Transcript Kein Folientitel - Facultad de Ciencias-UCV

3D Mammography
Ernesto Coto
Stefan Bruckner
Sören Grimm
M. Eduard Gröller
Institute of Computer Graphics and Algorithms
Vienna University of Technology
Motivation
 Breast cancer is the second leading cause of
cancer deaths in women today
 X-Ray Mammography is currently the primary
method of early detection
 Interpretation of Mammograms is difficult
 Radiologists often use CAD for prompting
X-Ray Mammography
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Motivation
 Capability of X-ray mammography is limited by its
2D representation
 Other imaging modalities can obtain a full 3D
representation of the breast
Rendering of MRI Mammography
showing cancer
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Introduction
 Cancer is a disease that causes cells in the body
to divide and reproduce abnormally without control
 Breast cancer refers to a malignant tumor that
has developed from breast cells
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Breast Anatomy
 The breast has two main components: glandular
tissue and connective tissue
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Suspicious Regions
 Calcifications are tiny calcium deposits within the
breast tissue
 Microcalcifications
 Macrocalcifications
 Masses
 Solid
 Liquid (Cyst)
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Doctors Questions







Is there a tumor in the breast?
Is the tumor benign or malignant?
Is it a mass, a cyst or a microcalcification cluster?
What’s the location of the tumor?
What’s the size of the tumor?
How dense is the breast?
What’s the extension of the cancer?
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Imaging Technologies






X-Ray Mammography
Magnetic Resonance Imaging (MRI)
Full-Field Digital Mammography
Nuclear Imaging
Tomosynthesis
Ultrasound
X-Ray
3D Mammography
Ultrasound
MRI
TU Wien
Institute of Computer Graphics and Algorithms
Dynamic Contrast Enhanced MRI
 A tumor is usually well vascularized due to its
strong growth
 The absorption of contrast-medium in suspicious
regions is perceptible
Without contrast
With contrast
Intravenous catheter
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Study of contrast agent’s flow
 Type I: steady enhancement (straight or curved)
 Type II: plateau of signal intensity
 Type III: washout of signal
Benign lesion
Possible malignancy
Strongly suggest malignancy
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Answers using DCE-MRI
 Is there a tumor in the breast?
 Segment the contrast agent
 Is the tumor benign or malignant?
 Study of contrast agent’s flow
 Extract features and use them to classify the
tumors
 Is it a mass, a cyst or a microcalcification cluster?
 Quantify segmented region(s)
 What’s the location of the tumor?
 Estimate position of tumor’s center
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Answers using DCE-MRI
 What’s the size of the tumor?
 Calculate largest tumor diameter
 Calculate volume based on segmented voxels
 How dense is the breast?
 Segment glandular tissue + quantification
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Answers using DCE-MRI
 What’s the extension of the cancer?
 Classify as invasive or non-invasive
Non-invasive
3D Mammography
Invasive
TU Wien
Institute of Computer Graphics and Algorithms
CAD for DCE-MRI
“Interactive Detection and Visualization of Breast Lesions Using Dynamic Contrast
Enhanced (DCE) MRI Volumes“, Computerized Medical Imaging and Graphics, Elsevier, July 2004.
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Questions
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms
Suggestions
 How can me make this better???
3D Mammography
TU Wien
Institute of Computer Graphics and Algorithms