You are here: >home >funstuff
Water Cooler Conversations
May 2019: Google search for "JMRI" returns "A Java Model Railroad Interface" website. Journal of MRI shows up on page 4. At the first glance, seems odd that model railroading floats to the top. But, it is probably a good thing that humanity is more interested in model railroading than MRI. It will take a long time before "comfort" of MR tube can compete with people's preference to enjoy their hobbies.
Are innovation and disruptive technology always good in health care?
Many years ago (about 10-15), medschool and residency mentors taught that listening to patients and performing a thoughtful physical exam will solve majority of health mysteries. It seemed to work. Even now, just looking at the age of the patient and reading a decent clinical history line, bring top two-three differential considerations to mind before the images start showing up on the screen.
As residency progressed, it seemed that less and less people were taught listening and exam skills, and more and more had "in-service" on how to order tests in the computer. We now live in a country where CT scan can be obtained during any hour of the day. And, sometimes before even seeing a physician. At many places MRIs are done at 3am routinely. Is it helpful to have all that fancy technology available?
Susceptibility weighted imaging (SWI) is a fairly new technique that gained traction in clinical practice recently. SWI can detect ridiculously tiny amounts of blood not detectable by other routine MR sequences. Some machines are even able to differentiate between tiny calcium deposits and tiny haemorrhagic products. The following question came to mind recently: Does finding a tiny hemorrhage in ischemic stroke change anything clinically? Current articles vary a bit on clinical recommendations. Some advocate adjusting thrombolytics lower to decrease the risk of large hemorrhage, if micro hemorrhages are discovered. Others advocate continuing current clinical practice until more research is done. One went as far as saying to extend the thrombolytic therapy time window in patients who show microhemorrhages. Confusing.
Some centers do not have access to MR scanner at all, others are able to find microscopic hemorrhages at 3am. Many stroke algorithms do not even mention an MRI. Yet, we are now in a situation where a random MR scanner can find a tiny hemorrhage in ischemic stroke that would have never been found five-six years ago. The practice of thrombolysis in carefully selected patients based on clinical exam and negative head CT scan has been established and studied for years. We have yet to see which one of the above recommendations is the most beneficial in the context of microscopic hemorrhages and to which patient population.
It is ironic that new thrombolytic drug has to go through numerous stages of testing before FDA approves it. Yet, a sequence on MR scanner, or use or not use of such sequence only needs to be practical and safe from MR safety perspective. Yes, there are regulations related to technology, but when was the last time you heard of "phase 3 trial" on new gradient sequence in 2000 patients? Yet, in practical situations use or non use of such sequence may result in withholding or administration of a drug based on interpretation of results. Similar situation exists in oncological imaging where current technology detects tiny metastatic disease. Once again, the studies that resulted in current recommendations were often based on imaging tests obtained with decades old equipment.
Is more regulation the answer to this problem? It is doubtful. More education and research would help a lot. Access to simple "help" files on what the sequence does, what are its limitations and what are its benefits would solve many clinician's and radiologists's dilemmas. Perhaps, one regulation could be useful: if manufacturer adds a new sequence to its scanner, "executive summary" about it on its website should be available. Until then, it's "Dr. Google," UpToDate and Pubmed consult.
These images were created using ray tracing program "POV-Ray." This program has been around for several years. The idea of creating 3D scenes with POV-Ray is little bit different from conventional 3D modelling. In POV-Ray, one needs to set up every scene from scratch using specific geometric instructions. It is a little bit of a puzzle to figure out how a ray will interact with any given set of instructions, so there is defintely an element of high-tech art on all of this. People hold competitions with different themes. The images above are couple of amateur adventures with something called caustics...