Volume 36, Issue 6, June 2009
Index of content:
- Industrial Physics Forum: Imaging Symposium: Room 210A
Advances in Photon Counting Detectors in X‐ray Imaging
36(2009); http://dx.doi.org/10.1118/1.3182239View Description Hide Description
Mammography is currently one of the most common x‐ray imaging examinations. More than 100 million women worldwide are screened every year and early detection of breast cancer through mammography has proven to be a key to significantly reduced mortality. The requirement on spatial resolution as well as contrast resolution is very high in order to detect and diagnose the cancer. Moreover, because of the large number of women going through this procedure and the fact that more than 99 % are healthy, it also becomes very important to minimize the radiation dose.
Photon counting may be one way to meet the demands and mammography is the first modality in x‐ray imaging to implement photon countingdetectors. FDA approval is still pending but they are currently in routine clinical use in more than 15 countries. The photon counting enables a discrimination of all electronic noise and a more optimum use of the information in each x‐ray. The absence of electronic noise is particularly important in low dose applications, in for example tomosynthesis a number of exposures from different angles are required and since the dose in each projection is just a fraction of the total dose for a mammogram the sensitivity to electronic noise will increase.
Using the spectral information for each x‐ray it is in principle possible to deduce the elemental composition of an object in the breast. This could for example be used to enhance microcalcifications relative to soft tissue and differentiate water from fat in cysts. Recently contrastmammography has attracted significant attention. In this application Iodine is used as a contrast media to visualize the vascular structure. As in breast MRI the cancer stand out because of the leaky vessels resulting from its angiogenesis. A photon countingdetector gives a unique opportunity to image the Iodine through spectral imaging by adjusting one of the thresholds to its K‐edge.
Challenges for photon counting in mammography are high rates of x‐rays, both to generate the required flux at the source and to handle the rates at the detector without pile‐up. Even more difficult to handle are the charge sharing between detector pixels which, if not corrected for, will compromise the energy information.
The current status of photon countingdetectors in mammography will be described together with strategies to overcome the pit‐falls. Also future possibilities with spectral imaging in mammography will be investigated and examples from ongoing clinical trials will be given.
1. Status of photon countingdetectors in mammography
2. Pit‐falls and opportunities with photon countingdetectors for mammography
3. Future applications based on spectral detectors for mammography
36(2009); http://dx.doi.org/10.1118/1.3182241View Description Hide Description
Photon‐counting detectors are now being introduced in medical imagingsystems. A major challenge for PC detectors is to enable high count rate operation required for medical imaging. High count rate operation is now available with state‐of‐the art pulse counting electronics. Pulse counting techniques can work together with direct‐conversion sensors enabling additional cost and performance benefits versus indirect‐conversion sensors.
PC detectors are easily configurable for multi‐energy acquisitions, enabling dual and triple (k‐edge) imaging techniques. The energy discrimination and binning by PC detectors competes well with the energy separation achieved with dual kVp imagingsystems and provides unique performance capabilities for triple/k‐edge imaging.
PC detectors have been shown to have negligible levels on electronic noise. The low noise performance enables new scanning techniques and ultimately reduces the radiation dose the patient, an important consideration in the design of X‐ray imagingsystems.
This lecture will provide an overview on the use of photon‐counting detectors in dual‐energy CTimaging and image quality on clinical CT applications.
1. Understand the operation of direct‐conversion, photon‐counting detectors
2. Application of photon‐counting detectors to multi‐energy CTimaging
3. Use of photon countingCTsystem in a clinical setting
MO‐D‐210A‐03: Energy‐Sensitive, Photon‐Counting Computed Tomography: Opportunities and Technological Challenges36(2009); http://dx.doi.org/10.1118/1.3182242View Description Hide Description
Recent advances in the development of direct‐conversion, energy‐sensitive x‐ray detectors stimulate research in the domain of pre‐clinical and clinical photon‐counting x‐ray computed tomography(CT). The ability to quantify the energy of individual X‐ray photons allows for novel approaches to improve the soft tissue differentiation, material decomposition and labeling techniques, the suppression of beam‐hardening artifacts as well as the potential reduction of radiation dose. Moreover, spectral data acquisition enables the selective and quantitative imaging of certain contrast media on top of the conventional anatomy, by tuning an energy threshold in the detector system to the K‐edge discontinuity of the contrast generating element in the agent.
The present lecture will provide an overview of both, the opportunities and the technological challenges arising in the context of clinical, energy‐resolved, photon‐counting CT. Some examples of potential future applications will be given together with the most challenging technical difficulties encountered in the use of photon counting detectors in CT. The problem of counting photons at the high flux conditions of clinical CT will be discussed as well as the degradation of energy resolution by effects like pulse‐pileup or charge sharing.
All authors are employees of Philips Research.
1. To understand the physical basics of photon counting detectors and the implications on their use in x‐ray computed tomography
2. To learn about a possible future application of energy‐sensitive CT in connection with the identification and quantification of contrast‐agents by means of the K‐edge discontinuity in the attenuation.
3. To obtain an overview of the technological challenges to be overcome in order to realize photon‐counting CT in clinical practice.
Developments in Breast Imaging ‐ In Memoriam of Carolyn Kimme‐Smith
36(2009); http://dx.doi.org/10.1118/1.3182359View Description Hide Description
This presentation in the symposium is focused on the recent advances in digital breast tomosynthesis (DBT). DBT is a three‐dimensional (3D) x‐ray breast imaging technique. Several projection images of the breast are obtained from different angles, and image reconstruction is used to generate cross‐sectional slices (with 1 mm thickness) that are parallel to the detector. Several DBT prototype systems have been developed by different manufacturers through modification of screening full‐field digital mammography (FFDM) systems. The x‐ray tube gantry typically rotates around the compressed breast with a limited angular range (15 to 60 degrees) and acquires a limited number of images (11 to 49). In this presentation, an overview of the following aspects of DBT will be provided: 1. Different approaches to system implementation of DBT, and their advantages and limitations; 2. Comparison of different image reconstruction methods, including filtered back projection (FBP) and iterative reconstruction; 3. Special considerations for DBT compared to conventional projection mammography: scattered radiation, x‐ray spectrum and detector performance; 4. Methods for the evaluation and optimization of DBT image quality; 5. Clinical results of breast cancer detection using DBT and its comparison with projection mammography; 6. Combination of DBT with other 3D imaging methods for multi‐modality breast imaging applications.
36(2009); http://dx.doi.org/10.1118/1.3182360View Description Hide Description
Recent controversy over the efficacy of screening mammography highlights the need for more effective breast imaging techniques. While the sensitivity of mass breast cancer screening with mammography is in the range 60% – 90%, numerous studies have demonstrated that this sensitivity is reduced to less than 50% in radiographically dense breasts. Hence there is need for an alternative screening technique, particularly in women with radiographically dense breasts. A recent ACRIN trial shows that whole breast ultrasound had comparable performance to mammography. Breast MRI performs very well in this population, but high cost and variable specificity limit its application.
Conventional scintimammography has a high overall sensitivity and specificity for the detection of breast cancer. However sensitivity drops to ∼50% for the detection of breast lesions less than 15 mm in diameter, making it unsuitable as a diagnostic or screening tool. Molecular Breast Imaging (MBI) refers to a new technique that employs small field of view gamma cameras specifically designed for breast imaging. These cameras utilize either multi‐crystal arrays of Sodium Iodide (NaI) or more recently, use semiconductor materials such as Cadmium‐Zinc‐Telluride (CZT). Using Tc‐99m sestamibi as the primary radiopharmaceutical, MBI has been shown to have a high sensitivity (∼90%) for the detection of breast lesions < 10 mm in diameter. Hence, MBI may be a valuable screening tool, particularly for women in whom the sensitivity of conventional mammography is reduced by the density of the breast parenchyma.
This presentation will review the basic principles of molecular breast imaging, and present current clinical results. It will also present details of prototype systems currently under development and future plans for this technology.
1. Understand the limitations of current imaging techniques of the breast, including mammography,ultrasound,MRI and scintimammography.
2. Understand the basic concept of multi‐crystal and semiconductor‐based gamma cameras.
3. Describe the relative advantages and disadvantages of scintimammography and molecular breast imaging.
4. Understand the potential advantages/disadvantages and future applications for this technology in the detection and understanding of breast cancer.
36(2009); http://dx.doi.org/10.1118/1.3182361View Description Hide Description
Near‐infrared (NIR)spectraltomography applies and records multiple wavelengths of light in the 700–850 nm range through a transceiving multi‐fiber array in order to estimate absorption chromophores and scattering parameters within tissue. The technique provides important functional information associated with local concentrations of oxygenated and deoxygenated hemoglobin, oxygen saturation, and water and lipid fractions but on a centimeter spatial scale. In an effort to improve the spatial discrimination of these molecular characteristics of tissue, the technique has been combined with high resolution anatomical information obtained from MRI. Technical developments have lead to a simultaneous multi‐modality breast exam which combines the spectralNIR information with the data obtained during a dynamic contrast‐enhanced breast MR study. Commercial and custom breast coils have been populated with a 16‐channel fiber optical array and concomitant data acquisition module to explore the clinical potential of the approach. Clinical breast exams involving women with asymptomatic and symptomatic screening and diagnostic ammography along longitudinal studies of women with locally advance breast cancer receiving neoadjuvant chemotherapy have been performed. Results will be presented from anecdotal case reports as well as statistically valid findings from small clinical study designs and phantom experiments which indicate that the technique produces quantitative estimates of the NIR functional parameters at substantially improved spatial resolution relative to NIRspectraltomography when used in the breast alone.
1. To learn the basic physical principles of NIRspectraltomography.
2. To learn the basic clinical and physical characteristics of combined MR and NIRspectraltomography.
3. To learn the latest clinical breast imaging results from the technology as developed by Dartmouth investigators.
36(2009); http://dx.doi.org/10.1118/1.3182362View Description Hide Description
MRI of the breast is a highly sensitive technique for finding breast cancer. Suspicious findings include masses with an early rapid enhancement and a late plateau or washout. When suspicious findings are found on MRI, breast biopsy is done with vacuum‐assisted core devices or preoperative needle localization. This presentation will review state‐of‐the‐art findings diagnostic of cancer and review MRI‐guided biopsy techniques.
1. To learn MRI features of cancer.
2. To identify methods of MRI‐guided biopsy.
3. To learn limitations of MRI for finding cancer.
Advances in Ultrasound Imaging and Therapy
36(2009); http://dx.doi.org/10.1118/1.3182398View Description Hide Description
Definitive clinical diagnosis of prostate adenocarcinoma (PCa) requires histopathological confirmation of a tissue sample drawn from a 2D transrectal ultrasound(TRUS) guided biopsy. The prostate biopsy procedure, however, is plagued by high false negative rates (up to 34%) as early‐stage PCa is generally not visible on ultrasound. As a result, a negative biopsy does not rule out a diagnosis of PCa, as many tumors are missed on initial biopsy. In such circumstances, patients will undergo multiple repeat prostate biopsy (RPBx) to find undetected PCa. In a repeat biopsy, the physician must either avoid previously biopsied tissue (in cases of prior negative biopsy) or target the same anatomical site for patients with non‐diagnostic, atypical small acinar proliferations (ASAP). In cases of ASAP, there is a 40–50% chance of finding PCa on RPBx in the same anatomic location, so accurate targeting of prior biopsy locations is essential.
3D TRUS is hypothesized to be superior to 2D TRUS for accurate guidance and recording of the prostate biopsy procedure. Patients undergoing RPBx might benefit most from the hypothesized improvements, as previous Bx core locations can be viewed in 3D and used to guide a RPBx. Accuracy is also important when suspicious findings exist on other diagnostic imaging modalities, such as MRI or PET, are used to direct a TRUS‐guided biopsy.
In this paper, we describe a new method to guide prostate biopsy procedures using 3D ultrasound and extend it to prostate therapy. This new approach allows us to guide the biopsy to specific 3D targets in the prostate, record the biopsy locations in 3D, and register the intra‐biopsy procedural 3D ultrasoundimage with an MR image to guide the biopsy to specific locations in the prostate.
1. Understand the limitations of conventional TRUS‐guided prostate biopsy
2. Understand the methods that can be used to overcome the limitations of TRUS‐guided prostate biopsy
3. Understand the advantages of using MR images fused with 3D ultrasoundimages to guide prostate biopsy
TU‐D‐210A‐02: Catheter‐Based Ultrasound Technology for Image‐Guided Hyperthermia and Thermal Ablation36(2009); http://dx.doi.org/10.1118/1.3182399View Description Hide Description
Catheter‐based ultrasound applicators have been developed for delivering hyperthermia and thermal ablation for the treatment of cancer and benign diseases. This technology includes interstitial applicators for tumorablation and hyperthermia delivery during brachytherapy, transurethral devices for conformal prostate hyperthermia and ablation, and an endocavity applicator integrated with an HDR ring applicator for intrauterine hyperthermia. The common design incorporates arrays of multiple ultrasound transducers which provide dynamic axial and angular control of hyperthermia and thermal ablation. Performance was evaluated in phantom, excised tissue,in vivo experiments in canine prostate under MR temperature monitoring, clinical hyperthermia, and 3D‐biothermal simulations with patient specific anatomy. Interstitial and endocavity devices can tailor hyperthermia to large treatment volumes, with multi‐sectored and multi‐transducer power control useful to limit exposure to rectum and bladder. Transurethral applicators include curvilinear transducers with rotational sweeping of focused heating patterns, and multi‐sectored tubular devices capable of dynamic angular control without applicator movement. Curvilinear transurethral produce target conforming coagulation zones that can cover either the whole gland or defined focal regions. Multi‐sectored transurethral applicators can dynamically control the angular heating profile and target large regions of the prostate without applicator manipulation. High‐power interstitial implants with directional devices can be used to effectively ablate defined target regions while avoiding sensitive tissues. MR temperature monitoring can effectively define the extent of thermal damage and provided a means for real‐time control of the applicators. Preliminary investigations have also demonstrated ultrasound strain imaging can be used to monitor the ablatedtissue. In summary, these catheter‐based ultrasound devices allow for dynamic control of heating profiles along the length and angular expanse of the applicator during therapy delivery, are amenable to MR monitoring, and provide a minimally‐invasive technique for true 3D control of hyperthermia and thermal ablation. (Support NIH R01CA122276, R01CA111981, & R41CA121740).
36(2009); http://dx.doi.org/10.1118/1.3182400View Description Hide Description
The acoustoelectric(AE)effect is a well‐documented interaction between local pressure and electric resistivity. In recent years, the phenomenon has been revived for applications in biomedical imaging and therapy. My laboratory is developing two new approaches based on this concept: 1) ultrasonic imaging of current flow and 2) novel detection and imaging of an acoustic beam.
UltrasoundCurrent Source Density Imaging (UCSDI) is a new modality for mapping electrical current deep into tissue. This approach combines moderate acoustic pressure with recording electrode technology to directly imagecurrent densities. We have demonstrated feasibility of UCSDI in saline, tissue‐equivalent phantoms, neural tissue and direct mapping of the cardiac activation wave in the live rabbit heart. Although potential applications of UCSDI are diverse, we are focusing on enhancing electrical cardiac mapping during ablation treatment of arrhythmias. There are several potential advantages of UCSDI over conventional electrophysiology and electrical imaging, highlighted by 1) enhanced spatial resolution determined by the size of the ultrasound focus (<1 mm); and 2) automatic co‐registration of UCSDI with pulse echo ultrasound, depicting current density maps superimposed on heart structure and motion.
In addition to UCSDI, we are also developing the AEhydrophone as a new device for detecting pressure and imaging an ultrasound beam. As clinical applications for ultrasound therapy continue to proliferate—from lithotripsy to ablation treatment of uterine fibroids and cancer—the need for simple, rapid, and accurate estimates of the acoustic field become increasingly important. The AEhydrophone does not depend on a piezoelectric material; instead, a small region with high current density is used as a gain mechanism for detecting ultrasonic waves. I will present early results from initial prototypes and compare with conventional hydrophones, as well as simulations. The AEhydrophone has attractive attributes not typically seen with other devices, including simple construction, low cost, decent sensitivity, and resistant to damage at high intensity ultrasonic fields.
36(2009); http://dx.doi.org/10.1118/1.3182401View Description Hide Description
Ultrasonic imaging is one of the most important and still growing diagnostic tools today. Ultrasonic imaging is more appealing as a clinical imaging modality compared to such imaging modalities as magnetic resonance imaging(MRI), nuclear imaging, and x‐ray computed tomography(CT) in that it is more cost‐effective, non‐invasive, capable of real‐time operation, and portable while providing images of comparable quality and resolution. State‐of‐the‐art ultrasonic scanners offer real‐time gray scale images of anatomical detail with millimeter spatial resolution superimposed on which is a map of Doppler blood flow information displayed in color. Clinical applications of these devices are still expanding and the operating frequencies of these devices seem to inch higher and higher.
High frequency (HF) ultrasound (>20 MHz) yields improved spatial resolution at the expense of a shallower depth of penetration. There are a number of clinical problems that may benefit from high frequency ultrasonic imaging. Intravascular imaging with probes mounted on catheter tips at frequencies higher than 20 MHz with the highest frequency being 60 MHz has been used to characterize atherosclerotic plaque and to guide stent placement and angioplastic procedures. The medical efficacy of ultrasonic imaging of anterior segments of the eye at frequencies higher than 50 MHz in diagnosing glaucoma and ocular tumors and in assisting refractive surgery has been demonstrated. Applications in dermatology for defining tumor involvement and for monitoring treatment, in vascular surgery for characterizing atherosclerotic plaques, and in small animal imaging are also being investigated.
To further expand the role of HF ultrasound, probe performance must be improved since current HF ultrasound imaging devices, dubbed “ultrasound biomicroscopes” or “UBMs” by a number of investigators use mechanically scanned single element transducers. Compared to electronic scanning with linear arrays, mechanical scanning of a single element transducer has three major drawbacks: poorer resolution, slower frame rate, and motion of the probe. Single element transducers can only produce beams with a fixed focus which means the spatial resolution of the device is best only within the depth of focus, i.e., in a very tight zone and degrades rapidly beyond the focal point. Mechanical motion of the transducer limits the frame rate, is unreliable, and may cause discomfort and, at worst, hazard to the patient.
My presentation will introduce several HF ultrasound transducers currently being developed by the NIH Transducer Resource Center at the University of California. Emphasis will be placed on the construction and implementation of high frequency array transducers.
1. Familiarize audience with high frequency ultrasound transducers and their applications.
Advances in CT
36(2009); http://dx.doi.org/10.1118/1.3182432View Description Hide Description
Purpose: Dual‐energy (DE) CT has received much attention in recent years. There are various approaches to the data acquisition, differing in terms of the simultaneity, geometric alignment, data collection efficiency and completeness, and flexibility. We present a fast‐kVp switching (FKS) method in combination with advanced imagegeneration process, and demonstrate the efficacy of our approach with phantom and clinical results. Methods and Materials: In FKS, the input to the x‐ray tube is rapidly changed between two kVp‐settings in adjacent views. To ensure the signal fidelity, x‐ray generator was redesigned to minimize its input impedance, new scintillating material was developed to reduce the primary speed to a small fraction of a millisecond, and the data acquisition system was designed to sample up to 7Hz frequency. To take full advantage of the co‐registered dual‐energy projections, advanced algorithms were developed to enable projection‐space pre‐processing, calibration,reconstruction, and spectral image displays. These techniques allow the production of mono‐energetic CTimages, as well as different material decomposed images.Results: Phantom experiments and computer simulations show that because of the near simultaneous data acquisition in terms of timing and orientation, mis‐registration due to patient motion is kept to a minimized. FKS also allows the projection‐space processing of the dual‐energy signals and results in a significant improvement in terms of beam‐hardening artifacts. Advanced algorithms allow superb noise suppression in the mono‐energetic images as well as material‐decomposed images. Clinical studies demonstrate the importance of beam‐hardening reduction, metal artifact reduction, and the accuracy of material decomposition. Conclusion: We present a fast‐kVp switching approach to the dual‐energy data acquisition and advanced algorithms to the dual‐energy reconstruction. Phantom and clinical experiments have demonstrated that such approach provides superior image quality and clinical utility.
36(2009); http://dx.doi.org/10.1118/1.3182433View Description Hide Description
The first investigations of dual energy methods for CT were made by Alvarez and Macovski in 1976. They demonstrated that using a conventional X‐ray source having a broad energy spectrum, one can still separate the attenuation coefficient into the contributions from the photoelectric effect and Compton scattering. Thereafter, several applications were reported utilizing dual energy CT, focusing primarily on lung,liver and tissue characterization. However, all of the approaches were limited in some manner and not able to be used routinely in clinical practice. The primary limitation was that data for the different tube voltages were acquired at two different times.
In the 1980s, it was possible to acquire dual energy data nearly simultaneously using a modified commercial CTsystem (Siemens DR scanner). During the rotation of the tube‐detector pair, the tube voltage was switched quickly for each detector reading between the high and low settings so that two sets of raw data (projections) were acquired nearly simultaneously at two different tube voltages. Unfortunately, the only practically and routine used Dual Energy application was bone densitometry, and this was not sufficient to let this system survive. The main limitation of the tube voltage switching approach of the DR scanner was that a) the delivered dose of the low kV readings was — compared to the dose delivered by high kV reading — too low, b) the necessary adaptation (increase) of the tube current at low kV was not adequately possible, and c) the impact on image quality due to the switching between high and low kV was too big.
The technical limitations of the DR system were overcome with the introduction of the first dual source system in 2006. In contrast to a single source system, dual source CTsystems have two separate tube/detector pairs that are mounted orthogonally on the rotating slip ring. This design provides the flexibility to adjust not only the tube voltage but also the tube current for both tube/detector pairs and allows simultaneous data acquisition. Although the raw projection data do not match identically because of the 90 degree offset between both systems,reconstructed images — although measured at different tube position — are acquired at exactly the same time. This is true for regular spiral or sequential scans and also for the gated scans used in cardiac examinations. Data from both tube/detector pairs are reconstructed separately into two different image stacks. Image‐based post processing then is used to extract the dual energy information.
In 2008 an even improved version of the established Dual Source CTsystem was introduced — the SOMATOM Definition Flash (Siemens Healthcare, Germany). The main changes are that the limited FOV of the second tube‐detector system was extended so that nearly all of the relevant anatomy fits into the FOV of the second system. In addition, a so called ‘selective photon shield’ was added to the high kV beam. This additional filtration (tin pre‐filtration) improves the dual energy separation significantly. The overlap of both spectra is minimized and therefore the Dual Energy separation e.g. of bone and iodine is increased by nearly a factor of 2; noise in dual energy post‐processed data, e.g. virtual non‐contrast images, reduced significantly.
Dual energy data from dual source systems are typically prost‐processed using image‐based DE methods.
Some of the DE techniques focus on the ability to separate two materials based on their CT‐values at high and low kV. For this purpose, pixel data are put into a 80–140‐kV diagram. Pixels containing a certain material (e.g. calcium or iodine) in different concentrations or at different densities are on a straight line. Drawing a separation line for example between the lines for calcium and iodine allows separating both materials in the 80–140‐kV diagram and therefore also in the DE CTimages. This technique is for example used for applications like bone removal or lung vessels. A similar principle applies to the differentiation of a vessel lumen filled with iodinated blood and calcified plaques.
Other image based Dual Energy methods focus on the quantification of iodine distribution in tissue. This allows, for example, the visualization of perfusion defects in the parenchyma in the case of a pulmonary embolism by displaying the iodine concentration of the different areas of the lungs. Another possible application might be the visualization of pure iodine enhancement in abdominal organs like the kidneys. In this case, additional information about the pure enhancement might be helpful for the differential diagnosis between a hemorrhagic cyst and a renal cell carcinoma. The basic method for post‐processing in those cases is the so called three material decomposition. As a first step, three material/tissue types are defined that are of interest and are found in a certain anatomical area. In the liver region, this might be, for example, “soft tissue”, “fat” and “iodine”. Based on values from literature or clinical experience, these materials are drawn into an 80–140‐kV diagram, where they span a triangle. To evaluate a certain area of interest, the respective CT values in the 80 and 140 kV image are plotted into the existing diagram. Projecting this point onto the line between “fat” and “soft tissue” allows the calculation of true iodine enhancement in the region of interest.
By applying this method to the whole field of view, an image showing true enhancement can be calculated. In addition to that, the image showing true enhancement can be subtracted from the weighted sum of the low and high voltage image and by doing so a “virtual” non‐contrast image be calculated.
Base on the above mentioned techniques in total 12 dedicated dual energy post‐processing applications are available for DE images from dual source systems. In addition, special visualization techniques like ‘optimum contrasts’ and ‘monoenergetic’ are available. Due to the recent improvements with the second version of a Dual Source CTsystem even more applications can be expected in the near future.
36(2009); http://dx.doi.org/10.1118/1.3182434View Description Hide Description
The emerging of fast‐rotating MDCT scanners, and the recent approach of the, so called, “Slice War” to its saturation, opened new opportunities for CT evolution. New routes to extend CT applications beyond anatomical imaging have been pursued. Quantitative functional imaging seems to become one of the main trends in this process, with perfusion applications and multi‐energy CT methods as the leading techniques.
To enable a simultaneous dual energy CT scanning, using a single X‐ray tube, without FOV, or sampling limitations, a special double‐layer detectorCT has been developed in PHILIPS Healthcare. The conventional CTdetection pixel has been reconfigured, to consist of 2 Scintillator layers, read simultaneously, by a double‐layer, side‐looking photodiode. The top‐layer scintillator has been chosen to contain, relatively, low atomic‐number elements, while having a very high light output, and very low afterglow. Its thickness has been optimized to achieve a maximum spectral separation between Iodine and Calcium. A 2‐mm layer of GOS has been used for the bottom Scintillator layer, to enable stopping of 99.8% of the X‐rays, transmitted through the top layer, without limiting the GOS light collection which is done sideways. The raw data of each of the detector layers is reconstructed separately, resulting in a Low‐Energy and a High‐Energy HU images. Thus, each slice‐pixel has two HU values assigned to it, the low and the high energy, respectively. In addition, a weighted sum of the two raw data is reconstructed to give the conventional CTimage, while the weighting factor can be modified to present the combined image at any desired mean energy, within the system energy range. The pixels of each slice are mapped on a 2D spectral scatter plot, HU_Low_E VS. HU_High_E. The physics nature of the radiation interaction in this energy range, determines that varied concentrations of the same material (same effective atomic number) are represented along a straight line on this map. Also, as will be shown, this method enables the separation and quantification of specific materials, like Iodine, from mixtures with other compounds.
The system capability of material identification and quantification enables plaque characterization, non cathartic virtual colonoscopy, virtual non‐contrast imaging, and an accurate quantitative mapping of perfused Iodine for advanced functional CT applications. More than 2000 human patients have been scanned in a fully operated system, as well as plenty of research animals, at the Hadassah University Hospital in Jerusalem. The animal experiments (rabbits) aimed, mainly, at testing the system capability with new, targeted contrast agents, indicating a great opportunity for the use of multi‐energy CT, and full spectralCT in the future.
1. Understanding the clinical opportunities associated with a dual‐energy CT.
2. Understanding the unique advantages, the tradeoffs, and the limitations of the double‐layer detectorCT approach.
3. Understanding the principles of the material identification and analysis method in the image domain, chosen for this system.
36(2009); http://dx.doi.org/10.1118/1.3182435View Description Hide Description
Dual energy X‐ray computed tomography(CT) is becoming more important, in part due to the availability of multi‐row scanners that collect data using two source energy spectra. An X‐ray source emits a spectrum of energies for each setting of the source. The energy dependence of Beer's Law leads to the relationship between the overall attenuation function and the mean measured data; the statistics of the data given the mean data determine a more accurate model. In many approaches, linear approximations to these inherently nonlinear phenomena are used to derive algorithms.
The data can be used to form two attenuation images that can be interpreted as coefficients corresponding to component attenuation spectra; given additional constraints such as nonnegative attenuation and a sum of fractional contributions equal to one, a third attenuation image can be recovered. Component attenuation spectra may correspond to physical processes (photoelectric and Compton effects), to known materials (for example, fat and bone), or to spectra that nearly span the set of spectra (for example, computed using a singular value decomposition).
Three classes of dual‐energy algorithms are discussed: prereconstruction, postreconstruction, and statistical iterative reconstruction (SIR), the latter recovering multiple images simultaneously using all data. The tradeoffs in performance of these algorithms are discussed. Analytical techniques for bounding achievable performance based on data models may be used in system analysis and design. Performance in actual systems may vary somewhat from the analytical predictions, due to inevitable differences between the model used for computing reconstructions and the clinical devices. This discrepancy in actual and predicted performance can be quantified in many situations. The approaches described extend in a straightforward fashion to data acquisitions using three or more energy bands.
1. Understand the assumptions underlying dual‐energy image reconstructions.
2. Understand the centrality of the nonlinear data model in dual‐energy CT.
3. Understand the different methods that may be used to compute the image reconstructions.
4. Understand the role of image performance prediction and limitations of performance prediction.