You can go for different approaches. First, secretion processes can be followed by calcium imaging and membrane depolarization. The last suggestion is not as easy as fura-2 recordings, but you can follow intracellular potassium increase by dye loading as well.
Another way might be to load with something fluorescent those vesicles. If you had GFP-insulin, then each vesicle should look as very bright dots inside the cells. Upon secretion, those dots will move towards plasma membrane and eventually disappear from cytoplasm, and simultaneously, you can detect green fluorescence at extracellular space.
calcium imaging and imaging of membrane depolarization are relativly easy to perform, but are unspecific ways of imaging secretion. you would need good control measurements and to correlate that with amounts of insulin secreated could be problematic.
i would go with his second approach of fluorescently labeled insulin combined with imaging of a marker for secretory vesicles. since you want to image secretion on single cell level rise in fluorescence in the medium would not be an option for you, since it would be affected by all cells secreting insulin. but the amount of insulin in secretory vesicles should directly correlate to the amount of secreted insulin.