Since the SDK is at the imaging plane level, you have to load multiple planes to aggregate all the neurons recorded in a given session. Depending on the analysis, you may want to align them in time (since each mesoscope plane is acquired sequentially). We should have a simple function or code block that shows how to do this.