Home › Guides
Radiation therapy image matching guides
Plain-English explainers on how setup-imaging registration actually works — the same skills you practice in the trainer. Written for radiation therapy students, therapists, and anyone learning IGRT.
What is IGRT?
The big picture: what image-guided radiation therapy is, why it matters for margins and dose, the main imaging modalities, and online vs offline correction.
Read the guide → CBCT · IGRTCBCT 6DOF registration explained
What cone-beam CT registration is, the six degrees of freedom, the matching workflow, window/level and fusion, tolerances, and the mistakes to avoid.
Read the guide → 2D / 2D · Portal2D/2D portal vs DRR matching
How orthogonal portal images are matched to DRRs, why it's a 5DOF correction, which bony landmarks to use, and how 2D/2D compares to CBCT.
Read the guide → Couch · 6DOFCouch shifts & 6DOF corrections
The six degrees of freedom, the Lat/Lng/Vrt and pitch/roll/yaw read-out, 5DOF vs 6DOF couches, and how action levels decide treat vs intervene.
Read the guide → Prostate · FiducialsProstate fiducial marker matching
Why gold seeds are implanted, how kV 2D/2D and CBCT marker matching works, why three markers give a full 6DOF read-out, and fiducial vs bony match.
Read the guide → Imaging · kV / MVkV vs MV imaging
How kilovoltage and megavoltage setup imaging differ — contrast, dose, the on-board imager and EPID, planar vs CBCT, and when a therapist uses each.
Read the guide → DIBH · Breath-holdDIBH deep inspiration breath-hold
Why breath-hold spares the heart in left-breast treatment, how the gating window and surface/respiratory monitoring work, and what good breath-hold coaching looks like.
Read the guide →Practice on real casesBrain, pelvis, thorax, breast, spine SBRT, lung SBRT, acoustic neuroma and prostate fiducials — 2D/2D and CBCT, with live residual-error feedback.
Open the trainerEducational use only. These guides and the trainer are for learning. This is not a medical device and not for clinical decision-making. All patient offsets and values in the trainer are fictional; imaging derives from de-identified, openly licensed research datasets (see image credits).