Pattern recognition in videos using Neural network

I'm new to neural networks and pattern recognition. Currently I have videos that I have to manually observe and look for certain behaviours/phenotypes. As these videos are recorded for weeks at a time, it quickly becomes an exhaustive process. I've looked into neural networks and pattern recognition and was wondering if it would be possible to train a neural network to classify observed behaviours in videos? I've seen the pattern recognition done for single images (tumours, etc.) in the Matlab demos, but I'm not sure how I would approach this for videos.

Does anyone have any idea how you would train a neural network to identify and classify patterns occurring over time in a video?

Any help would be great!