These modes convert various portions of the color spectrum to a gray-scale image to allow those with any type of color deficiency to perceive light of that frequency.
The Red/Green/Blue (RGB) color model is the basis of typical computer displays and image storage. Each of these three modes take one of the color channels and convert it to gray-scale. This may help reveal details hidden to those with color vision deficiency.
The LMS color model is a closer approximation to the three types of cone receptors in the human eye. Each of these modes displays the approximate stimulation to each of those receptors. This may help those individuals missing a particular receptor see information they have been missing.
This mode simply takes the full color image and converts it to monochrome. This image should appear essentially the same to everyone.