Information overload could swamp the imaging industry if data obtained by CTs, MRIs, and ultrasound scanners and available on networks overwhelm the processing power of computers. TeraRecon and Real Time Visualization have each come up with partial
Information overload could swamp the imaging industry if data obtained by CTs, MRIs, and ultrasound scanners and available on networks overwhelm the processing power of computers. TeraRecon and Real Time Visualization have each come up with partial solutions to the problem. Now these two companies are preparing to combine their technologies.
TeraRecon is buying Real Time Visualization (RT Viz), the division of Mitsubishi Electric that makes graphics accelerator boards for PCs. The strategy behind the union of these two companies is to integrate into PCs the graphics accelerators made by RT Viz with high-speed data processors made by TeraRecon. Together they will form a cost-effective computing engine with the potential to handle the flow of data from multislice CTs, ultrafast MRI scanners, 4-D ultrasound systems, and networks supporting 3-D images.
"With this acquisition, TeraRecon has added a key element for accelerating our plans to deploy comprehensive, innovative solutions for next-generation imaging systems," said Dr. Motoaki Saito, president and CEO of TeraRecon.
The two technologies of these companies are highly complementary. TeraRecon is scheduled next month to release the first of its XTrillion chips, which will allow the reconstruction of volumetric data moments after they are acquired by multislice scanners (SCAN 12/20/00).
RT Viz is already selling its VolumePro, which supports the manipulation of 3-D data sets. Combining the two technologies will allow the immediate reconstruction and inspection of volumetric images.
"It will allow true 3-D and 4-D ultrasound on a PC platform; it will lead to the ability to handle large sets of volumetric data coming from CT and MR," said Steve Sandy, TeraRecon vice president of marketing. "We can even combine that with patient data and store all those data together."
At the last RSNA meeting, RT Viz demonstrated VolumePro Net, a server-based product that distributes images over a network composed of desktop and laptop PCs. Sandy envisions a time when VolumePro Net is expanded to include XTrillion technology.
In such a configuration, imaging and patient data would be stored and routed over a network driven by two or more servers. A central server would hold the raw data, while other servers would be distributed among individual workgroups, in radiology, orthopedics, or surgery, for example.
On request, the central server would transmit raw data to these distributed servers, which would use XTrillion technology to instantly reconstruct the data. The slice or 3-D image would then be transmitted, as requested, to PCs outfitted with VolumePro to allow real-time manipulation of the processed images.
"This would provide real-time 3-D to people who wouldn't otherwise be able to afford it," Sandy said. "They would not be the 24/7 users. They might just want to be able to do surgical planning when needed, but it wouldn't be so often that they require a dedicated workstation."
This approach promises major benefits, he said. First, raw data take up less space than data processed into slices and 3-D models. This optimizes the use of the archive. Second, the transmission of raw data rather than processed data takes up less bandwidth, which speeds up the network.
"This could provide the ability to get real-time 3-D and 4-D into settings beyond radiology," Sandy said. "It would enable 3-D and 4-D in modalities as well as in PACS."
This vision will take some time to implement. First, the two companies must merge. Then they must go through a strategic planning cycle to understand business requirements, a process that may involve extensive discussions with OEMs. Finally, the merged companies must figure out the best way to combine the two technologies and decide whether they will be joined on a single PCI board or kept separate and plugged individually into the PC.
Similarly sketchy are details surrounding the purchase of RT Viz by TeraRecon. All that's known about the terms of the acquisition is that Mitsubishi Electric will become a minority shareholder in TeraRecon.
TeraRecon is privately held company headquartered in San Mateo, CA, with branch offices in Tokyo and Amsterdam. The company is an OEM supplier of processing and reconstruction solutions for advanced CT and ultrasound scanners. TeraRecon also develops and sells PACS components including workstations and network and archiving products. Chip and board design and software engineering are integral parts of company assets.
RT Viz was formed in 1998 as a business unit of Mitsubishi Electric. It has focused on the development of real-time 3-D imaging solutions for medical, geophysical, scientific, and research applications. The company's VolumePro family supports rapid processing of volumetric data sets (SCAN 6/9/99).
Mitsubishi Electric's divestiture of RT Viz is not a surprise. The business unit was something of a corporate misfit with Mitsubishi, which focuses on information processing, telecommunications, satellite communications, consumer electronics, industrial technology, energy, transportation, and construction.
"Clearly a lot of our chip development could not be used for other purposes," Sandy said. "We just didn't fit neatly in the long-term picture."
RT Viz does fit well with TeraRecon. But the merger's success is far from a sure thing. Cost issues external to the two companies may be difficult. XTrillion processors will be expensive, and their integration with VolumePro will add cost.
How the merged company will get the price of the combined product low enough to be of interest to OEMs has yet to be worked out. That will be one of the challenges when it settles in to develop a business strategy.
"We will look at the upper ends of markets, where performance and image quality are the important aspects of the sale, as opposed to cost," Sandy said.
Study Reaffirms Low Risk for csPCa with Biopsy Omission After Negative Prostate MRI
December 19th 2024In a new study involving nearly 600 biopsy-naïve men, researchers found that only 4 percent of those with negative prostate MRI had clinically significant prostate cancer after three years of active monitoring.
Study Examines Impact of Deep Learning on Fast MRI Protocols for Knee Pain
December 17th 2024Ten-minute and five-minute knee MRI exams with compressed sequences facilitated by deep learning offered nearly equivalent sensitivity and specificity as an 18-minute conventional MRI knee exam, according to research presented recently at the RSNA conference.
Can AI Enhance PET/MRI Assessment for Extraprostatic Tumor Extension in Patients with PCa?
December 17th 2024The use of an adjunctive machine learning model led to 17 and 21 percent improvements in the AUC and sensitivity rate, respectively, for PET/MRI in diagnosing extraprostatic tumor extension in patients with primary prostate cancer.