Image processing in the digital tomosynthesis for pulmonary imaging

Digital tomosynthesis makes it possible to reconstruct multiple tomographs from digital data obtained during a single tomographic motion and permits digital processing, which adds a number of special advantages to the well-known advantages of conventional tomography. We performed digital tomosynthesis with a fluororadiographic TV unit with tomographic function which was capable of producing pulsed low-and high-energy X-rays alternately, and we studied digital image processing to improve the image clarity of the reconstructed tomographs. To identify the optimal parameters for processing image data by means of spatial frequency filtration we evaluated the spatial frequency distribution of image data in linear tomographs of the lung, and on the basis of the results of this study we developed several types of digital image processing to reduce tomographic blur and system noise, to improve visualisation of faint opacities, to reduce resistant tomographic blur as well as overall blur, and to generate low-noise bone images based on dual-energy subtraction tomosynthesis. Correspondence to: S. Sone     

Digital image processing to remove blur from linear tomography of the lung

We have studied digital image processing methods to reduce blur in linear tomography of the lung performed with the Fuji computed radiography (FCR) system. One-dimensional unsharp (blur) mask filtering in the direction of tomographic movement was found to be effective in reducing blur. The appearance of the processed tomograms varied depending largely on the properties of the unsharp mask filters applied. With regard to the spatial frequency response of the filter, a high mid-frequency response and a low low-frequency response of the tomographic image data was most effective in obtaining high image quality tomograms. When the standard 2-dimensional unsharp mask technique of the FCR system was additionally applied to the one-dimensional unsharp mask processed tomogram, the clarity of the tomogram was further enhanced. These observations may also be helpful when considering image processing to obtain diagnostically informative digital radiography.     

Digital TV tomosynthesis--system development, digital image processing and clinical applications

We have designed and developed two types of high-resolution digital X-ray TV tomosynthesis systems which incorporate an X-ray TV fluororadiographic unit with tomographic function and a digital image processor, which handles image data from the image intensifier-TV camera chain. We have studied methods of improving image quality of the tomosynthesis images and their clinical applicability. As a result of clinical application of this method, a number of advantages have already become apparent, as listed below: 1) Low patient radiation dose per examination, short examination time, and early display of the image on the CRT. 2) Reconstruction of the tomograms at any desired depth and oblique tomograms. 3) Spatial frequency filtering to improve the image quality, particularly one-dimensional spatial filtering process to reduce linear tomographic blur. 4) Digital angiotomosynthesis with and without subtraction. 5) Dual energy subtraction digital tomosynthesis. 6) Electronic storage and transmission of image data. Clinical effectiveness of this method was shown in such areas as imaging of the ear, spine and lung, angiography of the brain and abdominal organs, and arthrography.     

Digital tomography: description and preliminary clinical experience

Digital tomography was performed in 30 patients using a new prototype system with a PET-Scope image tube. Twenty-eight exposures are obtained during a single revolution of the x-ray source over the volume of tissue to be imaged. The sum of the radiation produced by the 28 exposures is about equivalent to a single routine exposure of the part imaged. The data accumulated can provide up to 200 tomographic sections at intervals of 5 mm. Reconstruction time for each section is 3 sec. The raw images may be reconstructed as often as desired. Algorithms are being developed to allow axial and sagittal reformations. Further work is required to improve spatial resolution and contrast and to develop algorithms to allow subtraction of blurred images above and below the reconstructed site. The authors believe that digital tomography may eventually replace most routine tomograms.     

Interrelationship between cross-sectional and sagittal imaging in computer-assisted dental implant tomography

OBJECTIVE: To determine what changes may occur in computer-assisted dental tomography to cross-sectional and sagittal slices when the positions of one or both slices are adjusted. METHODS: A human skull with a ball bearing attached to the mandibular first molar was positioned off-centre in a CommCAT (Imaging Sciences International, Hatfield, PA, USA) tomographic machine. An occlusal projection of the mandibular dental arch was scanned into the computer. Correction measurements obtained from scout cross-sectional and sagittal slices were used to 'customize' cross-sectional and sagittal slices simultaneously and also individually. RESULTS: The resulting changes in position and blur of the ball bearing were analysed. When simultaneously and correctly customized, the resultant slices were at the correct locations with a sharp image of the marker. When the sagittal slice only was adjusted, the blur of the marker changed, but its position remained unchanged; at the same time, the blur of the marker in the cross-sectional slice remained unchanged, but its position changed. When the cross-sectional slice only was adjusted, the blur changed, but its position remained unchanged; at the same time, the blur of the marker in the sagittal slice remained unchanged, but its position changed. CONCLUSIONS: It is important for dentists evaluating tomograms for implant placement to be aware that with the CommCAT tomographic machine an interrelationship exists between cross-sectional and sagittal slices and that adjustment in one type of slice produces changes in either the blur or position of the marker in the adjusted and unadjusted slices.     

Linear vs. pluridirectional tomography of the chest: correlative radiographic anatomic study

The basic imaging capabilities of 15 degrees linear, 30 degrees linear, hypocycloidal, and trispiral tomographic movements were compared using a fresh frozen human cadaver. Large numbers of tomograms, 100 with each tomographic movement, were compared as a motion study developed by copying the four tomographic series on movie film. In addition, the integrity of the tomograms was compared with radiographs of representative 2 mm coronal sections sawed from the fresh frozen thorax specimen. The study demonstrated the theoretical conclusion that a linear tomogram is not a sectional image and does not truthfully represent the planar anatomy of the intended layer. The pluridirectional tomograms, on the other hand, do accurately depict the anatomic section. The more general use of pluridirectional tomographic techniques in the chest is encouraged and should be used as the basis for comparing conventional tomographic imaging capabilities to other new modalities, particularly computed tomography.     

Tomoscopy: dynamic layer imaging without mechanical movements

A tomographic system in which electronic movement of the x-ray source and the image receptor replace the mechanical movement of conventional tomography has been constructed and tested. Because the system permits very short, rapidly repeated exposures, viewing by television provides real-time fluoroscopic tomography, termed tomoscopy. Further, electronic adjustment of tomographic layer height permits rapid direct search for the optimum layer. The system requires no more than conventional fluoroscopic radiation levels and resolves 3 line pairs per millimeter (LP/mm) in the televised image and 4 LP/mm in the spot film images. It offers rapid, highly customized, high-resolution tomography under direct fluoroscopic control. Precise percutaneous biopsy with integrated needle-controlled equipment and tomography of contrast-filled blood vessels incorporating digital subtraction principles are theoretic possibilities.     

双能量减影骨组织图像在肋骨骨折诊断上的作用分析

目的:探讨双能量减影骨组织图像在肋骨骨折检查中的应用和诊断价值.方法:搜集171例临床怀疑肋骨骨折的病例,男97例,女74例,年龄7～89岁,平均41岁.运用美国GE公司生产的Revolution XR/d型数字X线成像设备(DR),采用双能量减影法摄正、斜位胸片.结果:在本组病例中,皱折状骨折在骨组织减影图像上的检出率达73.3%,高于常规DR图像的检出率60.0%,而裂隙状骨折及不全性骨折在骨组织减影图像上的检出率低于常规DR图像,完全性骨折在骨组织减影图像和常规DR图像上的检出率相同,骨组织减影图像与常规DR图像两者间总的检出率无显著性差异(P＞0.05);骨组织减影图像正位的检出率为24.1%,斜位的检出率为44.8%,常规DR图像正位的检出率为29.3%,斜位的检出率为48.3%.结论:双能量减影骨组织图像对不同形态骨折的显示程度不同,可提高细小肋骨骨折的检出.摄片体位影响骨折的检出,无论是骨组织减影图像还是常规DR图像斜位较正位更容易显示骨折.     

Retrocardiac phantom image on hypocycloidal chest tomograms

A phantom image produced by displaced blur margins of the cardiac silhouette and mimicking a paraspinous retrocardiac mass was noted in 50% of 125 hypocycloidal full chest tomograms. It was reproduced on tomograms of a chest phantom. Careful comparison with routine posteroanterior and lateral chest radiographs prevents misinterpretation.     

双能量减影骨组织像在肋骨骨折检查中的应用

目的探讨双DR能量减影骨组织像在肋骨骨折检查中的应用价值。方法搜集174例临床有外伤史并疑有肋骨骨折患者,其中男103例,女71例,年龄3～78岁。采用美国GE公司生产的双平板探测器DefiniumTM8000X线射线系统,摄取患者双能量减影胸部正位片,减影得到DR胸部标准像、软组织像、骨组织像三种图像。对标准像和骨组织像进行诊断分析和评价。结果 174例病例中,胸部标准像明确诊断肋骨骨折67例,占38.6%;骨组织像明确诊断肋骨骨折105例,占60.4%。结论双DR能量减影骨组织像较标准像更能清晰显示肋骨骨折情况,明显提高肋骨骨折的检出率。但需注意呼吸运动对图像质量的影响,尽量减少伪影的产生。     

Digital tomosynthesis: technique for electronic reconstructive tomography

A technique is described for obtaining tomographic images through retrospective reconstruction of digital data. The apparatus used for this technique, called digital tomosynthesis consists of a linear tomographic x-ray machine that has been modified by the addition of a fluoroscope and TV system, a video disk recorder, an analog-to-digital converter, and a small computer for data processing and manipulation. Video frames are collected and stored during a single tomographic sweep. The stored data are then digitized and retrospectively processed in the computer for reconstruction of any desired tomographic plane within the body. The major advantages of DTS include short patient study time, low radiation dose compared with conventional tomography, the ability to enhance the digitized image through manipulation of window and level display, and the applicability of this technique to dynamic studies such as angiotomography. Phantom studies show good diagnostic quality of the resulting images, and preliminary vascular studies in dogs indicate the clinical potential of this technique for use in digital subtraction angiotomography.     

Pulmonary nodules: detection of calcification by linear and pluridirectional movement in tomographic study

Calcification within solitary pulmonary nodules (1-2 cm range) was demonstrated by conventional tomography in a group of 20 patients. Each patient had a combined tomographic examination consisting of 15 degrees linear movement followed by 34 degrees hypocycloidal movement. Linear tomography was superior to pluridirectional tomography in 19 cases for visualization of calcium within the nodule. The results are attributed to differences in contrast sensitivity and "blur" phenomena between the two tomographic movements. This study, although comparing a specific linear movement with a specific pluridirectional movement, provides insight into the differences between tomographic techniques for a given task.     

Single photon emission computed tomograms of the liver: normal vascular intrahepatic structures

Because of the high target-to-background contrast obtained with single photon emission computed tomography (SPECT), normal intrahepatic vessels approximately 2 cm in diameter may appear as distinct focal defects in tomographic sections throughout the liver even though normal vessels rarely cause such defects on planar images. To assess this problem, five subjects without evidence of liver disease underwent tomography of the liver with Tc-99m sulfur colloid (TSC) and on a separate occasion tomography of the intrahepatic blood pool with Tc-99m autologous red blood cells (RBC). In each case, well demarcated defects were obvious in contiguous TSC liver tomograms in various planes. Direct comparison with RBC tomograms showed that all of these defects corresponded to intrahepatic veins, typically the right portal vein, its posterior branch, and the left portal vein. Knowledge of the intrahepatic vascular anatomy in a variety of tomographic planes, with examination of each defect in multiple orthogonal planes is necessary to avoid false positive interpretations. In some instances a study with RBC may also be required for more conclusive evaluation of defects seen on TSC liver tomograms.     

Conventional tomography with a digital vascular imaging system

Digital conventional tomography may be performed using a digital vascular imaging (DVI) system. Digital synthesis and removal of blur from multiple tomographic sections were accomplished by computer with a single motion of the source-receptor (x-ray tube-image intensifier) complex. Tomographic application of a DVI system was evaluated by phantom simulations.     

双能量数字减影胸片对肺内小结节检出的意义

目的探讨双能量数字减影胸片对肺内小结节检出的意义。方法27例病理证实恶性肿瘤伴肺内转移的患者,分别行增强CT扫描、双能量数字减影胸片与常规DR胸片。使用柯达质量控制检测仪比较双能量数字减影DR与常规DR胸片的图像质量。再以CT扫描结果为金标准,由2位高年资放射科医师采用双盲法对双能量数字减影胸片与常规DR胸片进行分析,比较两者对肺内转移瘤的检出有无显著差异。结果双能量数字减影DR与常规DR图像在噪声上(均匀度)无差异,但清晰度稍差。双能量数字化减影的胸片对肺内小结节的检出率为91.2%;而常规DR胸片对肺转移瘤的检出率为85.0%,两者之间有显著差异(P<0.05)。结论双能量数字减影技术可减少肺野内骨骼及其它钙化影响,对肺内结节的检出能力高于常规DR胸片。     

One-shot dual-energy subtraction imaging

Dual-energy subtraction imaging by a single x-ray exposure (one shot) can easily be performed by using computed radiography with scanning laser-stimulated luminescence. In a phantom study, a thin copper filter placed between two imaging plates produced a dual-energy subtracted image from a single x-ray exposure. One-shot dual-energy subtraction imaging was also useful in the diagnosis of thoracic lesions.     

下肢双能CTA去骨去钙化功能临床应用价值的研究

目的研究双能CT下肢动脉成像去骨去钙化功能的临床应用价值。资料与方法 43例患者进行双能CT下肢动脉检查,得到双能去骨后的图像(包括去除钙化斑块后的图像)及传统自动去骨后的图像,残留骨手动去除,记录各自去骨时间及对各节段血管的侵蚀度。判定斑块去除前后对血管狭窄度测量的影响。结果残留骨仅在未去除钙化斑块图像上观察到,双能去骨时间(1.86±0.28)min低于传统自动去骨时间(2.00±0.24)min(P<0.01,t=-4.382);共对822节段血管进行分析,血管侵蚀率前者低于后者,且最常发生于小腿部血管(χ2=13.13,P<0.01)。结论双能CTA明显优于传统多层螺旋CTA,双能自动去骨软件提高了工作效率,有助于小腿部血管的显示,自动去斑块功能可提高诊断准确性。     

The influence of tube voltage,X-ray exposure rate and thickness of filter material on image quality with dual-energy subtraction imaging in chest diagnostics

Digital luminescent radiography enables dual-energy subtraction imaging, because this computed system allows susbtraction of imaging data and image post processing, as for example special windowing or edge enhancement. In a special cassette a copper filter is placed between two imaging plates for energy separation by a single X-ray exposure. Image post-processing with subtraction of imaging data permits the elimination of either skeleton or soft tissue structures. The influence of filter thickness, tube voltage and the X-ray exposure dosage on image quality is examined by the use of an anthropomorphic quality is examined by the use of an anthropomorphic phantom of the chest. According to our initial results dual energy subtraction imaging in one-shot-technique seems to be useful in the diagnostics of skeletal lesions and especially pulmonary nodules.     

Flash双源CT脑血管成像双能量减影去骨比率值的优化研究

目的 探讨双能量减影去骨比率值的调节在脑血管成像中的临床应用价值.方法 回顾分析45例双源CT二期双能量脑血管检查患者资料.双能量法按不同去骨比率(Ratio)值分别分成A组(Ratio=1.6),B组(Ratio=1.7),C组(Ratio=1.8),D组(Ratio=1.9,默认值);100 kV Neuro-DSA减影法定为E组.采用双盲法按去骨程度及血管完整程度对图像质量评分(1~5分,1分为差,5分为优,3~5分为满意).通过自身对比分析不同双能量去骨Ratio值对减影图像质量的影响,并以Neuro-DSA减影为对照,比较各组图像信噪比(SNR)、对比噪声比(CNR)及图像质量评分的统计学差异.结果 双能量减影不同去骨Ratio值组的颅底动脉图像评分总体满意率分别为:A组73.33%,B组91.11%,C组78.52%,D组70.37%.Ratio值取1.7时,虹吸段及岩骨段动脉的评分达满意显示率最高,差异有统计学意义(P<0.05).B组、D组、E组3组间SNR与CNR的差异不显著(P>0.05).结论 双源CT单次增强双能量减影去骨Ratio值的调节优化后的图像质量与100 kV Neuro-DSA减影一致性较好,可弥补双能量减影颅底脑血管缺损的不足,推荐在临床急诊脑血管检查中应用.     

Coils of the cochlea.

Abnormalities of the cochlear portion of the bony labyrinth may often be suspected but are inadequately demonstrated by conventional tomography in the antero-posterior position. Further views are necessary. The authors demonstrate this by means of a comparison between tomographic and histological sections of the inner ear of an anencephalic with a unilateral Mundini defect. Submento-vertical tomograms give a good demonstration of the cochlea.