Research On Quantitative Determination Of Articular Cartilage Thickness Changes In MR Image

Osteoarthritis(OA),also referred to as osteoarthrosis or degenerative joint disease,is the most common of the rheumatic diseases.Cartilage degeneration has been identified as a major threat to our aging society with substantial implications for health care and as a major socio-economic burden.The aim of this dissertation is to quantify focal,local cartilage thickness changes in longitudinal study from magnetic resonance imaging(MRI)of osteoarthritic joint of the same patient.This technique may be applied to monitor the progression of joint diseases or related treatments and to evaluate the efficacy of new therapeutic compounds.To investigate whether the accuracy in thickness measurement of femoral cartilage is influenced by acetabular cartilage,we developed a mathematical model for two adjacent sheet structures,which simulated the femoral and acetabular cartilages in the hip joint.MR imaging process and post-processing for thickness measurement are also modelled and simulated.The simulation method is validated by comparison with the actual results obtained from the experiments using 3 phantoms,3 normal cadaver hip specimens and9 patients with osteoarthritis.The result of simulation measurements shows that considerable underestimation in thickness measurement occurred due to the influence of the adjacent sheet structure.In order to remove the influence of the adjacent sheet and calibrate measurement bias,we propose a measurement method based on a model of the MR imaging process for two adjacent sheet structures.The in-plane thickness is computed based on the boundaries detected using a model-based method.In order to correct the image plane thickness for overestimation due to oblique slicing,a three-dimensional(3-D)thickness computation approach is introduced.Experimental results show that the thickness measurement obtained by the new thickness computation approach is more accurate than that obtained by the existing thickness computation approaches.And,when tracing the variation of local cartilage thickness,we transfer the problem of point correspondence into task assignment problem.To improve precision of algorithm,first,we utilize an algebraic method to align the points approximately,and then applying Munkres assignment algorithm to cope with correspondence among the points.Finally,with regard to low precision of the algorithm,we change the problem of tracking variation of local cartilage thickness into a 3-D points rigid calibration problem,where we raise a globally optimal calibration algorithm.Given correspondence of all points,the problem of finding the optimal solution in the six-dimensional space composed of the rotation parameters and the translation parameters is divided into two steps:first,the optimal rotation parameters are found in the three-dimensional rotation space;The optimal translation parameters are then obtained by a transformation.In seeking process,we apply Munkres assignment algorithm to resolve corresponding problem among the points,and taking a globally optimal algorithm MCS to calculate the best rotation parameters.The algorithm guarantees the best solutions to 3-D points calibration problem,and it resolves the local convergence problem of traditional methods.At last,we prove the efficiency of method by both synthetic and practical MR data.