[Abstract] Objective To explore the diagnostic value of multilayer spiral CT image post-processing technique for bronchial tuberculosis and to analyze it against the diagnostic results of bronchoscopy. Methods Using a multilayer spiral CT scanner, 50 patients with bacteriological examination and confirmed bronchial tuberculosis by bronchoscopy were selected to perform CT scan of the lungs, and volume display (VR), maximum density projection (MIP), multiplanar reconstruction (MPR) and virtual bronchoscopy (VE) were performed by the workstation to observe the morphology and alignment of each bronchial segment. The results were 98% consistent with the imaging presentation of bronchial tuberculosis. Conclusion The detection rate of bronchial tuberculosis by multi-layer spiral CT image post-processing is close to that of bronchoscopy, which can provide a more adequate imaging basis for detecting relevant lesions and formulating clinical treatment plans. Hou Dai-Lun, Department of Medical Imaging, Shandong Chest Hospital
[Keywords] Bronchial tuberculosis, multilayer spiral CT, three-dimensional reconstruction
The value of post-processing images by multi-slice spiral CT reconstruction of the bronchial tuberculosis
［Abstract］Objective：To explore Multi-slice spiral CT Isotropic Scanning features of the bronchial tuberculosis, study the Clinical value. Method：Multi-slice spira CT images were performed in 50 adult bronchial tuberculosis patients confirmed by the bacteriological examination and bronchoscopy.They were performed lung spiral CT scanning. All original images were transferred to workstation for image processing.The best depictive MPR, MIP, VR , VE images were acquired for study. To observe all the image characteristics of the bronchial branches. Conclusion: Images of Multi-slice spiral CT could accurately demonstrate the bronchial tuberculosis, and ——————— Key words】：bronchial tuberculosis；multi-slice spiral CT；three-dimensional reconstruction In 1698 Morton [1] first proposed tuberculous bronchitis, lesions mainly occur in the trachea and bronchial mucosa or submucosa, also known as endobronchial tuberculosis (EBTB), in fact lesions can involve the muscular layer and bronchial cartilage, and can invade the trachea, now called tracheo-bronchial tuberculosis. The incidence of pulmonary tuberculosis has been on the rise in recent years, and about 10%-20% of patients with active tuberculosis can involve the trachea and bronchi [2], but because of the lack of specificity of its clinical manifestations, the positive rate of sputum smear test is low, and Lee et al. found that only 17% of 121 confirmed cases of bronchial tuberculosis showed positive sputum smear for antacid bacilli, leading to a high rate of missed diagnosis [2]. MSCT has high spatial resolution and temporal resolution, and has advantages for the display of bronchial lesions. Most of the bronchi travel longitudinally, and the original images are transverse, and most of them can only observe the short axis of the bronchi, which is not conducive to the comprehensive observation of the bronchial lumen and wall, and post-processing techniques can make up for the deficiency in this regard. The image post-processing technique allows observation of lung lesions from any plane, to observe the bronchial morphology in parallel planes in any bronchial long axis, to observe the bronchial lining, to see the thickening of the wall, narrowing of the lumen, increased density, unsmooth surface, obstruction and narrowing, distortion, deformation, etc., and also to observe the number and extent of lesions, the enlargement of lymph nodes and the concomitant situation in the lung from an overall perspective, which is more conducive to The diagnosis of bronchial tuberculosis can be made and detailed imaging information can be provided for treatment. Our study is an attempt to improve the sensitivity and specificity of MSCT for the diagnosis of bronchial tuberculosis through image post-processing. 1 Data and Methods 1.1 Object of study A total of 100 patients, 60 males and 40 females, aged 20-80 years, who were admitted to our hospital and underwent plain and enhanced chest MSCT scans, and were diagnosed with bronchial tuberculosis by bacteriological examination and bronchoscopy, were selected. 1.2 Imaging equipment and methods A GE LightSpeed16 CT scanner with spiral scan, matrix 512X512, tube voltage 120kv, tube current 120mAs, scan layer thickness 5.0mm, pitch 1.375 was used. Image post-processing methods: the original images were transferred to the workstation with a reconstruction layer thickness of 1.25 mm and a reconstruction interval of 1.25 mm, and the reconstructed image data were loaded into the 3D interface, and MPR was performed on the bronchi suspected of abnormalities in the transverse images, (1), coronal and sagittal MPR for observation of hilar lymph nodes and intrapulmonary TB foci; (2), by adjusting the angle of the MPR baseline to show The long axis of the “target” bronchus was used to observe the changes of lobe and segmental bronchial walls and lumen. The target bronchus was then subjected to virtual bronchoscopy to observe the bronchial lining and lumen status. In this study, only the main bronchus and each lobe and segmental bronchus were observed. The cross-sectional images were used as the control group, and the cross-sectional images combined with the post-processed images were used as the experimental group. Any case with bronchial lumen narrowing, wall thickening, pulmonary atelectasis, hilar lymph node enlargement, or intrapulmonary tuberculosis foci was considered as a positive case, and the percentage was calculated, and the positive rate was analyzed by SPSS statistical software. Then the target bronchus was processed by virtual bronchoscopy, analyzed, and the microscopic performance characteristics of bronchial endoscopy were summarized; at the same time, the enlarged lymph nodes were manually positioned on the transaxial images, and simulated images of the virtual bronchoscopic bronchial lumen were created to accurately locate the enlarged lymph nodes. 2 Results 2.1 Imaging performance of conventional CT transverse images and post-processed images of the chest (see Table 1) Image presentation irregular bronchial stenosis (see Figure 1) Irregular thickening of the duct wall (see Figure 2) segments and lobar pulmonary atelectasis (see Figure 3) Pulmonary hilar lymph node enlargement intrapulmonary tuberculosis foci No lesion was found control group 69 (69%) 56 (56%) of the 53 (53%) 51 (51%) 88 (88%) 18 (18%) experimental group 98 (98%) 90 (90%) 55 (55%) 51 (51%) 90 (90%) 2 (2%) The data obtained were compared by SPSS13.0 statistical software, and the positive rate of the two groups was significantly higher than that of the control group, P>0.05. 2.2 Comparison of virtual bronchoscopy performance with fiberoptic endoscopy findings (see Table 2)
Microscopic manifestations Ductal stenosis (Figures 1, 3)
Thickening of the duct wall (Figure 2) Pathological changes in the duct wall (ulceration, congestion, etc.) (Figure 3) Tracheal cartilage ring
Fiberoptic bronchial endoscopy
77 (77%)
0 75 (75%) 19 (19%) Virtual bronchoscopy 73 (73%) 36 (36%) 62 (62%) 2 (2%) The diagnostic results of virtual bronchoscopy were not significantly different from fiberoptic endoscopy for manifestations such as luminal narrowing and ulceration of the tube wall; uniform thickening of the bronchial wall was difficult to observe with fiberoptic bronchoscopy but could be observed with virtual bronchoscopy; fiberoptic endoscopy was significantly better than virtual bronchoscopy for the diagnosis of lesions in the cartilaginous rings of the trachea. 2.3 Bronchoscopy (see Table 3) Bronchoscopic manifestations Bronchial mucosa congestion and edema (see Figure 1d)
Bronchial mucosal ulceration, necrosis (see Figure 2d) Granulomatous proliferation (see Figure 4d) scar stenosis (see Figure 3d) Missing or broken cartilage rings in the wall Number of cases (%) 71 (71%) 65 (65%) 57 (57%) 49 (49%) 19 (19%) Bronchoscopic manifestations were mainly bronchial mucosa congestion, edema, caseous necrosis of the mucosa and/or submucosa, small ulcers or nodules, unequal narrowing of the lumen, tuberculous granulation tissue hyperplasia, thickening of the wall, scar narrowing, occlusion of the official lumen, and a few visible defects or fractures of the cartilaginous rings of the trachea and bronchial wall, and pathological examination of the tissue taken during microscopy to confirm the diagnosis of tracheal and bronchial tuberculosis. 3 Discussion 3.1 Regarding the diagnosis of bronchial tuberculosis. 3.1.1 Laboratory tests Bronchial tuberculosis (EBTB) refers to tuberculosis occurring in the trachea, bronchial mucosa, submucosa and outer lining (cartilage and fibrous tissue), and is an extrapulmonary tuberculosis. It was found that about 10-40% of patients with active pulmonary tuberculosis are complicated by EBTB [3], and about 60-70% of patients with pulmonary tuberculosis combined with EBTB are sputum positive, and about 25-30% are sputum negative, with the remainder being Patients with simple EBTB, i.e., EBTB with no obvious active TB lesions found in the lungs, account for about 5%-10% of all EBTB. Therefore, sputum positivity of TB in patients is not important for the diagnosis of bronchial tuberculosis. Those who have no obvious TB lesions in the lungs and have multiple positive sputum are helpful for the diagnosis of bronchial tuberculosis if they have clinical symptoms and imaging manifestations of bronchial tuberculosis. 3.1.2 Diagnostic role and defects of fiberoptic bronchial endoscopy for bronchial tuberculosis Fiberoptic bronchoscopy can visualize the narrowing and obstruction of the diseased bronchial opening, and can clearly detect mucosal congestion, edema, caseous necrosis, ulceration and granulation tissue formation, and can perform lavage and brush examination, take tissue biopsy and sputum for culture at the same time [5, 6], but at the same time there are certain defects, because the size of bronchoscopy is limited, for the distal lobes and segments of the bronchial opening and lumen situation observation It is difficult to observe the length of narrowed or obstructed trachea and the post-stenotic bronchial condition, and it is impossible to observe the bronchial condition as a whole and to detect the disseminated lesions in the lung. The recently developed ultra-fine bronchoscope with smaller outer diameter can observe the peripheral bronchi of the lung, but its aspiration ability is correspondingly reduced, and the small amount of blood and secretions around the airway also significantly reduces its visualization, which is complicated by the easy collapse of the peripheral airway during aspiration. 3.1.3 Imaging Conventional chest X-ray lacks specific signs for the diagnosis of bronchial tuberculosis, especially for patients with simple bronchial lesions, and patients with recurrent cough, sputum or with hemoptysis while chest X-ray shows normal should be recommended to perform CT or bronchoscopy for further examination to exclude the disease. From Table 1, it can be found that the detection rate of EBTB is only 82% for conventional CT transverse position scan, and there is a certain possibility of missing the diagnosis. 3.2 Advantages of MSCT image post-processing technique for bronchial tuberculosis diagnosis
In recent years, with the improvement and popularization of multi-row spiral CT and three-dimensional image reconstruction technology, CT volume scans with millimeter-level layer thickness of the whole lung are completed in a short time, and various image post-processing is performed. The information after CT conventional axial scans can be observed in different sections, angles and modalities of the diseased bronchus and its surrounding tissues through MPR, CPR, MIP, VR, VE and other image post-processing techniques The virtual bronchoscopic images generated by CT 3D reconstruction technique can be used as a supplement to bronchoscopy and can compensate for the difficulty of overall observation of the trachea by bronchoscopy, and Finkelstein et al. concluded that the sensitivity of virtual bronchoscopy for the diagnosis of airway obstructive lesions is 100% [7]. And CT images are more sensitive than fiberoptic bronchoscopy for showing calcification in the bronchial lumen and peribronchial calcification. The virtual bronchoscopy technology allows the observation of the trachea, bronchial tube wall and the lumen, and its fine level can meet the diagnostic requirements of EBTB. Some studies have concluded [8,9] that CT virtual bronchoscopic images are less effective for displaying subsegmental bronchi, Maniatis [10] et al. suggested -520Hu as the threshold for observing the central airway and -720Hu as the threshold for accurately displaying the bronchi and subsegmental bronchi in the lung segment, applying a 16-18 cm field of observation (FOV) for retrospective target reconstruction, spatial resolution The raw data are isotropic, which helps to improve the quality of virtual bronchoscopy images; comparing the images obtained with MSCT reconstruction layer thickness of 1.5 mm and reconstruction interval of 0.75 mm with those obtained with reconstruction layer thickness of 0.75 mm and reconstruction interval of 0.4 mm, the bronchi that can be distinguished have increased from 7.5 mm to 4.6 mm (P<0.0001) [11]. Virtual bronchoscopic images were also found to be useful in improving the diagnostic rate of transbronchoscopic fine needle aspiration and biopsy (TBNA) [12]. 3.3. The value of MPR technique for tracheal and bronchial tuberculosis. a. MPR is superior to cross-sectional images in terms of lumen and wall status sensitivity: MPR post-processed images revealed 98% bronchial lumen narrowing, significantly higher than the 69% observed in axial images; MPR post-processed images revealed 90% bronchial wall thickening, significantly higher than the 56% observed in axial images. specificity: comparing lumen and wall separately, MPR post-processed images were significantly more positive than conventional cross-sectional images for the diagnosis of luminal stenosis and irregular thickening of the wall. b. no significant difference in the observation of hilar lymph nodes. c. No significant difference in the observation of intrapulmonary lesions. 4. Virtual bronchoscopy versus fiberoptic bronchoscopy a. In the evaluation of luminal stenosis In 100 patients who underwent fiberoptic bronchoscopy, 77 cases were found to have luminal stenosis with varying degrees of stenosis, with a positive rate of 77%, while 73 cases were found to have luminal stenosis by virtual bronchoscopy, with a positive rate of 73%. There was no significant difference between the two positive rates when comparing The multi-layer spiral CT simulation bronchoscopy not only showed a high rate of tracheal and bronchial stenosis, but also had a better display of high degree of stenosis and post-obstruction bronchus that could not be passed by fiberoptic bronchoscopy. b. In evaluating the thickening of the tube wall In terms of evaluating bronchial wall thickening, fiberoptic bronchoscopy is difficult to directly observe bronchial wall thickness, and virtual bronchoscopy can be observed by VR post-processing technology, which can both directly display the longest diameter of the lumen through multiplanar reconstruction MPR technology and observe the lumen with CTVE, and also truncate the bronchus along the long axis in CTVE mode to compare and observe the thickening of the bronchial wall (Figure 5). c. In evaluating the pathological changes of the bronchial wall: (ulceration, congestion, etc.) In 100 patients who underwent fiberoptic bronchoscopy, 75 cases were found to have pathological changes such as ulceration, congestion, and granulation proliferation in the tube wall, with a positive rate of 75%, while a total of 62 cases were found to have a less-than-glossy luminal wall with visible necrosis and hyperplasia when they underwent virtual bronchoscopy, with a positive rate of 62%. The difference between the two positive rates was not significant. d. In the evaluation of the air cartilage rings In 100 patients who underwent fiberoptic bronchoscopy, 19 cases were found to have missing or broken cartilaginous rings in the trachea and bronchial wall, with a positive rate of 19%, while only 2 cases underwent virtual bronchoscopy and found cartilaginous ring changes, with a positive rate of 2%. Comparing the two positive rates, virtual bronchoscopy was significantly lower than fiberoptic endobronchoscopy. 5. To study the clinical significance of bronchial tuberculosis and its imaging manifestations Clinical studies have found that there are significant differences in the main contradictions, regression and prognosis in the treatment of active EBTB and inactive EBTB with combined bronchial stenosis, which should be treated differently during treatment. For patients with diagnosed EBTB, 3D CT images of the bronchus of the lesioned segment can accurately calculate the extent of lesion involvement and accurately grasp the degree of airway stenosis, or the presence of restenosis or bronchial dilatation distal to the stenosis. Multi-layer spiral CT post-processing technology can significantly improve the diagnosis rate of bronchial tuberculosis with easy operation and painless examination of patients. Combining the post-processed images with the bronchoscopic staging of EBTB, CT staging is developed, thus providing an important reference for bronchoscopy and the development of endoluminal interventional treatment plans, and clinical treatment plans are developed according to the intrapulmonary and bronchial and tracheal manifestations of different staging. 4 Conclusion In summary, based on multi-layer spiral CT image post-processing technology, which is displayed by multi-planar reconstruction and multiple imaging modalities, combined with bronchoscopy, the strengths and weaknesses are complemented to maximize the identification of lesion characteristics, make the diagnosis more accurate, and provide favorable assistance for clinical treatment. Figure 1 Left main bronchial lumen stenosis a. Axial cross-sectional images; b. MPR coronal images; c. MPR sagittal images; d. Fiber endobronchoscopic images; e. Virtual bronchoscopic images; f. VR transparency images showing bronchial stenosis Figure 2 Irregular thickening of the bronchial wall and luminal narrowing in the anterior basal segment within the lower lobe of the left lung a. axial transverse images; b. MPR oblique coronal images; c. MPR sagittal images; d. fiberoptic endoscopic images; e. virtual bronchoscopic images; f. VR images Figure 3 Bronchial stenosis in the dorsal segment of the lower lobe of the left lung with subsegmental bronchial occlusion and partial atelectasis of the lung lobe a. Axial transverse image; b. MPR coronal image; c. MPR sagittal image; d. Fiberoptic bronchoscopic image; e. Virtual bronchoscopic image; f. VR image Figure 4 Nodule-like protrusion in the lumen of the bronchus of the left lower lobe of the lung a. Axial transverse image; b. MPR oblique coronal image; c. MPR oblique sagittal image; d. Fiberoptic endoscopic image; e. Virtual bronchoscopic image; f. VR image Figure 5: Figure 5a. MPR image in long-axis position showing thickened bronchi connected to the cavity; Figure 5b. VR image showing smooth inner wall of the cavity and uneven thickness of the bronchi connected to it; Figure 5c. CTVE showing irregular inner wall of the bronchi with nodule-like protrusions visible; Figure 5d. Cross-sectional dissection on the basis of CTVE showing thickened bronchi walls with restricted thickening visible. Reference. [1]Jiang HN,Qu JM,He LX. Advances in the diagnosis and treatment of tracheobronchial tuberculosis [J]. Chinese Journal of Anti-Tuberculosis, 2000,(01). [2]Kim Y, Lee K S, Yoon J H, et al. Tuberculosis of the Tachea and Min Binchi: CT Fndings in 17 Patients[ J].AJR,1997,168:1051. [3] Huang MG, Qi M, Lei XY, et al. A controlled fiberoptic bronchoscopic and imaging study of endobronchial tuberculosis [J]. Journal of Practical Radiology, 2005,21(7):698-701. [4] Yu H, Li H, Shen C, et al. achea and Min Binchi: A study of multilayer CT diagnosis of tracheobronchial tuberculosis. Chinese Journal of Medical Computer Imaging, 2004,10(2):92. [5]Chinese Society of Medical Sciences, Branch of Tuberculosis. Diagnostic and therapeutic microscopy guide for pulmonary tuberculosis [J]. Chinese Journal of Tuberculosis and Respiration, 2001,24(2):70-74. [6]Liu L,Wang HX,Zhang FQ. Efficacy of trans-fiber bronchoscopy in the treatment of various types of endobronchial tuberculosis. Chinese Journal of Tuberculosis and Respiratory Diseases,2002,25:62. [7] Finkelstein SE, Summers RM, Nguyen DM, Stewart JH, Tretler JA, Schrump DS. Virtual bronchoscopy for evaluation of malignant tumors of the thorax. J Thorac Cardiovasc Surg. 2002;123(5):967C972. [8] Summers RM, Feng DH, Holland SM, et al. Virtual bronchoscopy: segmentation method for real-time display. Radiology 1996;200:857-862. [9]Lacasse Y, Martel S, Hebert A, et al. Accuracy of virtual bronchoscopy to detect endobronchial lesions. Ann Thorac Surg 2004;77:1774-1780. [10]Maniatis PN, Triantopoulou CC, Tsalafoutas IA, et al. Threshold selection in virtual bronchoscopy: phantom study and clinical implications. Acta Radiol. 2004 Apr;45(2):176-83. [11]Khan MF, et al. Virtual endoscopy of the tracheo-bronchial system: submillimeter collimation with the 16-row multidetector scanner. Eur Radiol 2004;14:1400-1405. [12]William E. Higgins1, James P. Helferty1,Kongkuo Lu,et al. 3D CT-Video Fusion for Image-Guided Bronchoscopy. Comput Med Imaging Graph. 2008 April ; 32(3): 159C173.