In the computer world there are two different ways of interpreting 16-bit data:
1) All 16 bits are significant, yielding values from 0 to 65,535
2) The high-order bit is a sign bit, yielding values from -32,768 to +32,767.
This is a decision made at design time for a particular piece of equipment.
In a photographic context, #1 makes more sense than #2, since you'll never have a negative light value. However, some engineers are too stupid to realize that.
If a particular MFDB used format #2 then it would be proper to show the data depth as 15 bits.
- Leigh