Digital Hepatic Vein Typing And Its Application In Liver Surgery

Research background:As early as1952, some scholars emphasized the importance of hepatic veins during liver surgery and the necessity for protection of the hepatic veins in partial hepatectomy. Since1957, the importance of hepatic venous drainage has come into minds of the surgeons in the discipline, for the serious consequences from hepatic venous obstructions induced from the venous drainage after partial hepatectomy.It is crucial to ensure the smooth blood circulation in hepatic veins of the remained liver. The obstructed veins contribute greatly to the surgical complications like postoperative delayed bleeding, hepatic functional disorders and bile leakage. Moreover, the absences of latent communicating branches at the base of hepatic veins in the major population result in the obstruction in venous drainage and the impossibility to relieve relevant complications by way of the latent branches.The impact of hepatic venous outflow obstruction on hepatic function from hepatic surgery is least understood at present. Maintenance of adequate venous outflow is widely recognized to be essential in avoiding hepatic dysfunction or failure from various degrees of hepatic venous outflow obstruction. However, what constitutes adequate hepatic venous outflow after liver surgery remains controversial. In order to prevent postoperative liver failure, normal liver tissues need to be reserved especially in particular cirrhosis patients. The inadequate residual hepatic venous return may lead to such complications as the occurrence of liver tissue congestion in remote lesions and hepatic cellular necrosis. The resection of normal liver tissue beyond the lesions, however, may increase the risk of liver failure. Therefore, the preoperative understanding of the individual variations of hepatic vein in patients is beneficial for planning the surgical operations, reserve the normal tissues to the most extent and reduce postoperative complications.In this study, the individual variations of hepatic veins were anatomically investigated using Medical Image3D Visualization System (MI-3DVS) by way of analyzing and treating epigastrical CT imaging data of200volunteers and80hepatic cancer patients. In return, the acquired results in the individual variations were applied in the simulated surgical system for diagnosis of hepatic cancer as well as operation planning. The model of individual variations has its values clinically for diagnosis improvement of hepatic carcinomas, direction for actual manipulations of hepatic surgery, conservations of more normal hepatic tissues and reduction of postoperative complications.Part1.The value of individual hepatic vein reconstruction by MI-3DVS and its clinical significanceObjective1. To study the characteristics of individual digitized hepatic vein imaging by MI-3DVS reconstruction;2. To investigate the normal and variant manifestations of digital hepatic veins and the practical significance;3. To explore the value of the normal and variant manifestations of individual digital hepatic vein in liver surgery.Methods 1. Objects and Materials:(1)200upper abdominal CT scan examination of healthy volunteers,102men, women98persons, with an average of22.6years;(2)①PHILIPS BRILLIANCE64-slice helical CT, Image process workstation;②binocular tube high pressure injector and contrast agent;③computer;④DICOM Viewer;⑤) ACDSee;⑥Medical Image3D Visualization System (MI-3DVS);⑦FreeForm Modeling System and PHANTOM.2. Imaging scan parameters:The CT scanner settings included a tube tension of120KV, a tube current of300mAs, a layer thickness of1mm, an interval of1mm, a pitch of0.984, a rotation speed of0.5S/cycle and the scanning horizon of40-50cm and matrix of512×512.3. Data collecting:Every volunteer drank500～1000ml water20-30min before examination and500ml just before examination to engorge gastrointestinal tract (as negative contrast agent). All the volunteers took breath training for controlling artifacts from respiration movement.(1) Plain scan:We used volume scanning with high resolution on submillimeter condition. During routine epigastrium plain scanning, volunteers lay supine and were scanned from diaphragm to pelvic cavity in the orientation from head to foot with the current at300mAs, the voltage at120KV, the thickness at5mm, the interval at5mm, the pitch at0.984, the rotation speed at0.5S/cycle, the scanning horizon at40-50cm and the matrix at512×512.(2) Dynamic enhanced CT scan:Pre-injection test at small dose:The contrast agent heated to37℃was injected with type20trocar from tube A of a double tube with a high pressure syringe via cubital vein at the speed of5ml/s. The dynamic scanning on the first hepatic hilar region was done within30s, with the current at300mAs, the voltage at120KV, the thickness at5mm, the interval at5mm, the pitch0.984, and the rotation speed at0.5S/cycle so as to attain time density curve of abdominal aorta and detect peak value of CT value as starting time of CTA scan. For CTA scan,70-120ml contrast agent (1.5ml/kg) was injected from pipe A at the same rate, then with20ml NS injected from pipe B. CTA scanning was started with the predict time, the volunteer asked to hold their breath (about6-8s). The CT scanner settings:the current at300mAs, the voltage at120KV, the thickness at5mm, the interval at5mm, pitch at0.984, and rotation speed at0.5S/cycle. Scanning was delayed20to25S after arterial phase,50to55S after venous phase. In order not to affect routine diagnosis, late artery scanning was carried out30-35s after injecting contrast agent and portal venous phase scanning was done50-55s after injection,6-8s for each period.4. Image Reconstruction:The original data was uploaded into the personal computer and translated into JPG format through DICOM viewer; then the JPG format was translated into JPG format, with the image size adjusted from512px×512px into304px×304px by ACDSee. The adjusted images were imported into Medical Image3D Visualization System (MI-3DVS) for procedure segmentation and three-dimensional reconstruction and then saved in a file of STL format. In the following, STL format documents were imported into FreeForm Modeling System for eventual reconstruction of hepatic veins by way of smoothening, denoising and digital painting.5. Statistical treatments:This study mainly involved morphological interpretations, but little statistical inference.ResultsHepatic vein reconstruction model, solid, distinct and vivid, presented the prevailing pattern of the left, middle and right hepatic veins thoroughly. In the reconstructed veins from200cases, the majority of left and middle hepatic shared the same trunks (122/200,61%) and for the rest cases, the left, middle and right hepatic veins respectively converged into inferior caval veins. In the reconstructed veins from the200cases, the right hepatic veins in78cases (39%) were viewed to be small and short, with outflows from other surrounding hepatic veins, though:42of them (21.0%), for example, manifested with greater posterior hepatic veins,17(8.5%) with right accessory hepatic veins and13(6.5%) with compensatory outflows of right liver blood from middle hepatic veins.103of them (51.5%) presented section VI hepatic veins:The segment VI veins in52of them (26%) ran into left hepatic veins, those of35(17.5%) into middle hepatic veins and those of16(8%) directly into inferior vena cava.Conclusions1. Reconstructed models of hepatic veins by MI-3DVS based on64-slice helical CT hepatic vein scanning data are clear, solid and vivid. It could provide a good platform for individual hepatic segmentation, preoperative planning, surgical rehearsal, assessment of surgical risk, selection of surgical approaches and clinical teaching.2. Attention should be paid to the venous variations like presence of section IV hepatic vein or greater posterior right hepatic veins in liver surgery. The reconstructed3D model based on CT scanning data may provide critical information concerning the venous hepatic variations.Part2The application of the MI-3DVS in the reconstruction of hepatic veins and the value in liver surgeryObjectiveThe individual hepatic veins reconstructed from the CT images of200cases were used for individual hepatic segmentation based on the hepatic variations and volume measurement of primary liver cancer, so that the surgical planning was made after assessing the resectability of the space-occupying hepatic lesions based on the reconstructed hepatic model. The value in clinical application was assessed from the angle of conserving the normal liver tissue to the largest extent and preventing liver dysfunction resulting from obstructions of hepatic veins outflows.Methods1. Objects and Materials:The imaging data of80liver carcinoma patients during November2009to April2012in our hospital received the abdominal64-slice spiral CT enhanced scanning, male57, female23. aged between24-81years, averaged47.3years.66of them were diagnosed with liver cancer,10with hemangiomas of liver, and4with hepatic nodular hyperplasia. Materials were the same as in the first chapter.2. Data Collecting:Every patient drank500～1000ml water20-30min before examination and500ml just before examination to engorge gastrointestinal tract (as negative contrast agent). All of them received breath training to avoid artifacts from respiration movement to the greatest extent.(1) Plain scan:Submillimeter volume scanning with high resolution was performed, the patients in a supine position and the scanning from diaphragm to pelvic cavity in the orientation from head to foot, with the current at300mAs, voltage at120KV, thickness at5mm, interval at5mm, pitch at0.984, rotation speed at0.5S/cycle, scanning horizon at40-50cm and matrix at512×512.(2) Dynamic enhanced CT scan:Small dose injection test:Headed to37℃,20ml contrast agent from pipe A of double tube high pressure syringe was injected with type20trocar into cubital vein at the speed of5ml/s. The scanning was performed after injection of contrast agent for30s with current at300mAs, voltage at120KV, thickness at5mm, interval at5mm, pitch at0.984and rotation speed at0.5S/cycle, so as to attain time density curve. The peak value of aorta abdominalis CT value was set as start time for CTA scan. After70-120ml contrast agent (1.5ml/kg) was injected from pipe A at the same speed, followed by injection of20ml NS from B pipe, CTA scanning was started at the predicted time, with the patients asked to hold their breath for about6-8s. 3. Methods for image reconstruction:The data from different phases (arterial phase, portal vein phase, venous phase and lag phase) were uploaded into the personal computer respectively and translated into JPG format through DICOM viewer; then the JPG format was translated into BMP format for adjusting the size of images with ACDSee. The adjusted images were imported into Medical Image3D Visualization System (MI-3DVS) for procedure segmentation and three-dimensional reconstruction. The reconstructed models were saved as STL format and then STL models were imported into FreeForm Modeling System for smoothening, denoising and digital painting.4. Diagnosis and resectability assessment:Preoperative rotation, enlargement and transparency of the reconstructed models were done to investigate the three-dimensional model of the individual liver as well as its internal lesions. From it, experienced hepatobiliary surgeons made the diagnoses based on the medical history. In accordance to the characteristics of hepatic and portal veins as well as’the presence of self-sustaining blood supply and blood outflows’, in the following, the livers were segmented and the practical surgical course was simulated based on the reconstructed3D liver model to measure the lesions, normal hepatic tissues and the residual liver after liver surgery. Finally, the venous variations in the infected livers, individual segmentation of liver according Couinaud segmentation, the percentages of residual liver by simulation surgery in the normal liver were worked out from the model.5. Comparisons with actual surgery:The space-occupying lesions in all the cases were explored and resected by experienced liver surgeons. The actual operative program was compared with the preoperative diagnosis and simulation surgery plans.6. Statistical analysis:This study mainly involved morphological interpretations, but little statistical inference. Results1. Individual digital hepatic vein and liver tumors:The reconstructed liver model with hepatic vein and internal lesions by MI-3DVS vividly presented not only the forms and ranges of the tumors and the adjacent normal hepatic tissues but also their clear adjacency with branches of hepatic veins.2. Hepatic vein variations in the affected livers of patients:The reconstructed model of hepatic veins was clear, vivid and solid. In the reconstructed hepatic vein models of80patients, left hepatic vein and middle hepatic vein shared one vessel by72.5%and in the rest, left, middle and right hepatic veins respectively converged into postcava. In the reconstructed veins from the80cases, the right hepatic veins in18(22.5%) were viewed to be small and short, with outflows from other surrounding hepatic veins, though:13of them (16.25%), for example, manifested with greater posterior hepatic veins and5(6.25%) with right accessory hepatic veins. Totally13(16.25%) had section FV hepatic veins, among which6(7.5%) entered left hepatic vein,4(5%) entered middle hepatic vein, and3(3.75%) directly enter postcava.3. Individual digital liver segmentation:The3D models of liver, hepatic vein and portal vein were displayed to perform individual hepatic segmentation according to the anatomical features of hepatic and portal veins as well as the specific changes from the space-occupying lesions. From the models, the liver was segmented into8segments in64cases, which accorded to Couinaud hepatic segmentation, into9segments in7cases and into10segments in3cases due to presences of section IV veins and posterior right hepatic veins and branching of right hepatic veins at the base, into7segments in1case due to atrophy or nonvisulization of hepatic and portal veins from intrahepatic space-occupying and crushing, into6segments in3cases and into5segments in2cases.4. Preoperational simulation operation and measurements of individual livers: Before operation, the3D data were used for respectability assessment. In case of positive respectability and patients’consents, the simulation operations from the models were performed, with lesions measured, the volume of normal liver and the percentage of residual liver in the normal one computed. Of the80case,17cases did not received surgical operations because the3D model showed infeasibility for hepatectomy due to invasion of the tumor into the great vessels or because the patients refused hepatectomy. The other63received operations, the results showing the normal liver ranged between785.1ml and2024.1ml in volume, tumors ranged between3.9ml and1163.6ml and the percentage between the residual liver and the normal part between33.5%-93.8%. No severe postoperative complications such as liver failure occurred.Conclusions1. Reconstructed models of hepatic veins by MI-3DVS are solid, vivid and clear. The anatomical structures of livers and internal ducts presented in the models could reflect the in vivo human hepatic construction. Therefore, it is of value for the practical use in anatomical research, preoperational diagnosis, selection of surgical approaches and clinical teaching.2. The three-dimensional models reconstructed by MI-3DVS can accurately directly present internal hepatic lesions as well as the adjacency of normal liver tissues and hepatic venous branches. It has great significance in maximizing the retention of normal liver tissues at resecting tumors, reducing hepatic failure and promoting the postoperative recovery of hepatic function.