Thursday, December 5, 2013

RSNA 2013: My top 10 on what’s new and what’s old, part 1.

It is a brisk morning, typical for November, and I am waiting for the bus to get me to the annual radiology circus: RSNA 2013 in McCormick Place conference center in Chicago. And a hustle and bustle it is indeed, especially around 10 am when the exhibition starts, and at 5 pm when it ends, and between scientific sessions when everyone is running to the next classroom to attend a presentation on the next latest and greatest technological innovation, or a lecture in order to get their CME credits.

Need my caffeine fix, even if it takes
20 minutes
Having a technical background, I can only comment on technological innovations and the curiosities at the meeting. These observations are obviously totally subjective and are offered as only a partial impression of what was shown. My findings are based on the OBWA method (Observations By Walking Around), i.e. seeing if there is something interesting that catches my eye and by talking with the many friends and colleagues and ex-colleagues, in the hallways, corridors and at the bar in the evening. This list is unlike the other RSNA news reports that are biased by published press releases or interviews from the industry “experts” with the inherent hype and vaporware.

I have split my observations into three parts, 1) what’s new, 2) what’s not so new (“new and old”) and 3) what’s old. So here are my first top 10 new ones:

A typical dual energy CT image,
showing the overlay with the 2nd image

  1. Multi-energy CT’s: Some of you might remember the very first generation of rotational CT’s, where a gantry containing the X-ray tube and detector were located in a rotating ring and images were taken slice by slice by rotating the gantry, and alternating clockwise and counterclockwise around the patient. Scan times could be 15-30 minutes with an additional 15 minutes or so doing the numbers crunching to create a set of maybe 50 images. Since then vendors introduced slip-ring technology to have a continuous rotation, multi-slice detectors to provide orthogonal voxels providing the source for excellent 3-D imaging, high-speed rotation of fractions of a second to allow for dynamic studies such as used for cardiac applications. Every time one might think the technology has matured, yet another innovation comes along to provide a complete new dimension and/or paradigm, such is the case with the CT multi-energy imaging capabilities. By processing two images that are acquired using a different energy spectrum, which can be done by using multiple X-ray tubes, switching voltage or, in the case Philips, multiple detectors, each with a different energy absorption characteristic, one can create different images that can provide more information than just the traditional attenuation information expressed in Hounsfield units. The images looked to me like fused PET/CT images however, the color images contain much different information using atomic numbers, which we will probably have to learn to interpret similar to when we saw the very first MR images. The CT images were created on a CT that is pending FDA approval so there is not a lot of experience available yet but who knows, this technology might become standard over the next few years, similar to the multi-slice capability of the recent CT’s.
  2. Micro-dose Mammo: There is also no disagreement that too much X-ray radiation can increase
    New mammo slit scanner technology
    the risk of cancer just as other factors such as nutrition, lifestyle, and genetic disposition can, and therefore any imaging solution that reduces the dose for digital mammography can only be applauded. With annual screening for women from let’s say age 50 to their average age in the US of 81 years resulting in a total of 31 exams, typically with two views or images taken for each breast. The good news is that the new Philips microdose digital mammo system uses a slit-scanner technology, which means that the tube rotates at an angle to scan the detector plate thereby reducing the dose by 40% according to the vendor. There are a few hundred of these systems installed in Europe, especially in Scandinavia and France, and installations in USA are just starting. Hopefully, this will challenge other vendors to either adopt this technology or rethink their implementations to achieve a similar reduction in dose.
  3. Social media for radiology: The use of facebook® for posting images is regarded by many as a
    novelty, something many, including myself, would never consider until a good friend of mine shared with me the fracture of his wife’s leg, which he had posted on his facebook page after she had a traffic accident. For those who have been ignoring facebook, I suggest you look a the “Radiology Signs” facebook page which allows people to post interesting cases, which has more than 400,000 “likes” as of today (are there even that many radiologists connected to facebook?). That is why at least one vendor is implementing a “facebook sharing” option in their PACS viewing software, obviously after making sure that the image is totally de-identified and stripped from any personal information. I believe that the power of social media cannot be underestimated and it might become not only a great education tool but also a forum for interaction and communication.
  4. Radiology-patient partnerships: The theme of RSNA President Dr. Sarah Donaldson’s address
    View from the top on Sunday, which
    was actually busier than usual
    was about partnership. I have seen partnerships between radiologists and other physicians in several institutions, notably those where physicians are on staff and paid by the hospital, which seems to avoid a lot of turf wars and breaks down silos between the different specialties. However, partnerships between radiologists and patients are a new concept as it is mostly the primary physician who gets a copy of the report and reviews it with the patient. As a matter of fact, unless a patient takes the effort to look at the person who signed the radiology report, he or she is almost never aware of who did the interpretation anyway. Therefore, even though it might be a good idea, I doubt that a radiologist would even return my phone call if I tried to call him or her about a diagnosis. I am fortunate in that I know several radiologists personally and if I want to ask a question about my or one of my friend’s or family member’s radiology exams, I send them a CD and ask for their (second) opinion. However, that is not an option for most people. Therefore, in my opinion, these are nice catchy phrases and make good headlines, but there has to be a major culture shift to make patient partnerships in radiology happen, if it ever will.
  5. NoSQL: Most people might not know what technology is behind managing those millions of
    images in a PACS or enterprise storage solution, but it is typically based on a relational database such as Oracle, Sybase, MySQL or other commercial or open source product. Databases were not designed with managing patient information in mind, except for the somewhat ancient MUMPS, which is both a language and a database, is the core of many of the popular EMR’s but is not typically used in PACS system databases. In addition, commercial database licenses are not cheap and therefore impact the system cost significantly. Therefore it is no wonder that vendors are looking for alternatives.
    Voila! the NoSQL, which stands for Not Only SQL, which indicates that they can still be accessed using SQL queries but also allow for other access methods. NoSQL databases were invented in the late 90’s and are very scalable and highly optimized for simple retrieval and updating operations such as used for medical applications. The nice thing about designing a product from scratch, which as an example, Karos did with their new VNA implementation, is that it allows you to use the latest technologies instead of porting or converting it from older technologies. NoSQL might become a good alternative to the commercial databases that are not as suitable and overkill for what a PACS or enterprise archive solutions are trying to accomplish.
  6. Analytics: Medicine is probably one of the disciplines that is the least measured and analyzed
    One of the many analytics companies
    with regard to efficiency and cost. Most institutions don’t have a good handle on how much it costs to perform a diagnostic procedure other than the amount of the reimbursement from an insurance company or medicare/medicaid. In order to analyze information we need to measure its input and make sure it is correct and accurate. 
    There are several companies offering analytics, most of them I had never heard of before, which showed information displayed on very nice dashboards, however, when talking with them I found that there is still a lot of missing and “dirty” information out there. There is also a lack of standardization of workflow and terminology, which SIIM is working on as part of their SWIM (SIIM Workflow Initiative in Medicine) project. Together with consistent terminology there needs to be a consistent implementation as well. For example, DICOM header information might contain one or all of the Attributes: Study time, Series Time, Acquisition time, and Content time. Some PACS systems even add a timestamp when an image is received by the PACS archive. Consequently, which time do you use to define a procedure length is at best a guess, especially if you want to compare modalities from different vendors who are using different Attributes in their image header. Similar problems occur when you want to record the report turn-around time: is the end time defined by the time that the radiologist signs the report electronically (assuming this is available), is it the time when it is sent to a report repository such as at the RIS, or when it appears in an EMR, it is faxed to a physician, or when it appears in his (secure) email inbox? In conclusion, there is going to be a major increase in analytics but we will have to do a lot of standardization of terminology and measurement, as well as data clean-up before we can trust the results of these tools.
  7. Compact CR: CR systems are getting more and more compact. The first generation CR I ever
    Couldn't be more compact
    encountered was a FUJI CR and it occupied a small room. That system included a printer as well as the CR technology and was ahead of the PACS infrastructure, which followed within the next 10 years. The first challenge was to make a CR small enough that it would fit on a table as a “tabletop” system. Having achieved that over the past 10 years, the latest technology allows it to be so small that it barely sticks out from the wall. The affordability also has come down so much that small practices, and even dentists, veterinarians, and chiropractors are now considering digital technologies instead of film. These are also great solutions for emerging and developing countries where there is an installed base of film X-ray equipment, which is a barrier to providing healthcare by itself as many can’t afford the film and associated developing costs. There will still be a need for larger high volume CR systems, but I would think that these small CR systems will become as ubiquitous as the many small copiers you find in offices in addition to a high speed copier in a central mailing room of an office.
  8. Wireless badges: When I used to work regularly with X-ray systems and/or visit X-ray
    Wireless X-ray badges
    departments, I would always carry my X-ray radiation badge with me. I remember that I occasionally forgot to take it out of my carry-on luggage at the airport, which caused a call at the end of the month from the radiation safety officer at my company questioning why the readout of the badge was much higher than normal. New badges are getting more sophisticated as they now can measure this information and send it wirelessly to a repository. This replaces the old collection and distribution system making it much easier and convenient, also allowing semi real-time monitoring.
  9. Less floor space: What was different this year was the reduction in exhibition floor space, which was somewhat of a mixed blessing. The smaller exhibit space was very welcome to many of those who in the past had to cross from the North-South location to the lakeside area exhibition halls. There were only two exhibit halls this year as vendors brought in significantly less “iron,” for example instead of a complete CT or MR gantry they would bring a
    Scale models instead of
    the real thing
    scale model and/or a monitor showing its images. Obviously the RSNA organization itself would have liked to rent out more space, but also if I were to buy a million dollar or more piece of diagnostic equipment, I might want to touch and feel it, similar to wanting to kick the tires when buying a car. Consequently, this year, if I had wanted to see the equipment prior to signing on the bottom line, I would have to travel to another facility to see it installed, or visit the manufacturer’s facility. I understand the cost savings to the vendors not having to haul these systems around, but I would feel somewhat cheated as a potential customer, especially after paying hundreds of dollars to be able to enter a tradeshow like RSNA.
  10. More lines: It seemed to me as if sequestration or budget cuts had an impact on the RSNA this year, as the waiting lines were much longer than it appeared to me in the past. Waiting for half an hour during the TSA check-in at the airport was to be expected for a holiday weekend, as well as 20
    This is a typical 35 minutes wait
    minutes to get a cup of coffee at the exhibition hall Starbucks, but having to stand in line for more than 30 minutes at 10 am in the morning just to pick up a badge and wait even longer for the bus getting back to the hotel at 5 pm while being exposed to CO2 exhaust fumes in the catacombs of the conference center was unexpected and in my opinion due to poor organization and customer service. I’ll definitely arrive earlier and depart later next year. Hopefully there will be more registration contractors and buses next year as well.
These were my observations about what’s new, and to be honest, there was not much earth shaking and or very innovative this year, as much technology has matured, which is why I will have a follow up on “what’s new and old” as well as “what old news,” in parts 2 and 3 of this report on RSNA 2013 (stay posted).

Herman O.