| Objective: To identify benign and malignant tumor of the bone and softtissue is very important, it relates to the treatment in clinical and determine theprognosis. The pattern of vascularity in benign and malignant tumor isdifferent. In the peripheral of malignant tumor, the angiogenesis probablybegins in venules and capillaries from the pre-existing vessels of neighboringnormal tissue. The vessels are mostly mature and the blood supply is abound.However, in the center of malignant tumor, there are mostly the neovascularityparticipated by the tumor cells. The vessels are immature and increasedpermeability, lacking the smooth muscle around the wall, so that it will resultin inadequate vascular supply and hypoxia. In contrast, benign tumor havewell vascularized central areas, the vascular is mature, and do not showdifferent blood supply between center and peripheral of the tumor. Someresearches have confirmed the difference perfusion model of benign andmalignant tumor. Blood flow of the tumor is instability. BOLD-fMRI candemonstrate the spontaneous flow fluctuations, and also can show the centerof the tumor has the lower response to the stimulation than the peripheral. Inthis study, we use rest BOLD-fMRI to detect the solid tumor of bone and softtissues, and analysis the feature of the signal to find out the difference betweenbenign and malignant tumor and the difference of BOLD signal in differentareas.Methods: The3D-T1WI images and the BOLD images of89patientswith bone and soft tissue tumor were preformed by the1.5Tesla magneticresonance scanner (Magnetom Avanto, Siemens, Erlangen, Germany).3D-T1WI images used the turbo FLASH (turbo fast low angled shot)squence with the scan parameters as follows: repetition time (TR),1900ms;echo time (TE),2.97ms; slice/gap,1/0.5mm; flip angle,15°; scanning layers,176slices; field of view (FOV),220mm×220mm.The SE-EPI-BOLD sequence was performed with the followingparameters: repetition time (TR),2000ms; echo time (TE),400ms; slice/gap,5/1mm; the voxel size was3.44mm×3.44mm×6.0mm; field of view(FOV),220mm×220mm; acquisition matrix,64×64; scanning layers,20;scanning time,6min.Using the MATLAB2010a software, the SPM8and the REST softwareperformed the data processing and remove the linear trends. After removesome limitations, there were48patients, including35cases malignant tumorsand13cases benign masses. Choose3ROIs in the center and the peripheral ofthe tumors of45patients and normal muscles of38patents respectively.Transformed the time series into the frequency domain with a fast Fouriertransform (FFT), and than obtained the power spectrum. Get the average ofthe3ROIs in the same area of any tumor. Divided the power spectrum into8frequency bands: band1,0~0.01Hz; band2,0.01~0.027Hz; band3,0.027~0.073Hz; band4,0.073~0.198Hz; band5,0.198~0.25Hz; band6,0~0.1Hz; band7,0.01~0.08Hz; band8, the whole frequency band,0~0.25Hz. Take the total power of1~7frequency bands divided by the wholefrequency band power respectively. And take the mean of the power of thewhole frequency band. We completed the statistical tests by the ANOVA andPost hoc pairwise comparison analysis. Furthermore, comparing the differencecentral power and peripheral power and the difference△P ofperiphery-center power of the tumor between malignant and benign tumor wasperformed all by Two-Independent samples t test or Two-Independent samplesNonparametric test. In this study two experimenters (wang and zhao) chosethe ROI according to the specific principles respectively. Wang had twooptions (wang1, wang2) week apart, independent of each other. Thencalculated the value of intrarater correlation coefficient (wang1and wang2)and interrater correlation coefficient (wang1and zhao; wang2and zhao) toevaluate the repeatability of choose of ROI.Result: The result of the ANOVA analysis shows at the frequency bands of0.073~0.198Hz (P=0.017),0.198~0.25Hz (P=0.024) and the wholefrequency band (0~0.25Hz, P=0.023), power of ROI in the center, theperiphery of the tumor and the normal muscle has the significant difference. Inthe Post hoc pairwise comparisons analysis, in the malignant tumor, at the0.073~0.198Hz and the whole frequency band, central power of malignanttumor is lower than peripheral (P＜0.05, corrected), and the peripheral powerhas the trend to higher than the power of the normal muscle (P＜0.05uncorrected, P＜0.1corrected). At the0.198~0.25Hz, the peripheral power ofmalignant tumor has the trend to higher than the central power of malignanttumor and the normal muscle (P＜0.05uncorrected, P＜0.1corrected).However, in the benign tumor, at the above three frequency bands, there is nosignificant different of the power among the center and the periphery of thetumor and the normal muscle. Furthermore, in all the frequency bands, themean power of the center and periphery has no significant difference betweenthe malignant and the benign tumor. The same as the central power, theperipheral power and the difference power (△P) of the periphery-center of thetumor. ICC analysis of repeatability shows that the value rang of intraratercorrelation coefficients of wang1and wang2is0.5183~0.9307, the meanvalue is0.6542, the repeatability is fair to excellent. The value of interratercorrelation coefficients (wang1and zhao; wang2and zhao) also show therepeatability is fair to excellent, but the reliability is lower, they are0.3901~0.8085and0.3372~0.8673respectively.Conclusion: At the0.073~0.198Hz,0.198~0.25Hz and the wholefrequency band, power of different regions has significant difference. Centralpower of malignant tumor is higher than the peripheral of the malignant tumorand the normal muscle. But in the benign tumor it do not show the difference.The central power, the peripheral power, the mean power and the differencepower of periphery-center all do not have significant difference betweenmalignant tumor and benign tumor. The result shows that BOLD could detectthe spontaneous blood flow fluctuations in the tumor, and effective to highfrequency fluctuation. The feature of blood flow in different regions in the malignant or benign is related to the blood supply model and thepathophysiological. Malignant tumor has the rich blood supply in theperiphery, the vascular is mature and has the smooth muscle around. Bloodsupply of the center is poor, the neovascularization is immature, lacking thesmooth muscle around the wall. Because the vasomotion in the periphery isbetter than center, and the blood supply is more, the blood flow fluctuation inthe periphery is more obviously than center. In the benign tumor, the vascularof the periphery and the center has not obviously different, the vessel is matureand has the smooth muscle. Furthermore, no neovascularization is formed inthe benign. The blood flow fluctuation between center and periphery of thebenign tumor shows no significant difference. ICC analysis shows therepeatability of this study is fair to excellent. As a new technology, BOLD canprovide the non-invasive, real-time imaging about the blood flow fluctuation.Furthermore, it has the high temporal and spatial resolution and the goodrepeatability, so we believe BOLD can be used in detecting the feature ofblood flow fluctuation of malignant and benign tumor. |
PDF Full Text Request