Introducing a New Mathematical Function for Prediction of Breast Cancer Growth in MRI
摘要
Much attention has been paid to the breast cancer seen in the diagnostic images. Breast cancers can be of different types or a combination of invasive and in situ cancer. Since breast tumors are grown increasingly compared to other types of tumors, quantitative appraisal using non-invasive imaging techniques like magnetic resonance imaging (MRI) is extensively utilized in the medical field.
MethodsHere, a new function is proposed that uses distinct mathematical relations on the relaxation times in MRI and in which the new maps are created by introducing a unique index. In our study, Transverse-longitudinal function (TLF) was defined as a combination of three parameters: T1, T2, and alpha, which can be zero for a certain alpha. When plotting the TLF, a maximum value is obtained, which we defined as a percentage of the maximum width at x value. Then, by taking the inverse of the TLF, an interval called FWxM was obtained. In other words, this parameter is the difference between two alpha-stars of minimum and maximum. Changing x value also changes the maximum and minimum alpha-star values, so we have a specific FWxM parameter for each pixel of the image. When this parameter was calculated for the entire image, a FWxM was consequently obtained for the entire image.
ResultsBy calculating the SNR, more details in the whole breast image become more clearly visible, and as the x value increases towards 1, the details of healthy and tumor areas are simultaneously better displayed. The SNR amount increased from 19.37 towards 31.24 for x = 1 and 0.01, respectively.
ConclusionsOverall, the maps were able to display breast structures alongside further details at greater x values.
Graphical Abstract