The Development Of Magnetic Resonance Pulse Sequences And The Applications In Orthopedic Research

Nuclear magnetic resonance（NMR）and magnetic resonance imaging（MRI）find extensive applications in the fields of chemistry,biology,and clinical medicine.Pulse sequences serve as the fundamental technique in NMR,enabling signal selection,enhancement,and quantification.It is of utmost importance to optimize and apply novel pulse sequences in emerging fields.Bones constitute the primary structural support in the human body,but their internal structure is intricately complex,filled with various types of tissues.There are few characterization methods available to obtain comprehensive and detailed information about the internal structure of bones.Currently,clinical bone density assessments only provide macroscopic mineral density information.NMR,based on the observation of atomic nuclei,can provide comprehensive information regarding chemical composition,spatial structure,and dynamic characteristics,making it of significant importance in orthopedic research and disease diagnosis.However,current NMR-related studies of bones primarily focus on trabecular and cartilaginous bones,with very limited research on cortical bones and bone marrow.Only macroscopic parameters such as water content or porosity can be obtained,failing to reveal the microstructure of cortical bones and the physiological role of bone marrow.In the first part,we conducted magnetic resonance imaging of cortical bones using ultra-high field（14 T）,obtaining micron-level images of cortical bones for the first time via MRI.In the study,by employing ultrashort echo time magnetic resonance imaging（UTE-MRI）,spectroscopy,diffusion,relaxation,and exchange analyses,the attribution of T₂*components within cortical bones was achieved,distinguishing collagen-bound water,free water in pores,and lipids.Based on this,the tubular structure within cortical bones was resolved,and the spatial distribution of these three components was determined.Through ultra-high field UTE-MRI and statistical analysis of osteoporotic rats with varying durations of the disease,the pathological development of osteoporosis was investigated.Similarly,the analysis of clinical samples differentiated local pathological features between human osteoporosis and osteoarthritis samples.Diffusion-weighted MRI can measure the diffusion characteristics of bone marrow,reflecting its physiological features.However,the bone marrow region exhibits significant heterogeneity and spatial magnetic field inhomogeneity,resulting in significant artifacts in MRI signals.Diffusion-weighted MRI based on novel spatiotemporal encoding（SPEN）is less affected by these challenges but is currently commonly used for brain imaging.Therefore,in the second part of this study,we focused on the characteristics of the bone marrow region and improved the sequence to obtain a high b-value diffusion-weighted SPEN sequence（hbd-SPEN）.By expanding the b-value range within a limited echo time,the hbd-SPEN sequence facilitated the observation of samples such as bone marrow,which exhibit fast relaxation and slow diffusion.After b-value correction based on local gradient effects,the diffusion coefficients displayed good isotropy in water phantoms and produced images with smaller artifacts compared to traditional DW-EPI sequences in various sample types.Based on this,we successfully measured the diffusion coefficients and DC-maps in isolated chicken femur and live mouse femur regions,both demonstrating larger diffusion coefficients along the femur shaft.Moreover,this sequence was also applied to the skull and spinal regions of live mice,exhibiting reduced artifacts.Steady-state free precession（SSFP）sequences improve the signal-to-noise ratio within a given sampling time and find extensive applications in the field of MRI.However,SSFP is limited by a relatively short TR,making it challenging to acquire spectra from samples with slow relaxation.In the third part of this study,we designed a novel linear-combination SSFP sequence（LC-SSFP）.By designing increasing phase increments and acquiring steady-state signals under each phase increment,the LC-SSFP sequence allowed for filtered analysis and frequency distribution spectra.The feasibility of LC-SSFP was theoretically analyzed,and through the design and optimization of filters,it was well-validated in simulations based on the Bloch equation under various scenarios.The sequence also achieved good results in single-peak experiments involving multiple types of nuclei and showed comparable or improved unit-time signal-to-noise ratio in some scenarios compared to single-pulse sequences.However,for complex samples and samples with fast relaxation,the reconstruction may include some errors.The feasibility of LC-SSFP was demonstrated through various types of experiments,and subsequent experiments on cortical bones and bone marrow confirmed its applicability to novel orthopedic research.