A Study On Method Of Quantitative Analysis In Computed Tomography Images Of Hepatic Sinusoidal Obstruction Syndrome And The Clinical Application

1 Background Sinusoidal obstruction syndrome(SOS),also known as hepatic veno-occlusive disease(VOD),is a disease caused by obstruction of outflow in sinusoids due to damage of sinusoidal endothelial cells.It is clinically characterized by ascites,painful hepatomegaly and even progressing to non-portal cirrhosis.Congestion and expansion of hepatic sinusoids,thickening,fibrosis,stenosis and obstruction of the hepatic venous wall which can be seen in liver biopsy.Ingestion of pyrrolizidine alkaloids containing products and hematopoietic stem cell transplantation are two common causes of hepatic sinusoidal obstruction syndrome.Quite a few of pyrrolizidine alkaloids related SOS have been suspected to be attributed to an exposure to Tusanqi(Gynura segetum),a traditional Chinese medicine.In the clinical investigation of 458 cases of hepatic sinusoidal obstruction syndrome organized by our research group,it was found that Tusanqi is the main cause of sinusoidal obstruction syndrome in China.CT examination is the main technical means for diagnosis of hepatic sinusoidal obstruction syndrome.Contrast agent is blocked in the end of portal vein branch and failed to enter or delayed enter into the liver lobe and segmental vein which can be seen in abdominal enhanced CT examination in portal vein phase and in equilibrium phase.And patchy liver enhancement and heterogeneous hypoattenuation were the two most unique manifestations in CT images.The uneven low-density sign or heterogeneous hypoattenuation is presented as patch-like distribution.Quantitative analysis of hepatic lesion volume in CT images in patient with PA-SOS by computer-assisted quantitative measurement,and analyze relationship with prognosis of patient with PA-SOS.Similar research has not been reported,and it is worth toconduct a research on it.2 Objective Based on CT images of patients with PA-SOS,the feasibility of measuring volume of hepatic lesions in CT images was explored,and then the measurements were compared with the clinical severity and prognosis of the patients.3 Materials and Methods This study is divided into two parts: the establishment of method for quantitative analysis of CT images in patients with PA-SOS,and the application of method in clinical work.The first part is based on CT images of 20 cases with disease,and explores the feasibility of the methods,Stereology and threshold based region growing algorithm,to quantitatively determine the volume of uneven low-density regions or heterogeneous hypoattenuation areas in CT images.The specific methods provides the basis of the clinical application and the superiority of the two methods were compared at the same time.The second part explores the relationship between measurements and the severity together with the prognosis of patients with PA-SOS based on threshold based region growing algorithm.The research contents are as follows:3.1 To retrieve CT images of 20 patients with PA-SOS confirmed by hepatic biopsy from the First Affiliated Hospital of Anhui Medical University,and explore two methods,Stereology and threshold based region growing algorithm,for quantitative analysis of heterogeneous hypoattenuation areas in CT images.Hepatic lesion volumes and operation time of 20 cases with the disease were analyzed,then the the advantages and disadvantages of the two methods were compared.3.2 Fifty-five consecutive patients diagnosed with PA-SOS were admitted to our hospital from December 1,2009,to July 31,2017.Patients who had other liver diseases or had incomplete clinical records and/or contrast CT examinations or patients who were lost to follow-up were excluded.Finally,25 patients with PA-SOS,including 12 males and 13 females,were enrolled.Heterogeneous hypoattenuation areas were analyzed by threshold based region growing algorithm in patients with PA-SOS,and the measurements were compared with the clinical severity and prognosis of the patients.4 Results The methods,Stereology and threshold based region growing algorithm,were performed for quantitatively measuring in CT images of 20 cases with disease,we found that the volumes of hepatic lesions were 1.001±0.327 L and 1.045±0.363 L respectively,with no significant difference(t=1.528,P > 0.05).However,the operation time of threshold based region growing algorithm(the shortest 500 s,average597.55±52.86s)was significantly less than that of the Stereology(the shortest 1075 s,average 1251.60±105.88s)(t=32.808,P < 0.05).Based on estimations using the threshold-based region growing algorithm of 25 patients with PA-SOS,hepatic lesion volumes in patients with mild PA-SOS were significantly lower than those in patients with moderate and severe PA-SOS(P<0.05).The ratio of hepatic lesion volume to liver volume(Ratio)varied based on clinical severity,and disease prognosis,and the differences were statistically significant(P<0.05).5 Conclusion5.1 Both methods,Stereology and threshold based region growing algorithm,may be applied to measure hepatic lesion volumes in CT images of PA-SOS.For measuring the lesion volumes,there was no significant difference between two methods.However,threshold based region growing algorithm may have more advantages as following: the less of the operation time,the clearer of the lesion segmentation and the simpler and faster of the operation.5.2 The ratio of hepatic lesion volume to liver volume of CT images,measured by threshold based region growing algorithm,in patients with PA-SOS is associated with a more serious clinical course and a poorer outcome.