When you get stuck during research?

Question

HI. I'm doing research right now using MRI. The problem is I'm stuck. I'm very very stuck and don't know how to continue. I'm not managing to do the same as the articles I've read. And yes I have e-mailed the authours but get no response. I've emailed tons of people asking for help. But I remain very much stuck. I need to do something (called deconvolution) in order to continue. I just don't know how to do it and go further. What do you do when you get stuck in research?

Answer 1

Well, I don't know much about MRI but I can tell you about deconvolution theorem:

Deconvolution is nothing more than the reverse process of convolution.

Here's a quote from Wikipedia about convolution:

"In mathematics and, in particular, functional analysis, convolution is a mathematical operator which takes two functions f and g and produces a third function that in a sense represents the amount of overlap between f and a reversed and translated version of g. A convolution is a kind of very general moving average, as one can see by taking one of the functions to be an indicator function of an interval."

Apparantly, convolution can be used to encode information about detecting the edges (parts of an image, for example, where there is a step in the contrast).

Basically, If we have a binary image (lets say its a 2D image of a jar of pickles, where the jar (and everything in it) is black and everything else is white) you could encode that information as a bitmap. That would be very costly for memory and time, so instead you can assign two functions (one to describe the white and one to describe the black) then perform a convolution function on them to produce a third function, which contains information about the boundaries.

Now, the reverse convolution (deconvolution) does the opposite; it takes that information from the third function and transforms it back into the image data.

The formula and theory are a bit to involved to type in here (including integrals, which there isn't a sensible way of representing with ASCII text) so I'll just give you the link to the page in the sources.

This is a very good page and should help you a lot.

Answer 2

found that in my own research, it helps to take a break away from the problem for a day, do something completely unrelated and then come back to it. I also try to remember that just because the answer I,m getting is not the same does not mean that it is wrong but merely that different variable will influence an experiment. any data is good data.

Answer 3

Ask colleagues, technicians, experts in the field. If you have the resources (funding), you should be able to reproduce anything someone else did. You might have to shell out for the expert help though.

Sometimes (especially for an undergrad or a young grad student), you'll just find that you're in over your head and have to find a more tractible problem. Ask someone in the department whom you respect for guidance, not just on how to solve the specific problem but on the big picture of what you're trying to do.

Answer 4

I walk away and do something else. then I come back to it later...