Advanced visualization has come a long way since its start as a tool for specialized studies such as cardiac CT and CT colonography(CTC). Today, thanks to the decentralization of visualization software from a thick client to a thin client together with the prevalence of tools on PACS workstations, advanced visualization (AV) is playing an increasingly important role in medical imaging. This trend toward accessing AV applications from a centralized server is becoming even more important as many radiologists continue to report remotely.
As thin-client AV technology has evolved, it has become more efficient in enabling remote access, said Frank Rybicki, MD, PhD, Director of the Applied Sciences Laboratory, Brigham and Women’s Hospital, in Boston. However, not all issues with thin clients have been completely resolved, he added, noting that compatibility with PACS and sites with limited licenses still needs to be better addressed. “At many institutions, there are only a certain number of licenses, so there are challenges with having the advanced visualization on all thePACS workstations,” Dr. Rybicki explained.
Certainly, radiologists are not fond of moving from one workstation to another for a specialized analytic task that is viewed as non-critical. This has helped to give rise to the prevalence of thin-client technology, said Michael H. Rosenthal, MD, PhD, radiologist at Dana Farber Cancer Institute and Harvard Medical School, in Boston. “The thin client on PACS provides immediate, low-resistance access to the tools.On the downside, PACS stations may not have the configuration or horsepower necessary to perform complex analysis tasks on the local client, so processing may need to be performed on a central server. This can reduce the sense of interactivity of the thin client.”
While the movement of AV to a thin client, or server-based environment, has enabled enterprise-wide access, the next integration hurdle will be to integrate imaging with EHRs for access at the integrated delivery network (IDN) and health information exchange (HIE) level,said Julie Kemp, Vice President of Global Marketing at Vital Images (Minnetonka, MN). The goal is to enable referring physicians to have access to 3D rendered images in their offices or exam rooms to share with patients.
“The trend is to go cloud-based with zero footprint capabilities,” said Kemp. “The challenge with a zero footprint is providing enough memory in the cache to enable quick access to images. This scenario becomes even more challenging with sites that do not have a vendor neutral archive or a PACS.”
Putting advanced imaging tools on a zero footprint viewer also enables their use in emergency or urgent care situations, said Sandra Stapleton, Executive and Director, Advanced Visualization, Claron Technologies (Toronto, Ontario). “Historically, post-processing tools have been used to review the results later. Now, there is a growing need for rapid assessment in the ER, for example in stroke cases. ER clinicians
want access to these tools in their environment.” In addition to accessing these tools virtually anywhere with thin client or zero footprint viewers, the industry is also trending toward a greater
array of visualization tools, particularly for modalities other than CT such as MR and ultrasound, Stapleton said. Whether these tools are for stroke, coronary plaque assessment, or contrast enhancement, they are helping radiologists derive even more information from advanced medical imaging.
“As we move toward precision medicine, individualized tools need to be more precise,” said Dr. Rybicki. “Radiologists need to be as precise as possible clinically, whether they are using these tools to establish a diagnosis or communicate findings with other clinicians. It’s implicit that if they can get more information from the image, they can make a better diagnosis.”
The tools are evolving with respect to the clinical exams that are being performed. One important area is image registration and fusion,added Dr. Rybicki. “Patients are often having multiple different studies. While it is still challenging to fuse images on a workstation, ther eare methods to do it. The value is that you can find something on one (type of) image, or see where it is at over time,” he explained. Even fusing two different CT studies taken at different times can provide new information for identifying or following a lesion.
Another area of development is in quantification tools. “Doctors want to quantify trends over time, and correlate that with other information about the patient,” Stapleton said. In oncology, this would include measurements on the tumor size; in cardiology it includes vessel analysis tools such as quantifying plaque or determining blood flow in areas of thrombolysis or the narrowing of vessels.
From a research perspective, Dr. Rosenthal said he sees a continued need for reliable, automated or semi-automated volumetric analysis.“Clinically, tools that simplify the task of measuring and tracking findings and indexing management standards should be in demand. “
While he said Carestream’s lesion management software performs particularly well on certain types of lesions—liver or lung metastases—Dr. Rosenthal has also found that he can use it very efficiently for volumetric research.“On the research front, volumetric measurement is the hot topic. While RSNA’s Quantitative Imaging Biomarkers Alliance (QIBA) is working hard on the validation of volume as a metric, it will not be broadly adopted until we have volume tools that can be used by a general radiologist,” Dr. Rosenthal said. He also does not believe that manual tracing will be viable in a general practice environment regardless of the results of the volume validation studies.
Dr. Rosenthal predicts the continued development of accessible, user-friendly approaches to lesion segmentation will further increase a radiologists’ use of AV. “When a general radiologist feels comfortable clicking on a lesion, getting a good result, and having it flow into the report, then advanced visualization techniques will have an opportunity for widespread adoption,” he added.
With widespread implementation, Dr. Rybicki said he envisions AV being able to interact with big data. “Typically, these tools work with small data sets,” he said. “The ability for these solutions to interact with a large number of patients is currently an unmet need.” Yet, before this can happen more clinicians will need to have AV tools available to them, he adds.
Stapleton also sees the need to develop algorithms with more intelligence so that in the future there can be a greater utilization of datamining. Claron has started to incorporate more artificial intelligence in its algorithms, particularly tailoring specific advanced algorithms fora specific organ of interest. The company is developing a suite of segmentation tools that quickly identify the organ or area of interest, identify structures within the organ, and then look for specific patterns based on that organ.
Kemp sees analytics as another key area for continued product development, particularly in the areas of utilization, dose and protocol management. “There is an increased need for dose management analytics, particularly with the requirements that will be in place in 2016,”Kemp explained, referring to Medicare’s announcement that it will be reimbursing 5% less in 2016 for CT scans acquired on systems that do not meet Standard XR-29-2013, or MITA Smart Dose, published in April, 2013. That figure jumps to 15% in 2017.
“In this area, our top priority is providing tools to filter grainy images to allow for very clear images regardless of dose,” Kemp said. Older scanners often do not have the same dose reduction tools as newer scanners, so there is an important role for AV tools in these instances.
Along with dose management analytics, protocol management is needed. Kemp explained that as protocols are defined for certain diseases in patients—pediatrics or adult—and certain body parts, they will need to be managed to help ensure consistency across scanner utilization.
Another current gap in the industry is the lack of tools for neuro-related diseases, such as Alzheimer’s and Parkinson’s, said Kemp. While it’s too early for Kemp to share details, she said this is an area of future interest at Vital Images. “If we can get an earlier diagnosis, by helping clinicians see a pattern that is consistent with early Alzheimer’s or Parkinson’s, then the patient can get on a treatment plan earlier in the disease process,” Kemp said.
The future of AV is clearly one of greater accessibility via thin client or browser-based solutions for widespread adoption, additional quantification and analytic tools, and the ability to interact with big data. However, there is one trend that some may find surprising: 3D printing.
While there is some disagreement—Dr. Rosenthal questions the role 3D printing will play in imaging despite the intuitive nature of visualizing a physical object—Dr. Rybicki is confident 3D printing will gain the interest of radiologists and other clinicians.
“Three dimensional printing is the future of advanced visualization,” said Dr. Rybicki. This doesn’t mean that all 3D images will be printed, he said, but rather there is no question that subsets of images are now being printed, and those subsets will grow over time and as data on the subject continues to expand.Back To Top
Technology Trends — Advanced visualization: The future is bright. Appl Radiol.