Computer generated transcript

Warning!
The following transcript was generated automatically from the content and has not been checked or corrected manually.

All right. My name is Patrice Cogwell. I'm a neuroradiologist at Mayo Clinic in Rochester, Minnesota, USA. In this section, we are going to be discussing reducing diagnostic variability in AI imaging interpretation. We will revisit Margaret, who has had a baseline MRI and has been started on anti-amyloid therapy. We will discuss how to perform reproducible imaging, appropriately identify and classify area, and convey an actionable report to the treating team. Multiple brain MRIs are required over the course of amyloid targeting therapy. A brain MRI is needed at baseline prior to the start of therapy. This examination serves to evaluate for findings that may make the patient ineligible for therapy, such as the presence of superficial cirrhosis or multiple microhemorrhages. Once a patient is started on therapy, this exam will also serve as a baseline for future comparison. We also have scheduled MRI's during therapy, that's at least 4 brain MRI's needed during the first year of therapy to evaluate for asymptomatic aria. And if a patient on therapy develops new neurologic symptoms, a brain MRI may be warranted to evaluate for aria, as well as other diagnoses. Across these exam indications and serial imaging in a patient, it is important that a consistent imaging approach is used to facilitate area identification and grading. First, we'll discuss using a consistent imaging protocol that should include the following key sequences, a flare for REE detection, a T2 star gradient echo, and or susceptibility weighted imaging for REAH detection, and diffusion weighted imaging to differentiate acute infarction, which has diffusion restriction from REAE prankyloedema that does not. In addition to a consistent imaging protocol, ideally a patient is imaged on a similar scanner over serial exams. This includes a similar field strength, ideally 3 Tesla, and a similar scan vendor to facilitate longitudinal comparisons. A current topic of interest around RA imaging is how to apply different hemosiderin sensitive sequences. The T2 gradient echo sequence has been used in clinical trials, and is the sequence on which the RA severity scoring systems and management pathways have been developed. Susceptibility weighted imaging is another sequence for heme detection and is more sensitive than the T2 star gradient echo, so may detect more microhemorrhages. And SWI is increasingly available and used in clinical practices. An open question in the field is if different thresholds for treatment eligibility, as well as different area severity scoring scales should be used for these different sequences. Just keeping in mind that other factors such as imaging parameters and field strength also affect our sensitivity for the detection of heme products. It has been assumed that if we apply existing thresholds to susceptibility weighted imaging, this may result in exclusion of more patients and higher RAH severity scoring than if if if the GRE was used, and further evaluation is needed in this area to gain more information and guide our future practice. So the first question in this section is that Margaret has a monitoring MRI that shows 3 punctate fossa of susceptibility on susceptibility rated imaging. The radiologist is not quite sure if these represent a vessel or a micro hemorrhage. They're not confidently seen on the gradient echo sequence from that same exam. And they are not present on the baseline MRI for which only the gradient echo is available. So what would be the best radiologic interpretation of this exam? Apply existing thresholds to the imaging findings on SWI and grade this as moderate RAH. B, ignore these foci susceptibility because of the current thresholds have been based on GRE. C. Report 3 possible new microhemorrhages. D, recommend permanent discontinuation based upon the SWI findings, or E, report no RAH because the patient has no symptoms. Give you a minute to think about your, what do you think is the best response in this case. And here with the most appropriate report would be C, reporting 3 possible new microhemorrhages. So we see that the reader is unsure about the findings on SWI. We'll note that there may be cases in which there are clearly 3 definite microhemorrhages on SWI and they may or may not be seen on GRE and it is appropriate to convey your level of diagnostic certainty. In each case, and in this case it would be their findings are are possible new micro hemorrhages. A, um, is incorrect because this would be mild RAH if there were 3 new microhemorrhages. And many practices are applying our current RA thresholds to susceptibility weighted imaging in their practices, and that's very appropriate, but again, in this case, if we were confident there were 3 new microhemorrhages, that would be mild RAH. B is incorrect as if we're going to acquire susceptibility weighted imaging, we should incorporate that into our reports. If practices would like to guide their management based solely upon gradient echo, that which is done in many cases, um, then only that sequence should be acquired as we should report everything, uh, that is done in the imaging. And D is incorrect as we won't be making treatment recommendations based solely upon the imaging. And here E is incorrect as R is an imaging finding regardless of the symptoms. So to uh find out our discussion on the comparison of GRE versus SWI and some of the considerations in our practices, and we must also um highlight the importance of using a consistent imaging approach for the detection of heme products. So that is choose SWI or GRE or both, whichever is appropriate for your practice and use that consistently. As changes in the imaging sequence of SWI versus GRE as well as other factors such as field strength, may change our sensitivity for microhemorrhage detection and alter apparent findings. As if there is a change in imaging technique between exams, such as from gradient echo to on the next time point susceptibility weighted imaging, new microhemorrhages may be seen and the radiologist may be unable to discern if there's been a change, new microhemorrhage, or if the findings are just better seen due to differences in technique, and it would be appropriate. To state that if that is the case. We've been discussing RH imaging, as we said, detected on GRE or SWI sequences, and REAH occurs in the form of microhemorrhages, also superficial cirrhosis, and rarely macro hemorrhage. There's also REE which we said is detected on the flare sequence that occurs in the form of vasogenic edema or sulcal effusion. And while we're discussing these as distinct entities, let's note that RAE and RAH can co-occur, and they may be identified at the same imaging time point, or in some cases, the appearance of RAH may lag that of RA E in that if RAH may be identified 1 to 1 to 2 month follow up of uh after RA E. When RA is identified, it should be given a radiographic severity score. This is a score of mild, moderate, or severe RAE, RAH microhemorrhages, and RAH superficial cirrhosis. RE is graded based upon the number of contiguous regions of flare abnormality and the greatest linear dimension of the largest region. RAH microhemorrhages is graded based upon the cumulative number of treatment emergent microhemorrhages and RAH cirrhosis by the number of regions of superficial cirrhosis. So our next question, Margaret has a follow-up MRI shows a new sulcal effusion measuring 3 centimeters. There's also two new microhemorrhages. Margaret also reports a new headache. Which is the best radiologic interpretation of the imaging findings? A, an abnormal MRI treatment toxicity not excluded. B, new RE and RH recommend clinical correlation.