| Part I The study of correlation between lumber intervertebral disc degeneration and Pfirrmann grading systemObjective:(1) To investigate the distribution of disc Pfirrmann grades in healthy volunteers and lay a foundation for further quantitative analyses.(2) To investigate the correlaitons between the degeneration in the lumber spine and the gender, disc level, age and suggest how intervertebral disc degeneration (IVDD) might be affected by physiologic factors.Materials and Methods:Subjects:88healthy volunteers (gender:36males,52females; age range=20-76years, mean age=33.5±14.0years) were recruited following approval from Institution Review Broad. Inclusion criteria consisted of:(1) no history of manual labor for a long time;(2) no significant backache within the past6months;(3) no history of spine trauma and deformity.MR:MR scanning was perfomed3.0T magnetic resonance imaging scanner with dual-source RF transmission(3.0T Achieva TX, Philips) and saggital T2-weighted images were acquired using a fast spin echo sequence with a spine coil for Pfirrmann grading in all patients. The parameters were TR=3000ms, TE=128ms, FOV=220mm(FH)×201mm(AP)×55mm(RL), slice thickness=5mm, matrix=448X448, flip angle=90°, All morphological evaluations were performed by two radiologists in consensus, both with more than10years of experience. All the degenerative discs on T2WI were classified according to the Pfirrmann grading system.Statistical methods:Statistical analyses were conducted using SPSS15.0software. Disc levels, age and sex of the volunteers and Pfirrmann grades were recorded and used for statistic analyses. We used (1) Kruskal-Wallis test between Pfirrmann grades and disc levels;(2) Wilcoxon test to calculate significant difference between Pfirrmann grades and sex;(3) A one-way analysis of variance (ANOVA) was employed to compare between age and Pfirrmann grades and the Spearman correlation coefficient was determined. Levene test for equality of variances was used to test homogeneity of variances among groups and under an equal condition, the significance of differences were tested using ANOVA with LSD as a post hoc test, otherwise a Dnnett T3test was used. P≤0.05was considered as significance.Results:The T2-weighted image-based on Pfirrmann grade classification results were as the following:69disks in grade Ⅰ,232in grade Ⅱ,90in grade Ⅲ,37in grade Ⅳ and12in grade Ⅴ.No statistically significant difference was observed between the grades (P=0.156)The difference in Pfirrmann scores between male and female was not obvious and not statistically significant (P=0.361). There was a significant positive correlation between age and degenerative grade (rs=0.668, P=0.001). The average age of each grade had statistically significant difference, but no difference between grade Ⅳ and grade Ⅴ.Conclusion:Pfirrmann grading system is the most common imaging classification method of IVDD but it provides only qualitative data that can be influenced by variable conditions. Part Ⅱ The study of correlation between lumber intervertebral disc degeneration and T1ρ-weighted imagingObjective:(1) To determine whether Tip-weighted imaging is quantitative biomarkers of IVDD and provide new technology for objective evaluation of degenerative discs.(2) To analyze the relationships between Tip and Pfirrmann grades, intervertebral disc levels, gender and age.Materials and Methods:Subjects:83healthy volunteers (gender:34males,49females; age range=20-76years, mean age=33.9±14.3years) were recruited following approval from Institution Review Broad. Inclusion criteria consisted of:(1) no history of manual labor for a long time;(2) no significant backache within the past6months;(3) no history of spine trauma and spinal deformity. All volunteers were subdivided into three groups according to age,<35years,36-50years and>51years, respectively.MR:Scanners and sagittal T2-weighted images acquisition were the same as mentioned in Part Ⅰ. T1ρ scanning sequence used3D steady gradient echo sequence and parameters were:TR=4.85ms, TE=2.39ms, FOV=220mm(FH)×201mm(AP)×55mm(RL), slice thickness=5mm, matrix=448×448, Flip angle=50°, spin-lock frequency=500Hz, TSLs=0/10/20/30/40ms.Image processing:All discs were graded according to Pfirrmann grading system. Tip values were calculated on a pixel-by-pixel basis to the exponentially decaying Tip fuction using IDL (Research Systems, Inc, USA) to generate Tip relaxation maps.The color-coded maps were fused with T2WI and ROIs were drawn. Three equally sized rectangular regions were manually subdivided on the middle slices and the center ones were chosen as the areas of interest.Statistical methods:Statistical analyses were conducted using SPSS15.0software.(1) ANOVA was employed to compare Tip values of different Pfirrmann grades and the Spearman correlation coefficient was determined. Furthermore, in order to determine the T1ρ boundary values between the Pfirrmann classification grades in the nuleus pulposus, receiver operating characteristic (ROC) analyses were perfomed between each grade which were significantly different.(2) ANOVA was employed to compare Tip values of different dick levels.(3) The independent sample t test was used to compare Tip values of different gender at the same disc level.(4) ANOVA was employed to compare between T1ρand age groups and the Spearman correlation coefficient between T1ρ and age was determined. Levene test for equality of variances was used to test homogeneity of variances among groups and under an equal condition, the significance of differences were tested using ANOVA with LSD as a post hoc test, otherwise a Dnnett T3test was used.P<0.05was considered as significant. Results:The T2-weighted image-based on Pfirrmann grade classification results were as the following:50disks in grade Ⅰ,188in grade Ⅱ,75in grade Ⅲ,37in grade Ⅳ and8in grade Ⅴ.Color-coded T1ρmap displayed normal nucleus pulposus could be divided into peripheral low values and core high values areas and fiber annulus showed lowest values. With the reduction of T2WI signal, nucleus pulposus values gradually were lower and values of fiber annulus increased.Tip values were statistically significantly different when comparing between every grade, but not when comparing between grades Ⅲ and Ⅴ. T1ρ was significant negative correlation with Pfirrmann grade (rs=-0.538). ROC analyses illustrate the Tip cut-off value between grade Ⅰ and Ⅱ was116.5ms, which corresponded to a sensitivity, specificity, and area under the curve (AUC) of70%,53.2%and0.623respectively. The cut-off value, sensitivity, specificity, and AUC between grade Ⅱ and Ⅲ were103.5ms,67.6%,66.7%,0.720respectively, and those between grades Ⅲ and Ⅳ were86.5ms,60%,75.5%and0.717respectively.T1ρ values were statistically significantly different when comparing between every level, but not when comparing between L4/5and L5/S1.The difference in T1ρ values between male and female in the nuleus pulposus of every level was small and not statistically significant.T1ρ values decreased significantly with increasing age and this correlation with age was moderate negatice (rs=-0.335). The T1ρ values remained relatively stable during20-45years and rapidly and continuously reduced after45years.Conclusion:Tip reflects changes in IVD composition, with comparably high spatial resolution. T1ρ imaging technique is more sensitive to molecular modification than T2WI and can be potentially used as a clinical tool to detect early IVDD and longitudinal followup. The method can create a reliable quantitative scale, as the result, the accuracy and repeatability of measurements are better than Pfirrmann grading system. ParⅢ The study of correlationship between lumber intervertebral disc degeneration with T2Mapping imagingObjective:(1) To determine the associations between T2and Pfirrmann grades, intervertebral disc levels, gender and age.(2) To assess, compare and correlate T1ρ and T2relaxation time.Materials and Methods:Subject:86healthy volunteers (gender:34males,52females; age range=20-76years, mean age=33.5±14.1years) were recruited following approval from Institution Review Broad. Inclusion criteria consisted of:(1) no history of manual labor for a long time;(2) no significant backache within the past6months;(3) no history of spine trauma and spinal deformity. All volunteers were subdivided into three groups according to age,≤35years,36-50years and≥51years respectively. Volunteers were investigated in the afternoon (5-10PM) and did not exercise strenuously and laid down about half an hour before scan.MR:Scanners and sagittal T2-weighted images acquisition were the same as the mentioned in Part I and T2Mapping parameters were:TR=1162ms, TE=n×20ms, FOV=220mm(FH)×201mm(AP)×25mm(RL), slice thickness=5mm, matrix=448×448, Flip angle=90°. To reduce the artifacts caused by respiratory and gastrointestinal moving, the saturation band was set in front of the vertebral bodies and automatic shimming and fat suppression techniques were applied.Image processing:All discs were scored according to Pfirrmann grading system. T2values were calculated on a pixel-by-pixel basis to generate T2relaxation map. The color-coded maps were fused with T2WI and ROIs were drawn.Statistical methods:Statistical analyses were conducted using SPSS15.0software.(1) ANOVA was employed to compare T2values of different Pfirrmann grades and the Spearman correlation coefficient was determined. Furthermore, in order to determine the T2boundary values between the Pfirrmann classification grades in the nuleus pulposus, ROC analyses was perfomed between each grade which were significantly different.(2) ANOVA was employed to compare T2values of different disc levels.(3) The independent sample t test was used to compare T2values of different gender at the same disc level.(4) ANOVA was employed to compare between T2values and age groups and the Spearman correlation coefficient between T2and age was determined.(5) The Pearson correlation coefficient was determined between T1ρ and T2values. Levene test for equality of variances was used to test homogeneity of variances among groups and under an equal condition, the significance of differences were tested using ANOVA with LSD as a post hoc test, otherwise a Dnnett T3test was used.P≤0.05was considered as significant.Results:The T2-weighted image-based on Pfirrmann grade classification results were as the following:66disks in grade Ⅰ,225in grade Ⅱ,81in grade Ⅲ,37in grade Ⅳ and11in grade Ⅴ. Color-coded maps displayed normal nucleus pulposus could be divided into peripheral low values and core high values area and the fiber annulus were lowest. Comparing with the T1ρ, the high values were lower and the values of nucleus pulposus and fiber annulus were more uniform. With the reduction of T2WI signal, the values of pulposus gradually were lower and fiber annulus changes were less significance than T1ρ.T2values were significantly different when comparing between every grade, but not when comparing grade Ⅰ with Ⅱ and grade Ⅳ with Ⅴ. T2values was significant negative correlation with Pfirrmann grades (rs=-0.665). ROC analyses illustrated the T2cut-off value between grade Ⅱ and Ⅲ was94.5ms, which corresponded to a sensitivity, specificity, and AUC of84.9%,71.6%and0.865respectively. The cut-off value, sensitivity, specificity, and AUC between grade Ⅲ and Ⅳ were67.5ms,75.3%,91.9%and0.906respectively.T2values were not significantly different when comparing between every level.Overall in all intervertebral disc level, T2values of male were higher than female’s, even there was significant difference between L1/2and L2/3.T2values decreased significantly with increasing age and this correlation with age was moderate negative (rs=-0.492). T2values during20-45years decreased slowly and increased slightly during45-50years, then the values fell rapidly again and stayed low stage after55years.T2values was moderate positive correlation with T1ρ value(r=0.445).Conclusion:The method of T2Mapping seems to be able to characterize different degrees of disc degeneration quantitatively and become a promising method for longitudinal followup. T2Mapping can monitor the changes in tissue composition and is less sensitive for early IVDD than T1ρ, on the other hand, T2Mapping demonstrate higher degree of accuracy for late stage. The moderate positive correlation between Tip and T2can suggest the decrease of PG content in nucleus pulposus lead to the water loss. PartIV The study of correlationship between intervertebral disc degeneration with diffusion weighted imagingObjective:(1) To determine the associations between ADC values and Pfirrmann grades, intervertebral discs levels, gender and age.(2) To assess, compare and correlate quantitative Tip, T2relaxation time and ADC values and finish the comprehensive analyses for the three quantitative methods.Materials and Methods:Subjects:42healthy volunteers (gender:16males,26females; age range=22-63years, mean age=38.2±12years) were recruited following approval from Institution Review Broad. Inclusion criteria consisted of:(1) no history of manual labor for a long time;(2) no significant backache within the past6months;(3) no history of spine trauma and spinal deformity. All volunteers were subdivided into three groups according to age,≤35years,36～50years,≥51years respectively. Volunteers were investigated in the afternoon (5～10PM) and did not exercise strenuously and laid down about half an hour before scan.MR:MR instrument, Tip and T2Mapping scan parameters were the same as the mentioned in Part1-3. DWI parameters were:TR=3000ms, TE=64ms, FOV=180mm(FH)×56mm(AP)×65mm(RL), slice thickness=5mm, matrix=288×288, flip angle=90°, b values=Osec/mm2and500sec/mm2. Fov and acquisition matrix of T2WI were the same as the ones of T1ρ and DWI, respectively.Image processing:All discs were scored according to Pfirrmann grading system. ADC values were calculated on a pixel-by-pixel basis using the scanner software. The color-coded ADC maps by using ImageJ software. At last the color-coded maps were fused with convention T2WI and ROIs were drawn.Statistical methods:Statistical analyses were conducted using SPSS15.0software.(1) ANOVA was employed to compare ADC values of different Pfirrmann grades and the Spearman correlation coefficient was determined. Furthermore, in order to determine the ADC boundary values between the Pfirrmann classification grades in the nuleus pulposus, ROC analyses were perfomed between each grade which were significantly different.(2) ANOVA was employed to compare ADC values of different dick levels.(3) The independent sample t test was used to compare ADC values from different gender of the same level.(4) ANOVA was employed to compare ADC values and age groups and the Spearman correlation coefficient between ADC and age was determined.(5) The Pearson correlation coefficients were determined between ADC with T2and T1ρ respectively. Levene test for equality of variances was used to test homogeneity of variances among groups and under an equal condition, the significance of differences were tested using ANOVA with LSD as a post hoc test, otherwise a Dnnett T3test was used.P≤0.05was considered as significant.Results:The T2-weighted image-based on Pfirrmann grade classification results were as the following:19disks in grade Ⅰ,106in grade Ⅱ,44in grade Ⅲ,29in grade Ⅳ and9in grade Ⅴ. Color-coded maps displayed normal nucleus pulposus could be divided into core low values and peripheral high values areas and fiber annulus were black which indicated extremely low values, as the result, the difference between nucleus pulposus and fiber annulus was more significant than other two kinds of color-coded maps. Comparing with the Tip and T2Mapping, the distribution of high values were different that laid on the peripheral area.ADC values were no significantly different when comparing between grade Ⅰ and Ⅱ, grade Ⅲ and Ⅳ. ADC was significant negative correlation with Pfirrmann grade (rs=-0.530). ROC analyses between each grade illustrated the ADC cut-off value between grade Ⅱ and Ⅲ was1.732×10-3mm2/s, which corresponded to a sensitivity, specificity, and AUC of50%,81.8%and0.673respectively. The cut-off value, sensitivity, specificity, and AUC between grade Ⅲ and Ⅳ were1.582×10-3mm2/s,64.1%,89.7%and0.813respectively.ADC values were statistically significantly different when L1/2comparing with L2/3and L3/4(P=0.014, P=0.050).No statistically significant difference was found between men and women in each disc level.ADC values decreased significantly with increasing age and this correlation with age was moderate negative (rs=-0.422). ADC values during20-45years decreased slowly and increased dramatically during45-50years, then the values fell rapidly again after50years.ADC values were significant positive correlation with T1ρ and T2values.Conclusion:The method of DWI seems to be also able to characterize different degrees of disc degeneration quantitatively and ADC values are significant positive correlation with T1ρ and T2, which can reflect the changes in tissue composition. T1ρ, T2and DWI are the promising methods for detecting early early and longitudinal followup ofIVDD. Part V A novel rabbit model of disc degeneration by needle puncture under X-ray guidanceObjective:To establish a reproducible rabbit model of disc degeneration by needle puncture using X-ray guidance, then determine whether MR quantitative technology can evaluate the disc degeneration of rabbits.Materials and methods:Five SP healthy New Zealand white rabbits were used in this study under the Insitutional Animal Care Use Committee approval. Nucleus pulposus and cartilage endplates were destructed respectively through anular puncture under X-ray guidance. The IVDD was evaluated using MR T2WI, T1ρ and T2Mapping techniques at the time points of pre-puncture, after30minutes,4weeks and8weeks. Pathological changes of the intervertebral discs were analyzed after8weeks.Results:T2WI and CT:After30minutes, MR and CT failed to find changes of intervertebral disc. On the contrary, after4and8weeks, signal intensity of the punctured discs decreased. Other disc levels intensity had not different with control group.Tip technology did not provide enough spatial resolution to detect the values changes of nucleus pulpous. T2Mapping showed values of nucleus pulposus were significant low after4and8weeks in the nuclers pulposus of punctured discs and the values of other discs remains no change.Conclusion:Nucleus pulposus destructed approach by needle puncture using X-ray guidance can result in a convenient, less-invasive, reproducible and cost-effective rabbit model of disc degeneration and be a useful choice for studying disc degeneration. T2Mapping can quantantatively evaluate the rabbit IVDD, on the contray, T1ρ technique is unable to detect the degeration in our experiment and more research investigation is required. SummaryThe relationship between Pfirrmann grades and three quantitative techniques: According to correlation coefficients, T2values had the strongest correlation with Pfirrmann grading, which can be resulted from Pfirrmann grading system based on disc signals of T2WI is closely correlated with water content in the nucleus pulposus. T2Mapping and DWI failed to discover the differences of grades I and II and it suggests the both techniques are less sensitive for early degeneration than T1ρ.Three kinds of quantitative values were not significantly different between grades IV and V. Two reasons can lead to the result:First, according to Pfirrmann grading system, the difference between grades Ⅳ and Ⅴ is only the space of the intervertebral disc, regardless of the signal changes. Second, it also may be due to the small sample size of grade V. Furthmore, the narrow space of intervertebral disc may be closely correlated with clinical symptoms since the low incidence of grede V in healthy adult populations.The relationship between disc levels and three quantitative techniques:Only T1ρ technology could identify the difference of L4/5and L5/S1from other levels with lower incidience of disc heniation.So we consider T1ρ is more sensitive for early degeneration and T2Mapping demonstrate higher degree of accuracy for late stage.Gender has little impact on the normal human intervertebral disc degeneration.The relationship between age and three quantitative techniques:All three methods showed moderate or significant age-associated changes but Tip values showed the weakest correlation and remained stable younger than45years old, which was consistant with fundamental research that nucleus pulposus matrix was hard to be repaired older than50years. So we think the most significant change in nucleus pulposus should be the loss of water content young and middle-aged populations. Hypothesis:Comparing three value curves of age-associated changes, ADC values fluctuated during45-50years old. We suggest ADC values are closely correlated with water and PG cotent. One one hand, PG attracted the water; on the other hand, PG restricts Brown motion of water molecules. Before45years old, PG remains stable and water gradually loses, which lead to ADC values decrease. During45-50years old, PG content diminishes dramatically and can not restrict the motion of water, as the result, ADC values increase although water content lose continually. After50years old, the loss of water dominates and ADC values decrease again. So studying the correlation between DWI and IVDD, more factors need to be considered. We think it is more important to note the tissue composition changes dramatically at the45-50years old stage and the finding can provide the evidence to support clinical intervention for delaying IVDD. We can induce a convenient, less-invasive, reproducible and cost-effective rabbit model of disc degeneration and T2Mapping is a promising method for investigating IVDD of rabbit model. |
PDF Full Text Request