The channels of MRI images

Hi everyone,

In AI4M Course 1 Week 3, there are MRI images with the 4 dimensional shape, let say (240, 240, 150, 4) for example. I have my own thoughts but I’m not sure they’re correct:

  • Are 240 and 240 here the width and the height of the image just like a 3D image?
  • I understand 150 layers here as the image was taken 150 times as it went deeper each time. Is that true?
  • And the final number 4, the channel, I think there are 4 different views, right? Like, the first view is the front view, then the rear view and then the 2 side views? (This seems wrong since I browsed the net, there are also 8, 16, 32 channels MRI images).

Thank you.

1 Like

It is mentioned in the the AI4M Course 1 Week 3, explore ML data that The image object has the following dimensions: height: {height}, width:{width}, depth:{depth}, channels:{channels}"

I had the same question. What do channels mean? Can you please explain. The lab doens provide any explanation.

@jkhushalani1
Channels means features. Number of channels means number of features.
    Think about a color violet. This color is made with adjustment of {red, blue, green}. These are three features.
    If input of an image is 240x250x4, then there are 4 individual features in the image.

1 Like

Thank you! That’s helpful.