Sensing mammographic density using single-sided portable Nuclear Magnetic Resonance.

This research paper presents a quantitative approach to sensing mammographic density (MD) using single-sided portable Nuclear Magnetic Resonance (NMR). It focuses on three main techniques: spin-lattice relaxation (recovery) time (T), spin-spin relaxation (decay) time (T), and Diffusion (D) techniques by testing whether or not the aforementioned techniques are in agreement with the gold standard and with each other when used for scanning breast tissue specimens with a variety of mammographic densities (MDs). The high mammographic density (HMD), intermediate MD, and low mammographic density (LMD) regions of each slice were identified according to the mammogram images. Subsequently, the grayscale values for these regions were quantified. One region was measured from the first sample while the remaining ones were measured from the second sample. The same areas were then exposed to portable NMR, and the sequences used as following: the stimulated echo sequence for diffusion (D), the Carr-Purcell-Meiboom-Gill (CPMG) sequence for T, and saturation recovery sequence for T. The correlations between the grayscale values and NMR techniques were strongly correlated. The Pearson correlation coefficient, R, of T (%) versus grayscale value, D (%) versus grayscale value, and T (%) versus grayscale value, was 0.91, 0.91, and 0.93, respectively. Furthermore, the relative water content of the breast slices based on T, T, and diffusion (D) measurements were strongly in agreement with each other. The Pearson correlation coefficient, R, of D (%) versus T (%), D (%) versus T (%), and T (%) versus T (%), was 0.984, 0.966, and 0.9868, respectively. The three pulse sequences can be employed in a portable NMR device to deliver continuous quantitative measurements of MD in breast tissue samples. As a result, the method demonstrated to be acceptable for determining the distribution of MDs among breast tissue samples without the need for additional qualitative analysis.