Hadi Tadayyon June 11 2012 Clinical challenge High risk for metastasis Characterized as gt 5 cm long axis Skinchest wall involvement Lymph node involvement Tumour grade a histological feature that is a prognostic indicator and is important for treatment design ID: 395872
Download Presentation The PPT/PDF document "Non-invasive breast tumour grading using..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Non-invasive breast tumour grading using ultrasound frequency-dependent backscatter analysis
Hadi TadayyonJune 11, 2012 Slide2
Clinical challengeHigh risk for metastasis
Characterized as:> 5 cm long axisSkin/chest wall involvementLymph node involvementTumour grade – a histological feature that is a prognostic indicator and is important for treatment designDetermined from pathological examination of biopsy sampleHigher grade higher degree of malignancy / poorer prognosis
Our goal: ultrasonically detect variation in tumour grades
Locally advanced breast cancer (LABC)Slide3
Midband fit, slope, and intercept used to differentiate:
Prostate cancer from benign tumours1Metastic from non-metastatic lymph nodes2Acoustic scatterer spacing used to characterize breast lesions as benign or malignant3
Quantitative ultrasound characterization of cancers
Feleppa
et al., 2004 IEEE Trans UFFC, 43(4), 609-619, (1996)
Mamou
et al., Ultrasound in Med. & Biol., 37(3), 345–357, (2011)
Y
.
Bige et al., Ultrasonics 44 , 211–215, (2006)Slide4
Can LABC tumours be characterized in terms of grade using quantitative ultrasound?Given:
Retrospective in-vivo clinical breast data (N=43)A diagnostic ultrasound machineResearch QuestionSlide5
Methods: data collection and classification
Tumour ROI
1 cm
Normal breast ROI
QUS
Midband
fit (MBF)
Slope (SS)
Intercept (SI)
Scatterer spacing (SAS)
GI (N=3)
GII
(N = 22)
GIII
(N = 18)
10 MHz
fc
linear array transducer (
Ultrasonix
, Canada)
4-7 cm depth
5 MHz 50% bandwidth
Normal tissue ROISlide6
Methods: Spectral analysis
Depth-dependent spectral normalization (reference phantom)
Variable bandwidth linear regression
Discrete depth spectral normalization (reference reflector)
Auto-regressive
(AR) spectral estimation and autocorrelation-derived
scatterer
spacingSlide7
Results: QUS distributions among tumour grades
Mann-Whitney test: p = 0.032Slide8
Results: parametric
images of scatterer spacingGI
GII
GIIINT
0
0.5
SAS (mm)
10 um
1 cmSlide9
The link of scatterer spacing to biology
Mean spacing between glandular islands = 200 um
100 umSlide10
A potential method to non-invasively characterize tumour grade was proposedScatterer spacing statistically different among tumour grades (ANOVA test & Mann-Whitney test)
Scatterer spacing is linked to spacing between glandular islands Small sample size for GIIn large population study – 362/1409 = 25%In our study, 3/43 = 7%Cannot evaluate classification due to insufficient parametersFuture directions: investigate other QUS parameters
Discussion & conclusionSlide11
Acknowledgments
Czarnota
Lab, University of Toronto
Dr.
Czarnota
Dr. Omar
Falou
Mike
Papanicolau
Sara
Iradji
Ervis
Sofroni
Ryerson University
Dr. Lauren
Wirtzfeld
University of Illinois
Dr. Michael
Oelze
CGSDSlide12
Tumour grade
Increasing risk of metastasisGrade I
Grade II
Grade III
Total score 3-5
6-7
8-9Slide13