Research Of Quantitative T₂ Mapping At Low Field Mr System

Magnetic resonance imaging(MRI) is a non-invasive diagnostic technique, which is based on the principles of nuclear magnetic resonance(NMR). With 2D or 3D imaging, it obtains anatomical and functional information of the inside tissues and organs of human body and provides diagnostic information of pathological structures. As a powerful tool for medical and biology research, MRI has been widely used in the field of clinical diagnosis and scientific researches.It is always important to measure the value of T2 relaxation time in MRI because T2 value is often used to identify the tumor and determine their nature. At present, most of the quantitative T2 mapping techniques are implemented in high-field superconducting MR systems. However, the superconducting MR scanner has not been popularly used in domestic hospitals due to its high prices and maintenance costs of superconducting systems. The low-field opened MR scanner is widely employed mainly due to its low price and maintenance costs, as well as many advantages of open magnet. It has become the mainstream of the routine clinical diagnostic system. Compared with high-field systems, measurement data based on low-field systems are scarce as a result of its shorter development period. Besides, some measurement techniques used for high-field systems are hard to realize in low-field ones because of its natural characteristics, such as low signal-to-noise-ratio(SNR) and poor stability. Therefore, it is quite meaningful for both clinical usage and low-field system to find a fast and accurate measurement method of quantitative T2 mapping.Considering the characteristics of low SNR in low-field system, a novel and more effective post-processing curve fitting method for quantitative T2 mapping is proposed. Adaptive curve fitting algorithm is achieved based on T2-to-noise-ratio(T2NR) optimization. Echo number used for curve fitting is adaptively optimized voxel wised with the estimation of T2 value in the corresponding voxel. Thus, T2NR is optimized, which qualifies higher SNR and accuracy improvement in quantitative T2 mapping. Computer simulation, phantom and in vivo study have all shown that more accurate quantitative T2 mapping, as well as higher SNR, is obtained by the proposed method in comparison with truncation curve fitting method.