Several Key Technologies Research In Multiple Dimensional Radiotherapy Based On Magnetic Reasonance Imaging

According to statistics from the world health organization,lung cancer is the leading status of morbidity and mortality in all malignant tumor.Radiotherapy is one of the important treatments.In recent years,radiation therapy technology has been developing rapidly,which mainly includes image guided radiation therapy(IGRT)and intensity-modulated radiation therapy(IMRT).However,therapeutic effect of lung cancer is not ideal,the main reason is tumor motion and heterogeneity.Respiratory movement can lead to tumor motion and is a complicating factor in radiation therapy.Respiratory organ motion can especially affect the radiation therapy,because it can lead to substantial degradation of the dose distribution.The existing approaches to take respiratory motion into account include choosing larger safety margins to ensure target coverage,voluntary or forced breath-hold,gating,or tracking of the target.Magnetic resonance imaging(MRI)can provide anatomical and functional information of the organization,and has no ionizing radiation.Therefore it becomes the research hot spot of IGRT.With the development of conformal radiation therapy and adaptive radiation technology,in addition to the tumor motion,the functional MR information is also worthy of being mentioned.Therefore,how to integrate the functional information of tumor tissue with radiotherapy which can improve the local control of the tumor,at the same time reduce the radiation injury,is becoming even more important.In conclusion,we need a robust model to accurately depict the respiratory motion and tumor motion in clinic,and reduce the injury of normal tissue in the process of radiation therapy.In order to predict the treatment sensitivity,evaluate the curative effect and show the injury of normal tissue,the analysis of functional image is used.In the future study,we hope the respiratory model can be used in the process of radiation therapy,which can reduce the dose of normal tissue received and boost the tumor received,eventually improve the tumor control probability.In this paper,we first carried out the research about respiratory motion based on MRI,summarized the existing motion model,put forward the improvement about existing models,established a noninvasive respiratory signal extraction model,retrospectively reconstruct Four dimensional MRI images(4D-MRI).In the next part,we used radiomics technique to analyze tumor heterogeneity,differentiate different pathological types of tumor and predict the curarative effect and overall survival.Research achievements and contributions are summarized as follows:Firstly,we developed respiratory signal recording system to search skin motion law through multiple points on the skin surface using dynamic 2D MR images.We use the location of multiple points on the surface to depict the respiratory status.Afterwards,we obtain 4DMR images retrospectively based on the respiratory status separation technique,and determine the position and shape of the tumor in different respiratory condition.We eventually developed a retrospective image reconstruction technique of 4D-MRI by recording the amplitude,velocity and tumor location.Secondly,we used radiomics technique to analyze tumor heterogeneity.We also did some research in the classification of the tumor by using the texture analysis technique based on MR images.In other study,we predicted the curative effect and overall survival for ESCC patients by using the same method on functional MR images.In the last part,we studied the influence of image intensity normalization methods on texture classification and explored the effects of slice thickness on the quantification of image features using a T1 post contrast MR.At the same time,we analyzed the stability about these texture features,and review the value of the texture analysis.