[Insight-users] Confused abour Optimizer Scales

[Luis Ibanez]

Hi Auron,

Following your numerals:

1) Great !

2) A negative value in the optimizer scale,
    definitely doesn't make sense.

    What it will do is to move your Transform
    parameter in the wrong direction (e.g.
    against the good direction for improving
    the metric.

    Something else is wrong in your registration
    setup, and we should track it.

3) Not to criticize too much your image acquisition
    methods,

    but...

    if the images in the screen shot are from consecutive
    slices, how come that they are rotated about 50 degrees ?
    They are not even covering the full same region of the
    tissue, so, the best you can hope is for a partial match
    based on the left half of the tissue sample.

4) You could avoid to have to do difficult image processing,
    if some basic improvements were made on the process
    of acquiring the images from the tissue samples.
    Normally in a microscopy stack you wouldn't see rotations
    larger than a couple of degrees.