It is based on decades-old tomogram technology; even more surprising is the concept behind it is inspired in part by formulas conceived by a mathematician on a blackboard at the turn of the 20th century.
But until recently, the “synthesis” portion of it had no engine. Today, incredibly high-speed computing and imaging technology is bringing tomosynthesis to the forefront of exciting applications for treating cardiovascular, kidney and lung disease. But in the near term, imaging companies are betting the farm and are currently pursuing Food and Drug Administration approval for what they believe is the biggest immediate promise of tomosynthesis: detecting, and one day, treating breast cancer.
Many manufacturers with whom Outpatient Care Technology spoke have large clinical studies currently underway at prestigious academic medical centers while awaiting FDA clearance to market their new tomosynthesis systems. That approval is anticipated near the end of 2008.
They see tomosynthesis technology as a natural progression of full-field digital mammography, and are quite optimistic of building on FFDM’s gains to go after the more than 20 percent of dense breast tumors left undetected by current technology. Digital breast tomosynthesis promises to provide a quantum leap in screening and diagnostic capability, including three-dimensional lesion localization and contrast-enhanced 3D imaging.
One major player, Bedford, MA-based Hologic, visually demonstrates the excitement in a compelling 10-minute video on its Web site, where part of a house is obstructed by a tree in a photograph, or intersecting shadows cast on a wall by people on a sidewalk show convincingly the limitations of the two-dimensional world of current digital mammography.
Spotting Disease in 3D
In the simplest of terms, tomosynthesis is a 3D imaging technology that involves acquiring as many as 15 different images of a stationary and mildly compressed breast at multiple angles during a short scan. The images are captured by moving an X-ray tube in an arc over the breast, and mathematically synthesizing the tomogram into a volumetric view to show tissue as layers. A tomosynthesis exam typically takes several high-resolution frames per second and lasts as little as five seconds.
Proponents of the technology claim that it overcomes the limitations of 2D imaging by eliminating the problems of tissue overlap and signal noise.
While the concept of tomosynthesis has been widely discussed and researched for many years for different parts of the human anatomy, it has until now lacked the means to capture images quickly and accurately.
“Tomography existed 40 years ago, basically doing a set of simple projections at different angles. What tomosynthesis is doing is taking the tomography application and translating and reconstructing projections not as one single set of images but as a whole volume,” explained David Caumartin, global general manager for the mammography business of GE Healthcare Clinical Systems, from his office in Paris. “With tomography, you’d have 10 slides coming from 10 projections. What we’re doing with tomosynthesis is using projections as a rogue data set to reconstruct them in the volume.”
More than 20 years ago, Loren Niklason was tinkering with tomosynthesis as a medical physicist in the department of radiology at Massachusetts General Hospital, where he witnessed firsthand the promising benefits of tomograms on gall bladder and kidney disease. Then, computed tomography was approved, and interest in conventional tomograms began to wane. But not with Niklason, who today oversees Hologic’s burgeoning tomosynthesis program. “The reason tomosynthesis wasn’t adapted until recently was simply that digital detectors just weren’t ready,” Niklason told OPCT in his trademark gravelly voice. “The tomogram was not a true 3D technology and you had to give an entire dose to get one plane; so to get multiple planes, you had to keep dosing the patient.”
Safer, Faster, More Accurate?
And therein lies one of the fundamental benefits tomosynthesis promises to bring to breast cancer detection: quantum increases in accuracy with less radiation exposure. “CT typically uses a very high dose, but with tomosynthesis, you can do one sweep with the tube; one exposure and you can get all the slices at once,” said Niklason. “With tomosynthesis, you’re reconstructing all the different planes by simulating that motion at what it would have been at each plane in the body. So basically, you back project to whatever plane you want to see and can reconstruct an image at that plane.”
Another benefit of tomosynthesis is much better localization of breast lesions, an inherent limitation of conventional mammography. “In traditional mammography, the dense parts of a breast are projected on a 2D image, and it’s hiding some cancerous masses and microcalcifications,” said Jonny Eser, women’s health product manager for Malvern, PA-based Siemens Medical Solutions. “By slicing the breast, you can get in-depth, and when you come to a plane where those masses and calcifications are, they really pop out on the screen when you look at it.”
Consequently, for the 40 to 60 percent of the female population who have dense breasts, the technology is very promising. “Tomosynthesis offers the potential for better outcomes with these women,” said Caumartin. “Not only does it have the potential to detect more cancer, it also has the potential to lower or at least keep recall rates steady.”
Noted Niklason: “Conventional mammography has two primary problems: One is that we miss cancers because they’re hidden, and the other is we call back a lot of women for further testing. Everyone expects tomosynthesis to be good with dense breasts, and that’s where mammography is weakest because many tumors can hide in dense breasts. We also think that the breast is very ideally suited for tomosynthesis because in the breast, all anatomical structures radiate or orient outward from the chest wall toward the nipple. In tomosynthesis, we can actually blur structures very well in all directions except the direction the tube moves. The other advantage is there are no big bony structures to create a lot of artifacts in the breast.”
Because of the speed and accuracy it promises, tomosynthesis also has the potential to vastly improve efficiency and throughput in an imaging center. For radiologists, this means faster case review and more confident readings, according to
Finally, tomosynthesis could actually persuade more women to get mammograms because exams will be faster – and more comfortable. Unlike conventional mammograms, breasts are only slightly compressed during a tomosynthesis exam because all that is needed is to keep them immobile.
With any emerging technology, there are possible downsides. For tomosynthesis, those include the high initial cost. “I think it will start at a fairly high price but certainly not extremely more than digital mammography,” said Niklason. “But the cost of the technology will gradually go down as it does with
There are other caveats to also consider, according to Siemens executives. These include incrementally higher demands for storage (data from just one tomosynthesis exam could easily exceed a gigabyte or more in size), the potential need to invest in computer-assisted detection (CAD) equipment and larger, more powerful displays.
In recent months, manufacturers have released promising results from clinical trials held at the University of Michigan Health System and Massachusetts General Hospital.
In the University of Michigan study of 30 women, researchers reportedly concluded that tomosynthesis resulted in better detection and fewer recalls. In fact, tomosynthesis spotted cancers missed by conventional mammography in nearly 20 percent of the same women.
In the Massachusetts General Hospital study, which involved more women and multiple sites, researchers studied the clinical value of using tomosynthesis as an adjunct to FFDM and concluded that the procedure resulted in fewer recalls and improved specificity in spotting lesions. The custom-built tomosynthesis system also reportedly sped up image reconstruction by a factor of 60.
Hologic released results of a multicenter, multireader study it launched in 2004 to demonstrate the value of using tomosynthesis as an adjunct to FFDM and also found that the procedure improved radiologists’ cancer detection rate and screening recall rate over conventional digital mammography.
Manufacturers with whom OPCT spoke expect the outpatient imaging market to warmly embrace tomosynthesis, provided it’s proven and affordable. “Tomosynthesis is a tool that will remain affordable, simple and efficient for screening because mammography is first and foremost about screening,” said Caumartin.
How quickly it is adopted in the outpatient market will depend on training. “For those with existing FFDM systems, the learning curve will probably be low,” he added. “From a workflow standpoint, only one compression will be needed, so exams will go faster. Patients will love it and likely will demand it. Providing these systems have the right productivity and visualization tools, radiologists will too.”
Niklason remains optimistic. “It’s hard to say what’s going to happen,” he noted. “We certainly believe that at least for mammography, tomosynthesis adoption is going to be widespread and a really big thing in outpatient imaging. The most promising application of tomosynthesis right now is in the area of breast care, and clinicians are very interested. For the past 10 years, people have been seeing these dramatically improved breast images from tomosynthesis. When we talk with some time to get through all of the regulatory hurdles.”
Others are more cautious. “Once it is approved by the FDA, I would say tomosynthesis’ adoption will initially be more in large academic medical centers and hospitals and research facilities because it has to make sense,” said Eser. “Once proven it provides better healthcare, the industry will have to begin considering better reimbursements. Most outpatient centers probably won’t invest this kind of money until they are comfortable getting reimbursed.”
Manufacturers are somewhat divided on the future of tomosynthesis in breast care. Some maintain its biggest promise will remain as a screening tool. Others argue as technological advances come into play, tomosynthesis may play a larger role in treatment.