That is a very thoughtful question! I have been thinking about this too. I am no expert, but I wanted to explore the idea of combining auditory cues with graphical passwords for people with disabilities. This involves utilising multiple senses, can involve additional paired devices, and can be susceptible to 'Man-in-the-Middle' attacks, but that’s as far as I could think of.

A user-recorded or pre-recorded audio clip can be fragmented into say 3 parts, and a blind user could assign these fragments to tiles on a screen. His main goal would now be to click the right tiles on the screen following audio cues.

Would love to hear your thoughts and ideas on the same!