Research Of Using BOLD To Identify Benign And Malignant Tumor Of The Bone And Soft Tissue

Objective: Musculoskeletal tumors originate from bone or adjacent soft tissue and present in variable fashion on clinical manifestation and pathology. Given that the tumor of the bone and soft tissue bring vast adverse consequences and their clinical treatments are different, so differentiating malignant tumor from benign tumor is important and then assist in formulating an appropriate treatment strategy as well as determine the prognosis. Some immunohistochemistry studies suggest that the pattern of vascularity in benign and malignant tumor is different: the characteristics of the malignant tumor vessels are missing vessel hierarchy, blind-ending vessels, the increased vessel diameters and heterogeneous intratumor vascular densities, for example, the vascular density is higher in peripheral, the angiogenesis probably begins in venules and capillaries from the pre-existing vessels of neighboring normal tissue. The vessels are mostly mature and the blood supply is abound; however, the vascular density is lower in the center, there are mostly the neovascularity participated by the tumor cells. The vessels are immature and increased permeability, lacking the smooth muscle around the wall, so that it will result in inadequate vascular supply and hypoxia; In contrast, the vascular of benign tumor is mature, the density is homogeneous, that is, have well vascularized in both areas, and do not show different blood supply between the two areas. Some researches have confirmed the difference perfusion model of benign and malignant tumor. Besides, some animal or human tumor studies suggest that the reactivity of vascular to stimulation(angiotensin II or anesthetic) in malignant tumor is differ from normal tissue. In addition, BOLD-f MRI can demonstrate the spontaneous flow fluctuations. In this study, we use rest BOLD-f MRI to detect the solid tumor of bone and soft tissues, and analysis the feature of the signal to find out the difference between benign and malignant tumor.Methods: The 3D-T1 WI images and the BOLD-f MRI images of 92 patients with bone and soft tissue tumor were preformed by the 1.5 Tesla magnetic resonance scanner(Magnetom Avanto, Siemens, Erlangen, Germany). 3D-T1 WI images used the turbo FLASH(turbo fast low angled shot) squence with the scan parameters as follows: repetition time(TR)=1900 ms, echo time(TE)=2.97 ms, thickness/gap=1/0.5 mm, flip angle=15°, number of slices=176 slices, field of view(FOV)=220 mm × 220 mm, matrix=256×246. The SE-EPI-BOLD sequence was performed with the following parameters: TR=2000 ms, TE=400 ms, thickness/gap=5/1 mm, the voxel size=3.44 mm × 3.44 mm × 6.0 mm, FOV=220 mm × 220 mm, matrix= 64 × 64, number of slices=20, scanning time=6 min. Preprocessing and analysis of f MRI data was carried out using DPARSFA v2.3, REST v1.8,MICA toolbox based on SPM8 and Matlab 2012 a. After removing the unqualified cases, there were 74 patients, including 56 cases malignant tumors and 18 cases benign masses.Transformed the time series into the frequency domain with a fast Fourier transform(FFT), and then obtained the power spectrum, the power spectrum finally constituted of 96 frequency points: frequency point 1=0.0026×1 Hz, frequency point 2=0.0026×2 Hz, …, frequency point 96=0.0026×96 Hz. We completed the statistical tests by Two-Independent samples t test on power(amplitude) between benign tumor and malignant tumor, and then the ROC analysis was performed.Result: The result of the Two-Independent samples t test shows that the power(amplitude) between benign tumor and malignant tumor at frequency point 58(0.1508 Hz,P = 0.036), 59(0.1534 Hz,P = 0.032) and 95(0.247 Hz,P =0.014) have significant difference. In the later ROC analysis shows that the area of under curve is 0.706(diagnostic accuracy is medium, P=0.009), the cut off value is 0.73130, the value which is higher than 0.73130 is considered the possibility of benign tumor, on the contrary, less than this value is considered the possibility of malignant tumor, the diagnostic sensitivity and specificity values are 83% and 59%, respectively.Conclusion: At the frequency point 58(0.1508 Hz,P = 0.036), frequency point 59(0.1534 Hz,P =0.032) and frequency point 95(0.247 Hz,P =0.014), the power(amplitude) of tumor between benign and malignant has significant difference: the power of the benign tumor is higher than malignant tumor. The ROC analysis of the tumor power has a good sensitivity for differentiating malignant tumor from benign tumor. The result shows that BOLD-f MRI could detect the spontaneous blood flow fluctuations in the tumor. The feature of blood flow fluctuation in the malignant or benign is related to the blood supply model and the pathophysiological. Malignant tumor has heterogeneous vascular density, the neovascularization is immature and lacking the smooth muscle cells around the vascular wall, blood supply is poor. On the contrary, the benign tumor has homogeneous vascular density(the vascular of the periphery and the center has not obviously different), the vessel is mature. Because the vasomotion in the benign tumor is better than malignant tumor, and the blood supply is more adequate in benign tumor, so the blood flow fluctuation in the benign tumor is more obviously than malignant tumor. As a new technology, BOLD can provide the non-invasive, real-time imaging about the blood flow fluctuation. Furthermore, it has the high temporal and spatial resolution and the good repeatability, so we believe BOLD can be used in detecting the feature of blood flow fluctuation of malignant and benign tumor, and then proceed to differentiate malignant from benign tumor.