Imaging is often discussed as though scans differ mainly in detail or cost. They differ in what physical property they detect, and that determines which tissues appear at all.
X-ray records what absorbs radiation
A plain radiograph passes X-rays through the body onto a detector, and the image records how much was absorbed along each path.
Absorption depends on the density and atomic composition of tissue. Bone absorbs strongly, soft tissue less, air almost none.
This produces excellent bone contrast and poor discrimination between soft tissues, which are similar enough to each other to blur together.
CT applies the same physics with reconstruction
A CT scanner rotates an X-ray source around the body, collecting absorption measurements from many angles.
Software reconstructs those into cross-sectional slices, removing the overlap that makes a plain radiograph hard to interpret in depth.
Soft tissue contrast improves substantially as a result, though the underlying property being measured is still absorption, and the dose of ionising radiation is higher than a plain film.
Ultrasound measures reflected sound
An ultrasound probe emits high-frequency sound and records the echoes returning from boundaries between tissues of different acoustic properties.
It involves no ionising radiation, is portable, and shows movement in real time, which is why it is used for the heart, for pregnancy and for guiding needles.
Sound does not travel usefully through gas or bone, so structures behind the lungs or behind the skull are largely inaccessible to it.
MRI measures the behaviour of hydrogen nuclei
MRI places the body in a strong magnetic field, briefly disturbs hydrogen nuclei with radio waves, and records the signal as they return to alignment.
The rate of that return differs with the local chemical environment, which is why MRI distinguishes soft tissues that look identical on CT.
It is slower, noisier and more affected by movement, and the strong magnetic field makes certain implants a contraindication.
Choice follows the question, not the ranking
No modality is generally superior. A suspected fracture, an abdominal bleed, a foetus and a ligament tear each point to a different technique.
Availability, urgency, radiation dose and whether the patient can hold still all enter the decision alongside the physics.
This is also why a normal result on one scan does not exclude a problem another would have shown, and why interpretation belongs with the clinician who knows what was asked.