Few specialties are as technology-intensive as radiology, and advances in the tools of medical imaging have played a key role in the discipline?s growing importance. To provide readers with a window on the near future, Medical Imaging asked five industry experts to share predictions for their modalities in this electronic roundtable discussion.

MRI: Jacques Coumans, PhD, is vice president of global marketing and sales, magnetic resonance, at Philips Medical Systems, Best, the Netherlands.

CR and DR: Robert Goldy is national marketing manager of imaging systems at FUJIFILM Medical Systems USA Inc, Stamford, Conn.

CT: Scott Goodwin is vice president of CT at Siemens Medical Solutions, Malvern, Pa.

Ultrasound: Gordon Parhar is director of the ultrasound business unit at Toshiba America Medical Systems, Tustin, Calif.

Molecular Imaging/Nuclear Medicine: Jean-Luc Vanderheyden, PhD, is the global molecular imaging leader at GE Healthcare, Waukesha, Wis.

MEDICAL IMAGING: What are the greatest technical hurdles that you must overcome to take your modality to the next level in spatial and temporal resolution?

Robert Goldy, FUJIFILM Medical Systems USA Inc.

GOLDY: For DR and CR systems designed for routine radiographic work, the current levels of spatial resolution, which typically range from a high of 5 line pairs (lp) per mm to 3.5 lp per mm, are adequate for routine imaging. Therefore, increases in spatial resolution will not lead to improved diagnostic content. One area of research that is showing promise is the use of higher resolution for pediatric applications for determination of child abuse, as well as neonatal imaging. For this application, a resolution of 10 lp per mm, as found in CR mammography-capable systems, may aid diagnosis.

CR, by its nature, is not capable of temporal capture. This will not have a negative effect on CR utilization rates in the coming years, based on its overall proven capabilities. For DR systems, temporal resolution is valuable only from the perspective of allowing new capture techniques, such as tomosynthesis. Although it is an interesting technology, it is still too early to determine whether this application will gain traction among DR users, in light of competing technologies. Dual-energy subtraction is also possible with DR, as a result of temporal resolution. Despite being available for many years as a CR solution, implementation of this technology has remained limited. It is my belief that, from a commercial perspective, tomosynthesis and energy subtraction will remain niche technologies, rather then seeing broad clinical use.

More important than focusing on a single measurement, such as spatial resolution, is for the industry to take on a more holistic approach that strives to find the best balance among detective quantum efficiency, spatial resolution, noise suppression, and image processing.

GOODWIN: The Somatom Definition sets the clinical standard for both of these imaging concerns. Physicians are not asking for more slices and the same resolution; they are asking to see more clinically relevant structures and pathology. It will be very important that clinicians not confuse coverage with temporal resolution. More slices are not better temporal resolution.

Jacques Coumans, PhD, Philips Medical Systems

COUMANS: The main parameters here are magnetic field strength and radiofrequency (RF) coil technology, along with high-performance gradients and smarter pulse techniques. Further developments at 3T will move the field toward even higher-resolution imaging and even faster scanning. Even today, however, we are delivering the capabilities of 2K x 2K imaging over fields of view of 20 cm. I do expect that what we will learn at 7T, specifically in the area of transmit SENSE technology to combat specific absorption-rate limitations and RF nonuniformities, could be of great help at lower field strengths, such as 3T, as well. Such techniques as kt BLAST help improve temporal resolution in free-breathing or single-heartbeat cardiac studies.

VANDERHEYDEN: Molecular imaging is a technique that uses sophisticated diagnostic imaging equipment and systems to visualize specific signal molecules, based on their chemical and biological properties. This enables physicians to peer into the living body in order to identify, treat, and monitor diseases; their progression; or medical conditions at a molecular level. Molecular imaging can be carried out using several different modalities, with the most commonly used one being PET/CT.

The limitation is the ability to see enough of the targeted cells inside 1 mm3. New developments in detector sensitivity, integrated with more precise specificity and sensitivity of signal molecules, will continue to improve, and, therefore, increase the probability of detecting the early presence or likelihood of disease.

Today, the focus of molecular-imaging research is on highly sensitive nuclear-medicine techniques, but physicians and scientists expect that all imaging modalities will one day be involved in molecular imaging. GE Healthcare has molecular-imaging products in development for other modalities, including MRI, ultrasound, CT, and optical imaging. In the future, imaging will be required to produce information about molecular processes as well as anatomy and morphology.

The PET/CT images that we are capable of creating today have vastly improved over what was previously available, in terms of spatial resolution and clarity. The sensitivity of cameras and the biological specificity of imaging agents will be crucial going forward.

PARHAR: These days, the big news in ultrasound is volumetric imaging. Acquiring volume data sets and using a workstation for postprocessing can greatly improve workflow and productivity. The quality of the volume data set acquired is dependent largely on the transducer. Current two-dimensional transducers do not provide the image quality necessary for state-of-the-art imaging. This is especially true in cardiac imaging, where improved sensitivity, improved parallel-beam?forming technologies, and high-speed image reconstruction are necessary. As these obstacles are overcome, physiological frame rates will be achieved. The net effects are an improved diagnosis and better patient outcome.

MEDICAL IMAGING: What role do you expect CAD to play in your modality’s future?

Scott Goodwin, Siemens Medical Solutions.

GOODWIN: CAD will be a key tool in opening new clinical avenues for diagnosis. Siemens Medical currently has two such programs: Colon PEV and Lung CAD (FDA approved). As spatial and temporal resolution improve and CT data sets continue to grow, we see demand for these tools increasing. The ultimate outcome for our CAD products is to improve detection accuracy and reduce false positives.

COUMANS: This will become increasingly important. As a consequence of the enormous number of images that can be generated in even a few minutes of MRI scanning, the wealth of information in such data needs to be presented in an easy format, and the clinician will be directed using CAD functionality. Already, such tools are available for breast imaging, and this will expand to other applications, such as prostate, heart, and brain imaging. The important workflow?related step will be to transfer this CAD functionality from the modality-specific workstation (where things will be optimized) to the PACS workstation?which is, more and more, becoming the preferred work spot for the radiologist. As a consequence, I expect that the concept of CAD plug-ins will take hold for these PACS workstations, which fits seamlessly with our company?s drive toward sense and simplicity.

VANDERHEYDEN: CAD software can reduce image processing and interpretation time significantly, while helping to improve the physician?s diagnostic confidence. CAD software can provide an efficient, standardized method for processing and interpreting these data. It highlights and color codes suspicious areas and automatically corrects the images for patient movement. The ability to use image interpretation with dedicated software and the combination of this information in the broad aspects of the patient?s electronic record is likely to make a lasting impact on the delivery of molecular imaging.

PARHAR: CAD has great potential for categorizing solid breast nodules. I see CAD playing an even bigger role as it is used with volumetric (four-dimensional) imaging. I am sure that CAD, when coupled with volume data, will improve workflow and productivity, and provide a more definitive diagnosis.

GOLDY: Mammography CAD is taking on an increasingly important role in supporting diagnoses from examinations produced with full-field digital mammography. Based on this continued success and growing acceptance, we believe that CAD will see significant growth in the next 10 years, in general. We also believe that, as it reaches a state of maturity, new applications will evolve to extend the capability to other anatomic regions.

Significant early developments in this area are advances made in chest CAD, which holds significant potential in the near term. We actually look at CAD as broader in scope than the current understanding in the market of where CAD is focused (primarily on pattern-recognition software). In our development initiatives at Fuji Medical, we include advanced image processing, optimized detector technology, energy subtraction, temporal comparison, pathology correlation, and anatomical segmentation as ares that need to evolve together to maximize the modality?s potential. One of our core competencies is image-processing technology, and we believe that CAD should include any image-processing technology that aids physicians to detect pathology?for instance, energy subtraction, temporal subtraction, and disease-specific image processing, such as a pneumothoraces algorithm that we developed with the Veterans Administration, Baltimore.

MEDICAL IMAGING: What improvements will software bring to your modality in the near future?

COUMANS: Most importantly, advanced algorithms currently are developed to transform MRI scanners toward one-click operation. Already, Philips Medical MRI provides completely automated planning for the most frequently scanned anatomic regions, such as the head, spine, and knee. Eventually, this SmartExam feature will be expanded to all anatomic areas. It is part of fully automated examinations using comprehensive ExamCards, which are clinically relevant compilations of individual protocols that enable automated planning, scanning, and processing. If you think about Cardiac MRI, for instance, it is to be expected that the concept of ExamCards will become sophisticated enough to prescribe which MRI data set will need to be processed for quantification, and what results should be included automatically in the concise cardiac report to the referring physician. So, workflow?supporting software will gain significantly in importance on MRI scanners.

VANDERHEYDEN: Molecular imaging also provides quantitative information, such as the metastasis of a tumor. Analysis of the metabolic or physiologic uptake of imaging agents by cells requires software with the appropriate algorithms. A system that combines a PET/CT with a computer and software package could help provide physicians with clinical information that can characterize disease and also suggest the most appropriate treatment options for each patient.

Gordon Parhar, Toshiba America Medical Systems.

PARHAR: As software improves and evolves, the performance and capabilities of the system also will improve. Typically, the number of hardware components is reduced, leading to smaller and smaller scanners. You can see the effect in the market. Hand-carried ultrasound scanners are ubiquitous and getting smaller, faster, and more affordable with each passing year, largely due to improved software algorithms.

GOLDY: Image processing is one of the core competencies in our modality. CAD software will contribute to physicians? performance, but we also see a trend toward improved image-processing capabilities, particularly in the ability to suppress noise at lower dose levels. Fundamental to the new techniques is the development of complex software algorithms to support image processing and image reconstruction.

GOODWIN: Workflow improvements will be key to enhancing the diagnostic process. Having the software tools where you need them, when you need them, will create increased efficiency for the patient and the diagnostic outcome. This also will help reduce overall health care costs. For example, a server-client solution will be an enterprise-wide clinical benefit for clinicians to access their patients? CT data, in real time, without delay.

MEDICAL IMAGING: What technical advances do you expect to improve your modality?s ability to deliver functional information?

Jean-Luc Vanderheyden, PhD, GE Healthcare.

VANDERHEYDEN: GE Healthcare is evaluating nuclear-camera design and new detector materials in an effort to create imaging systems that can detect disease early and generate images more quickly to ease the patient?s experience.

Imaging devices and software that process the specific information of each molecular-imaging agent will be crucial in ensuring that the technology is properly used.

PARHAR: Ultrasound can be used with molecular imaging. It is fast, affordable, and widely available. Ultrasound is able to image structure and function, and in the future, it might be used to deliver drugs to targeted lesions. The development of microbubble contrast agents has opened new opportunities, including new functional-imaging methods, the ability to image capillary flow, and the possibility of molecular targeting using microbubbles.

GOLDY: Going into the future, the overall role of DR and CR will remain as it is today: a strong diagnostic tool that provides significant information and benefit in relationship to its cost. We do not hold the view that functional imaging, as currently defined, will migrate into conventional radiography, as defined by CR and DR.

Advancements in image-processing technology, however, are going to be required to deliver enhanced information, such as quantification of granular density and cardiothoracic ratio. In addition, reducing noise and increasing temporal resolution will improve the accuracy of the measurement result so that it will become more reliable.

GOODWIN: Dual-source, spiral dual energy with increased spatial and temporal resolution will open many clinical doors related to function. For the first time, dual-source scanning provides the clinical benefit of being able to detect different attenuation levels of structures/pathology all in one scan?something single-source systems cannot do.

COUMANS: The establishment of high-field 3T MRI as the new high-end MRI platform permits microscopic resolution and near?real-time 3D imaging. In addition, novel, tissue-specific contrast agents and advanced processing will become available. As a result, MRI will increasingly provide quantitative information allowing functional assessment. The sophistication of the MRI pulse sequences, tuned to specific tissue characteristics, and the operation of pulse sequences and protein-specific tissue or disease agents, will increase MRI?s ability to deliver functional information.

MEDICAL IMAGING: What role, if any, will this modality play in the delivery of therapeutic interventions by 2010?

PARHAR: Ultrasound will be used more frequently for RF ablation. RF is a type of electrical energy that has been used in medical procedures for many years. Electrical energy is used to create heat. The heat is created in a specific location, at a specific temperature, for a specific period of time, and it ultimately results in the death of unwanted tissue (tumors). During an RF procedure, an ablation probe is placed directly into the target tissue. Ultrasound is a cost-efficient way to guide the ablation probe and monitor the procedure.

GOLDY: CR systems will continue to act as a means of verification, as they do today in such applications as portal imaging. CR also will remain a very effective solution to providing imaging capability in the operating-room environment and other areas outside radiology; it will remain an important support for interventions. With the recent introduction of portable DR solutions in addition to CR, there will be a significant increase in DR systems supporting interventions. CR and DR systems also are strong enablers for improved orthopedic templating, which will be the standard of care by 2010.

CR is now used in acquiring images in high-energy radiation, such as gamma-ray or electron-beam applications. Combined with advanced image processing and higher-resolution imaging technology, DR will play a major role in imaging and monitoring therapeutic interventions.

GOODWIN: CT is everywhere, with examples including PET/CT, SPECT/CT, DynaCT (angiography CT), and Primatom (linear accelerators and CT). All these hybrid devices have CT as a cornerstone technology. It is clear that as the clinical process changes to one of diagnosis and intervention, CT technology will be at the forefront.

COUMANS: Already, MRI-guided clinical interventions are performed; these include breast biopsies, deep-brain device placements, cardiac interventions in pediatric cardiac patients, and ablation of lesions, in addition to radiation-therapy planning.

It is expected that MRI will extend its functionality in these areas, and probably will play a substantial role in such areas as electrophysiology. MRI-guided high-intensity focused ultrasound will become an important tool for ablation, and possibly also for more advanced applications, such as enhanced drug therapy and targeted molecular imaging. Pediatric interventional applications will gain importance due to the intrinsic absence of a radiation dose in MRI.

VANDERHEYDEN: The possibilities for this modality are significant in both diagnosis and management of disease. We have agreements in place with pharmaceutical companies that are designed to help these companies create and validate therapies using our molecular-imaging agents.

We believe that this model will only be strengthened over time as the value and benefit of delivering the right treatment to the right patient at the right time are demonstrated. To be accepted, molecular imaging will need to prove itself with optimal patient outcomes, allowing safer, targeted treatments and more cost-efficient health care. MI

Kris Kyes is technical editor of  Medical Imaging. For more information, contact .