MRI Signal Measurement
ALMOST CORRECT.
Actually, all the answers are correct. Read through them, and then select "All of the above".
ALMOST CORRECT.
Actually, all the answers are correct. Read through them, and then select "All of the above".
ALMOST CORRECT.
Actually, all the answers are correct. Read through them, and then select "All of the above".

The transverse plane, and the net magnetisation vector at thermal equilibrium
CORRECT. The net magnetisation vector M is very small when compared to the external magnetic field, B (e.g. 1 μT vs. B = 1.5 T). It is an order of magnitude less than the earth's magnetic field (which is about 30 to 60 μT), and is virtually impossible* to measure when it is aligned, at equilibrium, with B along the z-axis. However, if M is rotated away from alignment with B, it precesses and the rotating component of the net magnetisation vector now in the transverse (xy) plane will generate a signal (a voltage) in receiving equipment.
Thus, the very small MRI signal can be measured perpendicular to the external magnetic field. This is important to note; much of the MRI process involves the manipulation of M (or part of M) into the transverse plane so that a signal can be recorded. Diagrams of the movement of the net magnetisation vector are usually made in the rotating frame (of precession at the Larmor frequency) to significantly simplify description.
* It is possible to measure longitudinal magnetisation at thermal equilibrium with a superconducting quantum interference device (SQUID).
Further reading on this topic:
Books: MRI From Picture to Proton p138, Spin Dynamics p36