[wikipedia]
A Computed Tomography (CT) machine acquires xray images from different angles around the head-to-toe axis of the patient.
The machine provides a stack of 2D axial xray "slices" of the patient, giving a 3D view of the internals.
Pre-operative applications are in diagnosis, modelling, and planning.
Intraoperative CT (iCT) can be used for tool guidance (e.g. brachytherapy and biopsy), but out-of-plane tool tracking is difficult as only a slice of the tool appears.
Some facts:
The CT is an xray machine that spins around the patient.
Below, $T$ is the xray tube, $X$ are the xrays, $D$ is the detector, and $R$ is the direction of rotation.
[radiopaedia.org]
CT machines have gone through several generations of development:
[radiologykey.com]
The raw output for one CT slice is a 2D "sinogram", $g(\theta,\rho)$:
Angles range in only [0,180] degrees because the rays at angles $\theta$ and $\theta + 180$ have almost the same attenuation. (Why not the same?)
The value in the sinogram at $(\theta,\rho)$ is the total xray attenuation along the corresponding ray, or
$\int \mu(t) \; dt$ ,
where the integral is taken betweeen the xray source and detector.
Recall the Beer-Lambert law:
$I_\textrm{out} = I_\textrm{in} \; e^{- \int \mu(t) \; dt }$
Then the total xray attenuation stored in the sinogram is
$\int \mu(t) \; dt \; = \; \ln I_\textrm{in} - \ln I_\textrm{out}$
$I_\textrm{in}$ is the known energy of the emitted xray.
$I_\textrm{out}$ is the detected energy at the detector element.
Question: What does the sinogram of an isolated fiducial look like?