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

Objective: Bone tumors have the origin from the bone or its subsidiary organizations (blood vessels, bone marrow and nerves, etc.). Soft tissue tumors have the origin from the mesothelium, muscle, fat, the mesoblastic fibrous tissue, vessel and synovial membrane. Identifying benign and malignant tumors of the limb bone and soft tissue is very important to clinic, and many scholars have used many different methods. For example, DTI has been used to compare fractional anisotropy (FA) and volume ratio anisotropy values (VrA) of various tumors'central and peripheral regions to identify. Using CT or MR enhancement perfusion images, they compared the perfusion ratios on the central and peripheral regions between the benign and malignant tumor. BOLD-fMRI was first used on the cerebral function and cognitive neuroscience research, and in the recent years its applications are becoming more extensive such as evaluating oxygenation level of different typical tumor on different states, as well as studying BOLD signal changes of the limb skeletal muscle system on different states. Using a sleeve tourniquet to compress the proximal ends of the limb bone and soft tissue tumors to change the tumors'blood supply within a short time, we scanned tumors with continuous resting state, transient ischemic state and reactive hyperemia state 6 minutes respectively through using BOLD-fMRI to observe if there are different BOLD signal changes in three states between the benign and malignant tumors, as well as in the different regions of the same tumor.Methods: 46 patients (31 males, 15 females) with limb bone and soft tissue tumor were preformed by BOLD-fMRI with Germanic Siemens 1.5T superconductive MRI machine (Siemens, AVANTO). Using the sleeve tourniquet to compress and slack rapidly the proximal ends of the tumors, the blood oxygen saturations of the tumors and the surrounding regions are changed. The lesions were scanned by BOLD-fMRI in resting-state, transient ischemic state and reactive hyperemia state for 6 minutes respectively. BOLD-fMRI scan parameters: TR 2000ms, TE 400ms, FOV: 220mm×220mm, slice: 5mm, interlaminar distance: 20%, scanning layers: 20, matrix 64×64, voxel size 3.4 mm×3.4mm×5.0mm, scan time: 6 minutes (177 time points). We preprocessed the data through SPM5 (Statistical Parametric Mapping, http://www.fil.ion.ucl.ac.uk/spm/) software, including the corrections of slice timing and realign. We excluded the data whose mobility is greater than 3mm. By MATLAB software we calculated the signal ratio ofâ–³SI (â–³SI₁ = the signal intensity ratio of the BOLD signal intensity of ischemic to resting state,â–³SI₂ = the signal intensity ratio of BOLD signal intensity of reactive hyperemic to resting state) of the central and peripheral regions'ROI of the different tumors in 6 minutes (177 time points). We obtained the curve connected through 177 time points in the ischemic and reactive hyperemic periods. We obtained the peak time of theâ–³SI₂ in every ROI of the reactive hyperemia to calculate ?SI₂ peak time difference (?T) of the tumor center -periphery. We analyzed the BOLD signal characteristics of the benign and malignant tumors in the resting state, transient ischemic state and reactive hyperemic state. SPSS16.0 software was used for statistical tests to the obtained data.Results: There is difference in the ?SI₁ signal curve in different tumors or in different areas of the same tumor or in the normal tissues in the ischemic period. The curves of normal tissues are descended gradually or descended gradually after fluctuation. They were significant differences on ?SI₂ peak time between the tumor central regions, the peripheral regions and the surrounding normal tissue areas respectively in the reactive hyperemic period (P value= 0.0038, 0.000, 0.003 respectively),it is the most significant between the tumor central regions and the surrounding normal tissue area. There were significant differences on ?SI₂ peak time between the malignant tumor central regions, the peripheral regions and the surrounding normal tissue area respectively(P value= 0.0017, 0.000, 0.023 respectively),it is the most significant between the central regions and the surrounding normal tissue areas. There were significant differences on ?T of the center- periphery region, center- surrounding normal tissue area between the benign and the malignant tumor (P value=0.002, 0.005 respectively). There were significant differences of the BOLD signal coefficient of variation under quiescent condition between the tumor central region and the surrounding normal tissue, the tumor peripheral region and the surrounding normal tissue(P value=0.006, 0.017 respectively). There were significant differences on the coefficient of variation of BOLD signal in the quiescent condition between the malignant tumor central region and the surrounding normal tissue, the malignant tumor peripheral region and the surrounding normal tissue(P value=0.006, 0.0105 respectively), as well as between the benign and malignant tumor peripheral region (P value=0.038). There were no statistics differences of others.Conclusion: The results show that there were significant differences of BOLD signal changes under reactive hyperemia between the central region, the peripheral region and the surrounding normal tissue of the limb bone and soft tissue tumors respectively. There were significant differences on ?SI₂ peak time between the malignant tumor central regions, the peripheral regions and the surrounding normal tissue areas respectively. There were significant differences on ?T of the center- periphery region, the center- surrounding normal tissue areas between the benign and malignant tumor. There were significant differences on the coefficient of variation of BOLD signal in the quiescent condition between the tumor central regions and the surrounding normal tissues, the tumor peripheral regions and the surrounding normal tissues. There were significant differences of the coefficient of variation of BOLD signal in the quiescent condition between the malignant tumor central regions and the surrounding normal tissues, the malignant tumor peripheral regions and the surrounding normal tissues, as well as between the benign and malignant tumor peripheral regions. These may be related to the seldom stable states of the tumor oxygenation and the different changes of various tumor oxygenation in resting state. BOLD-fMRI using the body's own hemoglobin as contrast agent to achieve non-invasive effects and higher time resolution to provide a new way to identify the limb bone and soft tissue tumors for the clinic in the future.