication of two images Lays background mask on top of original image Input 1 Input 2 Output

23 RADBuilder Process

24 SoXIPLinearAttenuationCoefficients
Determines difference between effective attenuation of dense tissue and fat This Attenuation Difference Coefficient is chosen from a table [4] Based on DICOM header information Anode target material Filter material kVp Breast Thickness -our module - Anode filter target and filter material of MAMMOGRAM DICOM HEADER

25 RADBuilder Process

26 SoXIPBreastDenseTissue VolumeCalculation
hd(r) : height of dense tissue at pixel r µd, eff - µf, eff : Attenuation Difference Coefficient g(r) : current pixel value gf : fatty tissue reference value [4]

27 SoXIPBreastDenseTissue VolumeCalculation
hd(r) : height of dense tissue at pixel r µd, eff - µf, eff : Attenuation Difference Coefficient g(r) : current pixel value gf : fatty tissue reference value [4]

28 SoXIPBreastDenseTissue VolumeCalculation
hd(r) : height of dense tissue at pixel r µd, eff - µf, eff : Attenuation Difference Coefficient g(r) : current pixel value gf : fatty tissue reference value [4]

29 SoXIPBreastDenseTissue VolumeCalculation
hd(r) : height of dense tissue at pixel r µd, eff - µf, eff : Attenuation Difference Coefficient g(r) : current pixel value gf : fatty tissue reference value [4]

30 SoXIPBreastDenseTissue VolumeCalculation
hd(r) : height of dense tissue at pixel r µd, eff - µf, eff : Attenuation Difference Coefficient g(r) : current pixel value gf : fatty tissue reference value [4]

31 Where does fatty tissue reference value, gf , come from?
Maximum pixel value (minimum attentuation) in the interior of the breast (this should represent a pure fat pixel). Calculate a histogram of the image. Select the pixel value at 80% (empirical determination) of the maximum pixel values (avoids skin edges; low attenuation due to incomplete thickness).

32 RADBuilder Process

33 SoItkBinaryThresholdImageFilter
Adjust parameters to detect breast edge instead of entire region Input Output

34 RADBuilder Process

35 SoXIPBreastTotalTissue VolumeCalculation
h(r) : total height of tissue at pixel r H : compressed breast thickness d(r) : Euclidean distance to the edge between breast tissue and background [4]

36 SoXIPBreastTotalTissue VolumeCalculation
h(r) : total height of tissue at pixel r H : compressed breast thickness d(r) : Euclidean distance to the edge between breast tissue and background [4]

37 SoXIPBreastTotalTissue VolumeCalculation
h(r) : total height of tissue at pixel r H : compressed breast thickness d(r) : Euclidean distance to the edge between breast tissue and background [4]

38 SoXIPBreastTotalTissue VolumeCalculation
h(r) : total height of tissue at pixel r H : compressed breast thickness d(r) : Euclidean distance to the edge between breast tissue and background [4]

39 SoXIPBreastTotalTissue VolumeCalculation
If Euclidean distance is less than half of breast compression … use first formula in calculation of total tissue volume [4]

40 SoXIPBreastTotalTissue VolumeCalculation
If Euclidean distance is greater than half of breast compression… use breast compression in calculation of total tissue volume [4]

41 RADBuilder Process

42 Calculator Module in XIPBuilder Dense Tissue Volume
Total Tissue Volume x 100 = Percent Density

43 RADBuilder Process

44 Separator Module in RADBuilder Displays output In this case, text

45

46 Methods II 20 CC full-field digital mammograms
Prior manual breast density assessment using ImageJ Determine breast density using Cumulus™, a popular interactive thresholding program Determine breast density with the automated XIP solution

47 Methods III Compare the three measurements of breast density to determine Kendall’s Coefficient of Concordance Use that coefficient to determine a χ2 that allows testing of the null hypothesis: There is no agreement in the assessment of breast density by the three methods

48 Results The overall Kendall’s Coefficient comparing the three systems was 0.502 For 20 cases and three systems, the χ2 is which allows the rejection of the null hypothesis (χ2 of 30.14, 19 degrees of freedom, α =0.05).

49 Conclusion It was possible to develop, in XIP, an automatic software application to measure volumetric breast density in mammograms. The automatic measurement agreed well with two, independent, manual thresholding based techniques in common, current use.

50 Future Work MLO view capability Anisotropic Filter
Improve fatty reference value identification Validate this and other techniques against ground truth (MRI)

51 References Boyd, N., Lockwood, G., Byng, J., Tritchler, D., Yaffe, M., “Mammographic Densities and Breast Cancer Risk”. Cancer Epidemiology, Biomarkers & Prevention, 7, Marias K et al. “A Mammographic Image Analysis Method to Detect and Measure Changes in Breast Density”. European Journal of Radiology. 2004; 52: Vacek, Pamela M., Geller, Berta M. (2004) A Prospective Study of Breast Cancer Risk Using Routine Mammographic Breast Density. Cancer Epidemiology, Biomarkers & Prevention. 13(5) Van Engeland, S., Snoeren, P.R., Huisman, H., Boetes, C., Karssemeijer, N., (2004) A Prospective Study of Breast Cancer Risk Using Routine Mammographic Breast Density. Cancer Epidemiology, Biomarkers & Prevention. 2006; 13(5): Wolfe JN. “Risk for Breast Cancer Development Determined by Mammographic Parenchymal Pattern”. Cancer. 1976; 37:


Download ppt "Rachel Embree and Christina Sillery David Channin, MD, Advisor"

Similar presentations


Ads by Google