Ultrasonic Estimation of Effective Scatterer Diameter for Micro-Macro Parameter Based Improved Breast Lesion Classification
Year: 2019
Type: preprint
Abstract: ESD is an important micro-parameter for ultrasonic tissue characterization. However, classifying between benign and malignant breast lesions from a broad dataset remains a challenge. In this work, we propose a new technique for the estimation of ESD of breast tissues from the diffuse component of backscattered RF data. This allows us to combine ESD with MSS and other ultrasonic macro-parameters for binary classification of lesions. In order to separate the diffuse component from the coherent component of the backscattered data, EEMD is performed. K-S test is used to automatically select the IMFs responsible for diffuse scattering. To ensure proper minimization of the system effects, a multi-step process is adopted where the RF data is deconvolved and filtered as the first two steps prior to normalization. As the ESD is supposed to have a fairly consistent value over a small tissue subregion, it is estimated, from the slope of the average regression line, computed from the exponential weighted average of lines, fitted over the log-power spectra of the neighboring blocks. Using our method, the ESD for malignant lesions is estimated to be 123.05{\pm}8.85 {\mu}m and that for fibroadenomas is estimated to be 98.88{\pm}9.48 {\mu}m. ESD values estimated for inflammatory breast tissues (~76 {\mu}m) are close to normal tissues. When the ESD values are used to classify 159 benign and malignant lesions, we obtain sens., spec., and acc. values of 91.07%, 96.11%, and 94.34%, respectively. ESD is also combined with the MSS and 27 previously reported macro-parameters, estimated from the UB and UE, for classification. This results in high sens., spec., and acc. values of 96.43%, 98.06%, and 97.48%, respectively. The proposed classification approach based on this hybrid feature set, therefore, show great potential to be used as a CAD tool for breast lesion classification.
Source: arXiv (Cornell University)
Language: English
Subfield: Radiology, Nuclear Medicine and Imaging
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W2914933271