Procedures · Head, Spine and Pelvis

Skull Positioning Series

Compare the four routine skull projections: Towne, lateral, Caldwell, and straight PA.

Skull

StudyBucky

Loading model…

Hover a bone, name it, then click to check · drag to rotate

The routine skull series

The projections this region is examined with. Set them up yourself in Practice positioning.

AP axial skull (Towne method)

Anteroposterior axial projection · central ray 30° caudad to the orbitomeatal line

Demonstrates: The occipital bone, the petrous pyramids and the foramen magnum, with the dorsum sellae and the posterior clinoid processes projected inside the shadow of the foramen magnum. Symmetric petrous ridges prove no rotation — a ridge narrows on the side the head is turned toward. Seeing the dorsum sellae and the posterior clinoids WITHIN the foramen magnum is what proves the central-ray angle and the neck flexion were both right: too little angle or too little flexion throws the dorsum sellae above the foramen magnum, and too much drops the posterior arch of C1 over it and foreshortens it. Clinoid processes shifted off to one side within the foramen magnum mean the head was tilted.
Part position: Erect or supine, with all metal and plastic removed from the head. Depress the chin until the orbitomeatal line lies at 90° to the image receptor, adding a radiolucent support under the head if it helps; when the neck will not flex that far, bring the infraorbitomeatal line to 90° instead and take the alternative central-ray angle. Align the midsagittal plane to the central ray and to the midline of the grid, with no rotation or tilt of the head, and check that the vertex of the skull falls inside the collimated field.
Central ray: Angled 30° caudad to the orbitomeatal line, or 37° caudad when the infraorbitomeatal line is the one brought to 90° to the image receptor, centered to the midsagittal plane 2½ inches (6.5 cm) above the glabella to pass through the foramen magnum at the level of the base of the occiput.
Why it is positioned this way: A straight AP stacks the dense petrous pyramids over the occipital bone and closes the foramen magnum. Swinging the tube 30° caudad throws the petrous ridges and the facial bones downward out of the way, opens the foramen magnum into the middle of the image, and drops the sellar structures neatly inside it. The angle is measured to the orbitomeatal line rather than to the table, which is exactly why the second number exists: the infraorbitomeatal line sits 7° to 8° away from the orbitomeatal line, so a patient who can only reach that line needs 37° to keep the same 30° between the beam and the skull. The two positions give the same anatomic relationships, and only the number on the tube changes.

Also covered by this projection

The PA axial alternativeA PA axial with the central ray 25° cephalad to the orbitomeatal line shows the same occipital bone, petrous pyramids and foramen magnum. It exists for patients who cannot manage the Towne — including a limited range of cervical motion — and it lowers the dose to the facial structures and the thyroid gland, because the beam exits rather than enters through them. It is centered 1½ inches (4 cm) below the inion and exits 1½ inches (4 cm) above the nasion. The catch is magnification: the occipital bone is now the structure farthest from the receptor, so this is NOT the projection to choose when the occipital bone itself is the area of interest.

How far the angle can goThe central ray for an AP axial should not be pushed past 45°, because beyond that the distortion costs more than the extra clearance buys. If 30° to the orbitomeatal line cannot be reached before that ceiling — which happens with a patient in a cervical collar whose neck is extended, so the usual line relationships no longer hold — the dorsum sellae and the posterior clinoid processes end up projected above the foramen magnum instead of inside it.

Trauma comes first, and it is not thisCervical spine fractures, subluxations and dislocations must be ruled out before the head or neck is moved at all. A trauma skull is a horizontal-beam lateral plus an AP and an AP axial taken with nothing moved and the collar left on, Rapid CT is available in most hospitals that treat head injuries, and using it routinely to triage the minor and mild ones — deciding who needs admission or surgery and who can safely be discharged — is now advocated.

Lateral skull

Right or left lateral position · central ray 90° to the image receptor

Demonstrates: The whole cranium with the parietal bones superimposed, and the entire sella turcica — anterior and posterior clinoid processes and dorsum sellae — with the sella turcica and the clivus shown in profile. The two position errors separate different structures in different directions, which is what makes this image readable: ROTATION separates structures that should superimpose side to side, such as the mandibular rami and the greater wings of the sphenoid, while TILT separates structures that should superimpose top to bottom, such as the orbital plates and those same greater wings.
Part position: Erect, or recumbent in a semiprone position, with the side of interest closest to the image receptor and the body obliqued as far as comfort needs. Align the midsagittal plane parallel to the image receptor with no rotation or tilt, bring the interpupillary line to 90° to the image receptor, and adjust neck flexion until the infraorbitomeatal line sits at 90° to the front edge of the receptor, which lays the glabelloalveolar line parallel to that edge. Palpating the external occipital protuberance behind and the nasion or glabella in front, and checking that the two are the same distance from the receptor, is the practical test for a true lateral. A recumbent patient usually needs a radiolucent support under the chin, a broad-chested patient one under the whole head, and a thin patient one under the upper thorax.
Central ray: 90° to the image receptor, centered to a point 2 inches (5 cm) above the external acoustic meatus — or midway between the glabella and the inion for skull shapes where that landmark reads better.
Why it is positioned this way: Every structure in the cranium is paired, and on this projection every pair is superimposed, so the whole image is one long symmetry test — which is what makes a lateral skull so unforgiving and so useful. A little rotation pulls the mandibular rami apart sideways; a little tilt pulls the orbital plates apart vertically. Two different mistakes, two different directions of separation, and one image that names which one happened. The sella turcica sits on the midline in the middle of the skull, so a true lateral is also the only projection in this routine that shows it and the clivus in profile.

Also covered by this projection

Which side goes downThis projection is printed as a right OR left lateral, and the side of interest goes closest to the image receptor — the structure nearest the receptor is the one imaged with the least magnification and the most detail. This is the RIGHT lateral, so the right-side bones are the ones against the receptor and nearest you. The skull is close enough to symmetric that the choice is made by the clinical question rather than by convention, unlike the spine, where the routine names a side outright.

The trauma lateral is a horizontal beamFor a trauma patient this projection is taken with the patient supine and the tube swung to send the beam horizontally, without turning the head. Beyond keeping the cervical spine still, the horizontal beam is what makes an air-fluid level visible: fluid settles and its surface stays level, so blood in the sphenoid sinus shows as a line across it — a sign of a basal skull fracture. Angling the tube down at a supine patient instead would lose that line completely.

What the sella turcica is worth hereA pituitary adenoma is investigated with CT or MRI, but an enlarged sella turcica or an eroded dorsum sellae still turns up on plain images, often as an incidental finding — and the collimated lateral and Towne are the two projections that show it. That pairing is a good way to remember the two: the lateral gives the sella in profile, the Towne gives it projected inside the foramen magnum.

PA axial skull (Caldwell method)

Posteroanterior axial projection · central ray 15° caudad

Demonstrates: The frontal bone, the greater and lesser wings of the sphenoid, the superior orbital fissures, the frontal and anterior ethmoid sinuses, the supraorbital margins and the crista galli. At the routine 15° the petrous pyramids land in the lower third of the orbits and the supraorbital margin is clear of superimposition. Rotation is measured by comparing the distance from each midlateral orbital margin to the lateral cortex of the skull on that side, and the comparison also tells you which way: if the right gap is the larger one, the face is turned toward the left.
Part position: Erect or prone, with the nose and forehead resting against the imaging device. Flex the neck as needed to bring the orbitomeatal line to 90° to the image receptor, and set the midsagittal plane at 90° to the receptor so the head neither rotates nor tilts.
Central ray: Angled 15° caudad, centered to the midsagittal plane to exit at the nasion.
Why it is positioned this way: On a PA the frontal bone lies against the receptor, which is where you want the bone you are trying to see — but a beam with no angle drops the petrous ridges straight across the orbits. Fifteen degrees caudad pushes those ridges down into the lower third of the orbits and clears the supraorbital margins, and it does that without lifting the frontal bone off the receptor, so the bone of interest keeps its position and its sharpness. The angle and the orbitomeatal line work together rather than separately: a steeper angle OR a deeper chin tuck sends the ridges lower, which is why the position has to be described by both.

Also covered by this projection

The steeper alternative on a skull seriesThe routine also allows 25° to 30° caudad as a straight substitute for the 15°, exiting at the level of the mid orbit. The extra angle drives the petrous pyramids to or just below the infraorbital margin, so the entire orbital base is visible, and it brings out the superior orbital fissures and the foramen rotundum beside each infraorbital margin.

The facial-bone number is 30°, not a rangeThe Caldwell also belongs to the facial-bone routine, and its steep option there is stated as a single figure: 30° caudad, used specifically when the ORBITAL MARGINS are the area of interest, which projects the petrous ridges below the infraorbital margin. Carrying the skull series’ 25° to 30° range over to an orbit question is a common way to get it wrong — the orbit figure is 30°.

When the patient cannot be turned face-downA patient who cannot be positioned PA — a trauma patient on a board, most often — gets the mirror image instead: an AP axial with the central ray 15° CEPHALAD and the orbitomeatal line at 90° to the image receptor. The beam now enters the face and exits the back of the head, so the same structures appear, magnified, at a higher dose to the eyes and thyroid gland.

PA skull

Posteroanterior projection · central ray 90° to the image receptor, parallel to the orbitomeatal line

Demonstrates: The frontal bone, the crista galli, the internal auditory canals, the frontal and anterior ethmoid sinuses, the petrous ridges, the greater and lesser wings of the sphenoid and the dorsum sellae. The petrous portion of the temporal bone FILLS the orbits, with the petrous ridges sitting at the level of the supraorbital margins — on this projection that is the correct appearance and not a positioning error. Equal distances from each lateral orbital margin to the lateral cortex of the skull prove no rotation, and the anterior and posterior clinoid processes appear just above the ethmoid sinuses.
Part position: Erect or prone, with the nose and forehead resting against the imaging surface. Flex the neck until the orbitomeatal line lies at 90° to the image receptor, and set the midsagittal plane at 90° to the midline of the imaging device so the head neither rotates nor tilts — both external acoustic meatuses the same distance from the surface.
Central ray: 90° to the image receptor and parallel to the orbitomeatal line, centered to the midsagittal plane to exit at the glabella.
Why it is positioned this way: This projection is kept in the routine for one stated reason: to show the frontal bone with the least distortion of anything in the series. The frontal bone lies against the receptor and the beam runs straight through it with no angle at all, so nothing about it is elongated or foreshortened. The price is that the petrous ridges land squarely across the orbits. That is why the routine carries the Caldwell AS WELL as this, rather than instead of it — same patient position, one 15° angle, and the orbits clear, at the cost of the small distortion the angle introduces. Reading the two tabs against each other is the fastest way to see what a central-ray angle actually buys and what it costs.

Also covered by this projection

PA against Caldwell, in one lineSame position, same landmarks, one difference: the PA has no tube angle and the Caldwell has 15° caudad. So the PA gives the frontal bone with the least distortion and the petrous ridges filling the orbits, and the Caldwell gives slightly more distortion and the orbits clear down to their lower third. Neither is a better version of the other; they answer different questions, which is why both are in the routine.

The centring points are different tooThe PA exits at the glabella and the Caldwell exits at the nasion — a small distance apart on the same face, and an easy pair to swap under exam pressure. The glabella is the smooth prominence between the eyebrows; the nasion is the depression at the bridge of the nose where the frontal bone meets the nasal bones, one step lower.

Procedures lessonSkull / facial / sinuses