image analysis of brain image

i have an PSOCT data, and I want to analize the depth profiles for reflectivity and retardance based on the en-face images, and compare between the white matter contrasts for different solutions used.

[Walter Roberson]

What is your MATLAB question?

[puma]

1) How to average reflectivity signals in dB scale by fit with a least-square first-order polynomial function.

and how to obtain attenuation from the above.

[Walter Roberson]

polyfit() is the easiest way to get a least squared polynomial for numeric input.

[puma]

i have an image which is in x and y axis, and for the same image i have x and z axis. from these two images is there a way i can obtain the depth based on the pixels, by taking an average of 10 axis lines. ?? how can i do it in matlab. at the end i wanted to see the depth at a perticular point and compare the results.

[Walter Roberson]

No, you cannot determine depth from two perpendicular views.

[Walter Roberson]

[Bjorn Gustavsson]

@puma: What @Walter Roberson illustrates with this image is that one image is not enough - unless you know a fair bit more, like for example also the shadows of the objects illuminated from 3 different spot-lights on three different planes - which effectively gives us something comparable to four individual images of the objects.

[puma]

Thank you, i will look into it.

[Walter Roberson]

Suppose you have a near rectangle

+------------------+

| |

| A************ |

y | B************ |

| C |

| |

+------------------+

x

Now you want to know what the depth is at the C .

According to your hypothesis, you also have an x z image giving depth. So we look in the x z image for x = 4 and you get back a vector of numbers. But... the vector is in the z direction, and includes no information about the y dimension. So you cannot tell the difference between depth for A, B, and C: they all have the same x and the x-z image has no y information. It is like looking down on the image from the top.

Now, if you had two images taken at different angles that were not a multiple of 90 degrees apart, and you had some calibration information, then there might be the possibility of using perspective calculations to estimate depth.

[puma]

Now i have a video of around 800 scans of an image. and they merge together to form a 2d image.

can i obtain the depth of a perticular nerve. from the video image? and plot it ?

[Walter Roberson]

@puma

Is it possible that you have a video of around 800 scans, each at a different Z coordinate, and they merge together to form a something by something by 800 volume?

[Bjorn Gustavsson]

@puma - Walter is still guessing what you have. For anyone to be able to suggest solutions you need to describe what you have. In detail. PSOCT - does that stand for polarization sensitive optical coherence tomography? If that's the case what data, does the video contain? Does the images give information from varying points around a scanning-sweep of the camera? How does the polarization cause images to change? by what means does the coherence influence the image intensities? That's 5 questions that will determine what suggestions might help you solve your problem.

[puma]

yes PSOCT is polarization sensitive optical coherence tomography. the data is a scan of brain cell(cerebellum part) which is imaged from the top to see the deeper regions, like the white matter and gray matter of the brain. and i have to focus on obtaining how there is change in depth in white matter, for that we used two different solutions(pbs and glycerol) and imaged the cell. the machine scans the brain in all directions and give the final image.

my video has the reflectivity retardance and cross polarization en face images. which scans like an x ray.

the shown video, is for one solution. the left side images are the scaning and how each scan joins together to form the right side images.

for example i have to take tile 42, (pause it in the video, title at top). and stop. and that left side white spikes shows the white matter. i have to see its depth, and compare it with a different video (same kind of image). and see which solutions lets me knows how deep and clear i can see.

given bellow is the link.

https://drive.google.com/file/d/1s_532txf5YaN3K5x7LVGgpDgSoQ3Fxjg/view?usp=sharing

Thank you.