Real Spatial Morphology Reconstruction Based On An Electromagnetic Intravascular Ultrasound Composite Catheter And The Accuracy Validation By Comparative Histology Study With A Self-developed High Throughput Software

Atherosclerotic plaque rupture and subsequent thrombosis are the main causes of acute cardiovascular disease events.Intravascular ultrasound is important for the identification of vulnerable plaques.In the intravascular ultrasound imaging technique,all cross-sectional information of the entire vessel segment can be observed.But comparing the IVUS image and histological results will involve a whole length of the arteries with large amounts of images and slices,and the workload is huge.In the histological study of atherosclerotic plaques,a colorant is usually used as a component for staining.At present,most commercial software provides a macro histogram analysis tool.Even in this batch process,the parameters include the threshold of the pathological signal color,the ROI boundary,and so on,it is still necessary to manually adjust the optimal result for each different case.Obviously,a large number of histogram analysis systems to minimize human intervention will be of great help to improve work efficiency and reduce subjective factors.In this study,we use the clustering analysis principle and Lab color space,and the Matlab environment to develop a high-throughput tissue slice image analysis software,and verify the accuracy of the software by comparing with image-pro plus software analysis results.2.Methods2.1 Software developmentThe software was developed in the Matlab environment.The boundaries of the arterial intima and the outer membrane were manually delineated and stored.The color of each pixel was dispersed to the CIE-LAB color space.Based on the components a and b,the k-means clustering The analysis method will cluster each pixel.The best clustering number of k is set by comparing the results of a cluster based on a default range from 3 to 5 in the contour analysis.2.2 Animal modelTen male New Zealand white rabbits weighing 1.5 to 2.0 kg were fed with 1%high cholesterol diet for 1 week.After anesthesia with 3%pentobarbital sodium(1 ml/kg),the endothelium of the abdominal aorta was denuded by inflated balloon.The high cholesterol diet was continued to 24 weeks.2.3 Histology and immunohistochemical stainingAll tissue sections were made with a thickness of 5 um.Oil Red O staining,Sirius-Red staining,anti-a-SMC staining and anti-RAM-11 staining were performed to illustrated the lipid,collagen,smooth muscle cells and macrophage components respectively.10 and 20 magnitude microscope images were acquired and stored in TIFF format.2.5 Statistical analysisThe percentage of collagen fibers,smooth muscle cells,lipids and macrophages in the plaques were measured for each group of samples using IPP software and cluster anslysis-based software.Quantitative data were expressed as mean ± standard error,and SPSS13.0 software was used to analyze the data by independent sample t test or single factor analysis and Newman-Keuls test,P<0.05 was statistically significant.3.resultsA total of 50 plaque samples were collected from the abdominal aorta from 10 rabbit models.The samples were stained with oil red O staining,Sirius red staining,anti-a-SMC staining and anti-RAM-11 staining,and observed at 10 and 20 magnitude microscopy in turn.Image-Pro Plus software and the self-developed software measurements correlate to each other very well.In this experiment,the appropriate cluster analysis classification number is:oil red O(10X),oil red O(20X),Sirius red(10X),Sirius red(20X),anti-smooth muscle cell(20X)are 3;anti-RAM-11(10X),anti-smooth muscle cells(10X)are 5;and anti-RAM-11(20X)is 5.That is,for each staining,accurate clustering analysis results can be obtained by selecting the appropriate color classification(usually 3 to 5,but not limited to 3 to 5).Bland-Altman and linear regression analysis further confirmed that the two analysis software have a good correlation with the correlation coefficient R2 between 0.72-0.99.4.Conclusion:In this study,we successfully developed the software for automatic analysis of atherosclerosis tissue slices based on cluster analysis and Lab color space.By selecting the appropriate number of color analysis,accurate results can be obtained by the self-developed software.The developed software reduces the human interference factors,improve the efficiency,and is suitable for a large number of analysis of pathological sections.Recent studies have validated that aherosclerosis(AS)plaque rupture and the subsequent thrombosis is the main pathogenesis in most acute cardiovascular and cerebrovascular events.AS plaques occurredwithin the arterial wall.So,techniques like coronary angiography illuminated only the arterial lumen can not meet the clinical needs.Intravascular ultrasound imaging(IVUS)makes up the drawbacksof coronary angiography.It shows the plaque morphology,plaque load,and the composition of the plaque.As the two-dimensional vascular cross-sectional images can only provide local disease information,3D reconstruction of intravascular ultrasound images has become the current research focus.The prevalentreconstruction algorithm overlaps series of IVUS images acquired by automatic pullback of the catheter,which results ina straight tube model.This method does not take into account the curvature of the blood vessel itself during image acquisition.Another method is to use the multi-angle CAG to extract the vascular skeleton and sequence the IVUS images.However,based on the reconstruction of the multi-angle CAG image,it is generally assumed that the shape of the lumen is elliptical.Impeding the registration of DSA images from IVUS images.These reconstruction methods also ignore the warping of the image plane caused by the off-plane motion in the process of vessel skeleton retraction,which affects the accuracy of the reconstructed image,especially the virtual endoscopic environment.In this study,we put forward a new method of 3D IVUS reconstruction based on electromagnetic positioning information.Firstly,the IVUS image is transformed into three-dimensional coordinate system according to the information on the image coordinates and angle obtained by the electromagnetic positioning system.Then,the outlines of the intima and adventitia of the artery were extracted.By fusionof the spatial information obtained by the electromagnetic positioning system and the lumen outline information obtained from the IVUS imags,the accurate three-dimensional reconstruction of the artery of interest can be realized,and the diagnosis and treatment of the vascular disease.Purposes:1.Developmentthe composite IVUS catheter based on electromagnetic positioning and tacking system.2.To validate the accuracy of atherosclerotic plaque volume quantification by the developed composite catheter.Methods:1.Twenty-fivevessel phantoms were prepared using PVA cryo-gel(PVA-CG).Eleven AS plaques were harvested from Atherosclerotic rabbit abdominal aorta.2.Development of the composite catheter:the electromagnetic coil was installed at 0.2 mm below the IVUS probe and fixed to the outer sheath wall of IVUS catheter.The two parts were combined into a composite catheter in the composite catheter jacket sheath.IVUS and electromagnetic coil are connected to the host separately,but they can be advancedor retreated simultaneously.3.Accuracy verification3.1 Verification of Electromagnetic Positioning AccuracyThe vibrating of the electromagnetic coil in static state at different distance to the magnetic field generator was recorded.The dynamic displacement was generated by displacement platform.The results show that with the increase of the distance between the coil and the magnetic field generator,the vibrating amplitude increases.The maximum vibrating amplitude of less than 0.155mm could be maintained with the noise reduction algorithm at the distance less than 40cm from the magnetic field generator.In the dynamic state,excellent agreement between the electromagnetic system and the displacement platformcould be obtained when distance to the magnetic field generator is less than 30cm.The Bland-Altman plot also confirms that 95%of the confidence interval(LOA CI)at 15 cm,30 cm from the magnetic field,and the confidence interval does not exceed the acceptable threshold range,ie 0.2mm.3.2in vitro experiments:3.2.1 phantom experiment:The 7F sheath was installed from the sealing valve mounted on a tank filled with deionized water..The ends of the vascular model were fixed to the sheath.The phantom was wound on a cylinder with a diameter of 84 mm.With the guidance of the 0,014mm guide wire,IVUS composite catheter was inserted into the phantom through the arterial sheath.The IUVS and the electromagnetic positioning information are integrated by software,and the three-dimensional structure of the blood vessel is obtained.The volume of the vascular model was measured.3.3Atheroscleroticrabbit model:3.3.1 The atherosclerotic rabbit model was established by denudation of the endothelium of the abdominal aorta by an inflated balloon and the high cholesterol diet for 24 weeks.3.3.2Separation of the abdominal aorta,ligation of abdominal aortic collateral,and then wrapped around a cylinder with diameter of 84mm.With the guidance of the 0.014mm guide wire,IVUS composite catheter was inserted into the artery through the arterial sheath.The electromagnetic tracking information and IVUS images are integrated to obtain three-dimensional information.3.3.3 Pathological staining:the above blood vessels were cut into 5mm length segments,and fixed by the 4%paraformaldehyde,4 ? 24-48 hours.Oil red O and picorus Sirus Red staining,and anti-macrophage,anti-actin immunohisotological staining were applied.3.3.4The volume of atherosclerotic plaqueswere calculated based on histology and the 3D reconstruction.3.4Statistical analysis:SPSS 13.0 software and MedCalc software were used for analysis.The data were expressed as mean ± SD(mean ± SD),and P>0.05 was not statistically significant.Results:(1)composite catheter examination can accurately measure the volume of atherosclerotic plaque.There was no significant difference in the volume of atherosclerotic plaque between the estimation by 3D reconstruction with the composite catheter and the phantom and histology(P>0.05).(2)The 3D reconstruction by composite catheter can show the true spatial morphology of the blood vessel,and can realize a multi-angle observation and positioning of the vascular lesion conveniently.Conclusion:(1)The volume of atherosclerotic plaque can be calculated accurately by 3D reconstruction with the developed composite catheter.(2)3D reconstruction with the developed composite catheter can acquire the intravascular ultrasound images and three-dimensional spatial information simultaneously.