Dkhar, Winniecia and Kadavigere, Rajagopal and Paruthikunnan, Samir Mustaffa (2020) Image Quality Analysis for Optimization of b-value in Diffusion Weighted MRI of Breast. Malaysian Journal of Medicine and Health Sciences, 16 (4). pp. 126-131. ISSN 1675-8544

PDF 10569 DisplayPdf.pdf - Published Version Restricted to Registered users only Download (258kB) | Request a copy

Abstract

Introduction: Diffusion-weighted (DW) is an evolving technique that provides both qualitative and quantitative data about the diffusion of water molecules and b-value plays a significant role in DW images. The aim of this study is to find the optimal b-value in diffusion-weighted MRI for the detection of breast lesions. Methods: The study included 124 subjects. T2-FS sequences were acquired for lesion measurement. DW was acquired with b-values 300, 600 and 1000 s/mm2. DW images were evaluated by two radiologists under the criteria's - signal-to-noise ratio (SNR), contrast differentiation of lesion from the glandular tissue, and contrast differentiation of glandular tissue from fatty tissues. Relative Signal intensity ratio (SIR), SNR and contrast-to-noise ratio (CNR) were calculated. Results: Kappa value is highest for a b-value of 600 for all the three criteria, for subjective SNR the kappa value was 0.906 which indicates that the inter-rater reliability rate is an almost perfect agreement. Differentiation of lesion from the fibro-glandular tissues, the kappa coefficient is 0.904 which interprets that the agreement between the readers is almost perfect and the kappa coefficient to differentiate the fibro-glandular from fatty tissues is 0.875 which shows that there is substantial agreement. Relative SIR, SNR, and CNR is highest for a b-value of 600 with a mean value of 4.6, 14.7, and 44 respectively. Conclusion: Optimal b-value for good image quality is 0, 600 s/mm2. Standardizing an optimal b-value is highly significant since it improves the overall image quality and enhances multi- center diagnosing with an accurate report

Actions (login required)

View Item