What is our stance on questions on medical imaging algorithms? Are they on-topic or off-topic?

Example: Automatic Quality Control of MRI

Is there any standard process to assess the quality of MRI images? The goal is to automatically detect really poor quality data (and then manually confirm the defects).

I found some studies about automatic quality control of MRI:

Yet, I am looking for an easy to implement solution (taking the raw - unprocessed - images as input), or an existing tool available for researchers.

Is there a state of the art algorithm or tool for this purpose?


One of the topics listed as on-topic here is:

diagnostic and prognostic methods

So I'd say that's a yes, technical questions about medical testing are on topic. Do I think questions like that will ever get a useful answer? Nope, sure don't. Anybody who really knows how to answer the question is probably working for private industry and under an NDA. But I can't think of any reason why the question itself should be off topic.

  • I disagree with "Anybody who really knows how to answer the question is probably working for private industry and under an NDA." - it's likely the answer would come from the academic research community. – Bryan Krause Jan 13 '20 at 23:37
  • @BryanKrause I dunno.... I can see MRI manufacturers investing a lot of money in this sort of thing but not a lot of academic grants. For example, the company I work for manufactures lab instruments such as liquid chromatography equipment, mass spectrometers, and so on. The majority of innovations in that arena come out of industry, not academia. – Carey Gregory Jan 14 '20 at 0:05
  • Maybe, but this is the sort of thing that is particularly useful with research data sets, where someone is automatically extracting some quantitative signal and the main time-consuming part is artifact rejection. The customers for MRI manufacturers are mostly diagnostic labs. The equipment might be used for research, too, but the researchers aren't the primary customer. – Bryan Krause Jan 14 '20 at 0:10

I don't think that it is really on topic. Using the MRI as a diagnostic tool is definitely on topic as Carey Gregory shows in his links from the on topic reasons, but how to program the MRI to produce imaging I don't think should be on topic.

The site is intended for clinicians, allied health, students of the same. None of them are interested in programming the MRI machine, just that it works.

  • Note: This is a dissenting opinion to provide choices, I don't really feel strongly one way or the other. I don't think it is on topic, but I could see a case being made for it. – JohnP Jan 13 '20 at 14:07
  • I know this is just meant as a dissenting opinion, but I'll still quibble with it a bit. The question isn't about programming an MRI machine, it's about post-processing. A clinician, medical student, or anyone else involved in research using MRI images would likely find use in this question and answer. – Bryan Krause Jan 13 '20 at 23:39

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .