Virtual Imaging In Medicine And "Virtual Biopsy" Techniques In The Noninvasive Diagnosis Of Lung Cancer

Part â… : Clinical application of helical CT virtual bronchoscope and related virtual imaging in central lung cancerObjective: To evaluate the clinical application of CT virtual bronchoscopy ( CTVB ) combined with other related image reconstruction in the diagnosis of central lung cancer .Materials and methods: 3 healthy volunteers and 41patients of central lung cancer were examined with multirow helical CT, Then translated the pictures and the data to the workstation and obtain the real-time virtual bronchoscopic pictures by reconstruction. Results of CTVB was compared with the findings in fiberoptic bronchoscopy(FOB) , operation and the pathologic results.Result: The findings of tracheal wall, lesions in lumen and tracheal lumen from CT virtual bronchoscopy were similar to the results of FOB. Furthemore CTVB was better in revealing the distant broncheal obstruction. 3 healthy people had normal finding in CTVB as in FOB. 41 patients had a total of 52 abnormal FOB findings.The overall sensitivity of CTVB was 84%.The sensitivity of CTVB for the detection of obstructive,endoluminal and mucosallesions were 100%,89% and 0%, respectively. CTVB found 7 lesions which FOB could not find because of size.Conclusion: Virtual bronchoscopy together with additional related 3D/2D imaging may be a promising and noninvasive modality for identifying bronchial obstructions and endoluminal lesions,as well as demonstrating the extraluminal extension of tumors. Combining the advantages with accurate of 2D and intuition of 3D ,CTVB is a important noninvasive technique in examing central lung cancer.Part II: Diagnosis value of Multi-slice spiral CT 2D and 3D reformation methods in peripheral pulmonary carcinomaObjective: To assess the diagnosis value of multi-slice spiral CT(MSCT), two-dimension and three-dimension reformation methods in peripheral pulmonary carcinoma(PPC) , and through the description of bronchus sign(BS) in PPC, to evaluate the contribution of bronchus sign -guided bronchoscopic multiple diagnostic procedures(BMDPs) in diagnosis of peripheral lung cancer.Materials and Methods : A prospective study was carried out in 63 lung cancer patients with a 2-6cm diameter solitary nodules or masses . Theyunderwent multi-slice spiral CT(MSCT). Images were reconstructed withMPR (CPR) to evaluate the relationship between PPC and bronchi and identify the type of BS. in each of 60 patients,bronchial brushing(BR), transbronchial lung biopsy(TBB),transbronchial needle aspiration(TBNA) and bronchial washing(BW) were done respectively . In 43 of 63 patients, we assessed together with the 3D reconstruction image (mainly including SSD).Result: The combined diagnosed yield of BMDPs reached 59%. In 27 of 34(79%) with BS and in 10 of 29 without BS. Fiberoptic bronchoscopy established the diagnosis(p<0.05).The diagnostic yield of TBB and BR was significantly higher in nodules characterized by BS type I and type II .While TBNA was higher in BS type III and type VI than TBB and BR in BS type I and type II(p<0.05%).The larger the diameter of nodules.The higher the diagnostic yield(p<0.05). All in all, the 3D image is direct which make them wholly apprehended while the 2D image clearly displayed the endoluminal lesions as well as extraluminal extension of tumors and adjacent structures of PPC.Conclusion: These results confirmed CT BS can be utilized as a guiding sign to direct FB diagnostic procedures and enhanced the yield of peripheral lung cancer. MSCT and its reformation methods provide more malignant messages than conventional CT in confirming the diagnosis of lung neoplasms.part III: Noninvasive Virtual biopsy of pulmonary neoplasms : MSCT image reconstruction and its image fusion with PET.Objective: To provide new diagnose information for clinican by using medical image fusion techniques: MSCT image reconstruction matching with PET.Materials and Methods: A prospective study was carried out in 47 patients with a 2-12cm diameter pulmonary lesions on CT scans of chest . Each of 47 patients was checked by CT and PET within 2 weeks separately . Afterward the CT images were transferred to the PET workstation with CDR .A fusion method was then developed to match the slices of volume PET images with the correlative slices of CT images, and we got the fusion images of CT and PET by fusion software .Results: 54 positive lesions on PET images were accurately located on the corresponding CT imags of 36 lung cancer patients, 15 lesions of which are neglected in CT reports . In addition,axial CT detected 5 lesions of 2 patients which PET could not display or with low uptake but were finally proved as benign. Contrasted with the results of pathology , image fusion produced its diagnostic value: Sensitivity was 100%, Speciality 82%,NPV(negative predictive value) 100%, PPV(positive predictive value) 94%.Conclusion: Multi-modal medical image fusion provides useful information ,which help clinician to obtain a full understanding about thequality and anatomic details of the lesions. It overcomes internal defect of noninvasive techniques and makes " Virtual biopsy " come ture.