Ultrasound in Radiology - is there a future?

Throughout the world there is an increasing demand for ultrasound (US) exams. This is true also for US in radiology. However, there is no evidence that ultrasound is relieving the burdens of CT or MRI - on the contrary, there is an ever increasing demand also for the services provided by those modalities. It seems that ultrasound is merely being added to the plethora of other exams rather than being a true and dependable cornerstone within radiology.

Anybody working in an average radiology department is well aware of the fact that there has been a remarkable development in their CT and MRI departments in the past ten years. But what has happened to their US departments? Apart from better US scanners, don't the procedures and indications look very similar to those back in the early 90's, except for the fact that CT has taken over virtually all cases of acute abdomen as well as most planned examinations where the diagnostic results are really important? Has the overwhelming impact of PACS systems in radiology had any substantial impact on the documentation of ultrasound exams, or are we all still depending on the written report as concluded bedside? Have we reached the possibility to draw new conclusions from the images of an old US exam? Unfortunately most radiologists will find the answers to those questions discouraging on behalf of US, to say the least.

In my view there will be an inevitable downward spiral for advanced US in radiology in the overwhelming competition against other PACS-dependent modalities, if the US community does not acknowledge that drastic measures are necessary in order to reduce the gap between the modern PACS image documentation of CT and MRI on the one hand, and the prevailing limited still image documentation of US on the other. Clinicians become increasingly accustomed to the conspicuous images that are produced by CT and MRI, and will generally not continue to settle for the non-documentation of US in cases where imaging really matter. In hospitals where there is a fruitful cooperation between US and the clinicians, the sonologists are personally well known by the clinicians. This trust in a few sonologists is of course a very fragile basis for enduring and dependable US services. And regardless of the skills of the sonologist, his judgement on a piece of paper will never outweigh the ruthless truth of retrievable CT or MRI images that say something else. And his still images are no good for comparison when the patient comes back for a check-up, since new pathology can never be proven not to have been there on the first occasion.

Is there no way that US can approach the dependability and consistency of other modalities in radiology? Can US retrieve lost turf, and again become a cornerstone within radiology? Chances are that it can, but it will require a whole new way of thinking by radiologists and all other involved. Without learning from radiology's built-in quality assurance mechanisms and applying them to US, it will never happen. We simply can not afford to have top notch diagnosticians at each transducer for the large bulk of US exams in order to find rare pathology. We need to be able to capture exams in a way that they can be read and diagnosed by experts at workstations, even if the examiners are not fully trained diagnosticians but good scanners who capture pre-defined cine loops of the regions of interest.

What gives me the right to make such statements? Being a radiologist myself I dare expose these ideas, but they would be worthless without clinical experience of an alternative way of organizing a US department. Being head of the US section of the Radiology Department, Linköping University Hospital, Sweden, I and my colleagues reorganized ultrasound completely in June 2002. We installed a US dedicated PACS system and abandoned still images completely. Instead, we replace still images with 5-10 seconds long cineloops. We developed organ- and structure-specific standardized scanning patterns, capturing the entire organs or targets of interest on such cineloops, scanning at a steady pace in one direction. All involved - senior sonologists, radiologists and residents - began scanning all exams according to such standardized patterns, so that the organs were displayed equally on the US workstations regardless of who performed the exams. The US workflow essentially became equivalent to that of the rest of the radiology department, with capturing of full organ volumes, workstation readability, double checks and finally storage in PACS for later retrieval. Following a short period of time during which everyone got accustomed to this new practice, we found out that it was much easier to discuss pathology at the workstations when the scanning patterns were obvious and natural to all. For the first time it was possible to compare old and new exams side by side and actually show to us and the clinicians that there was new pathology in areas that were normal on the first exam. Our younger colleagues experienced a steeper learning curve, since they had the possibility to see the entire production of the US section on our workstations, all scanned in a way they were familiar with. They were very pleased with the competent feedback they could be given at the workstations. And, as a routine, all out of hours work was re-read by a senior sonologist the following regular working day for feedback and corrections.

Now, almost five years and more than 30.000 exams later, I dare say that the outcome of this reorganization in the direction toward a "radiological workflow" has exceeded all expectations. Now we know for certain that all examiners, regardless of skills, are biased by various parameters bedside, and often oversee information that is actually captured in the cine loops. The dependability of the reports has become better since the examiner can reassess his exam at a workstation while reporting, thus finding details that were overlooked bedside. It is also quite easy to discuss details of our exams when we want a second opinion. No reports by residents leave the US section without previous reading of the examination by a sonologist at the workstation. The accuracy and security of the standardized scanning patterns reached such a level that we trained two radiographers to become "sonotechnicians", who learned how to scan the bulk of average US exams for later reporting by the sonologists at the workstations. Despite the fact that the sonotechnicians are not pathology trained, their exam technique is excellent and their exams are presented on the monitors just like those of the experienced sonologist. Their exams are read at an average rate of 12 per hour by the sonologists, which proves that the technique is very time efficient for the sonologists.

However, of course there also are advanced US exams that require the skills of a sonologist or sonographer bedside for the understanding of the pathology, but documentation according to the basic principles of standardized cine loops, and reviewing them at the workstation, provides an increased understanding of the findings in many cases. This is especially true for structures with a complex anatomy, which can be better understood by slowly reviewing the scans back and forth at the workstation before the final decision. To date there has been no 3D- or 4D technique presented that allows for volume- or multiplanar rendering of such scans without loss of subtle parenchymal pathology, but considering the fast technical development, such volume rendering may soon be a helpful adjunct to the basic scans.

The conclusion drawn by our clinicians as well as ourselves is that our procedures including standardized cineloop scanning, workstation reading and PACS storage - which we call "Sonodynamics" - has a very positive continuous impact on the reliability and consistency of the US section's services. We feel that Sonodynamics is unparalleled when it comes to US education and training. A large percentage of advanced diagnostic challenges are resolved without the use of other modalities, and the employment of sonotechnicians provides time efficient workstation reading of the bulk of common US requests. I am sure that US in radiology would benefit very much if all of us in the US community would reconsider old habits and unite to work for standardization and full documentation. If we don't, I am afraid it will not be long before US looses all examinations of importance to other radiological modalities.

For further information on the background of Sonodynamics, and for details regarding our standardized Sonoexams, you are most welcome to visit the Sonodynamics website www.sonodynamics.com