Residency · Residency · Radiation Oncology
Image-Guided Radiation Therapy (IGRT): Technologies and Workflows
Overview
Image-Guided Radiation Therapy (IGRT) involves the use of imaging at the time of treatment to verify and, if necessary, correct the patient’s position before or during radiation delivery. The primary goal of IGRT is to reduce geometric uncertainty, which allows for smaller planning target volume (PTV) margins and thereby improves the therapeutic ratio by maximizing tumor dose while minimizing normal tissue exposure. IGRT has evolved significantly over time, progressing from weekly port films to daily volumetric imaging and even real-time tracking during treatment. The choice of IGRT strategy depends on the specific clinical scenario, the target site, and the expected sources of geometric uncertainty, ensuring that the imaging modality and workflow are tailored to optimize accuracy and patient safety.
IGRT Technologies
2D Planar Imaging
kV Orthogonal Radiographs
Kilovoltage (kV) orthogonal radiographs are pairs of images taken at right angles using an onboard kV imaging system (OBI). These images are compared to digitally reconstructed radiographs (DRRs) generated from the planning CT scan. The acquisition of kV orthogonal images is fast and involves a relatively low imaging dose. This technique is best suited for matching bony anatomy because it does not provide adequate visualization of soft tissue targets. It is particularly useful for treatments involving the spine, brain (when immobilized with a frame or mask), and extremities.
MV Portal Imaging (EPID)
Megavoltage (MV) portal imaging uses the treatment beam itself to create radiographs via an electronic portal imaging device (EPID). This method verifies the shape of the radiation field and the alignment of bony anatomy relative to the treatment aperture. However, MV imaging delivers a higher imaging dose and offers poorer soft tissue contrast compared to kV imaging. One advantage is that portal images can be acquired during treatment delivery in cine mode, allowing real-time verification of beam-target alignment.
3D Volumetric Imaging
Cone-Beam CT (CBCT)
Cone-Beam CT is the most widely used 3D IGRT modality. kV CBCT is acquired by rotating the onboard kV imaging system around the patient, typically collecting over 200 projections across 200 to 360 degrees. MV CBCT uses the treatment beam but provides inferior image quality; however, it is available on all linear accelerators equipped with EPID. CBCT provides volumetric images that enable soft tissue matching rather than relying solely on bony alignment. The typical imaging dose for kV CBCT ranges from 1 to 5 cGy per scan. Limitations include respiratory motion artifacts, a limited field of view, lower image quality compared to diagnostic CT, and scatter artifacts. The standard acquisition time is between 30 and 60 seconds.
CT-on-Rails
CT-on-rails systems integrate a diagnostic-quality CT scanner into the treatment room on a rail system. The patient remains on the treatment couch while the CT gantry moves into position for scanning. This setup provides superior image quality compared to CBCT and allows for true Hounsfield unit accuracy, which is essential for adaptive dose calculation. Despite these advantages, CT-on-rails systems are less common due to their higher cost and the specific room design requirements they impose.
Surface-Guided Radiation Therapy (SGRT)
Optical Surface Monitoring
Surface-guided radiation therapy uses structured light projection and camera systems to create a real-time three-dimensional surface map of the patient. This live surface is compared to a reference surface obtained during simulation, providing six degrees of freedom (6DOF) position verification without any additional radiation dose. Common systems include AlignRT (Vision RT), Catalyst (C-RAD), and IDENTIFY (Varian). SGRT is particularly useful for breast deep inspiration breath hold (DIBH) monitoring, where real-time respiratory tracking is critical, frameless stereotactic radiosurgery (SRS) for intrafraction motion monitoring, open-mask treatments, and tattoo-free workflows where patient setup is performed without permanent skin marks.
Electromagnetic Tracking
Transponder-Based Systems
Electromagnetic tracking involves implanted transponders, such as Calypso beacons, which emit radiofrequency signals detected by an external array. This technology provides real-time three-dimensional localization of the target, most commonly the prostate, during treatment with sub-millimeter accuracy. Continuous tracking enables detection of intrafraction motion and allows for beam gating based on target position. Although highly accurate, electromagnetic tracking has largely been supplanted by MRI-guided approaches that provide real-time soft tissue visualization.
Fiducial Markers
Fiducial markers are radiopaque implants, such as gold seeds or carbon coils, placed in or near the tumor. These markers are visible on kV imaging, CBCT, and fluoroscopy, allowing localization of soft tissue targets independent of bony anatomy. Fiducials are commonly used in prostate cancer (gold seeds), liver stereotactic body radiation therapy (SBRT), and lung treatments with CyberKnife tracking. Limitations include potential marker migration, complications related to implantation, and the fact that markers serve as surrogates for the tumor rather than providing direct visualization.
MRI-Guided Radiation Therapy (MRgRT)
MR-Linac Systems
MR-linac systems combine an MRI scanner with a linear accelerator, such as the Elekta Unity or ViewRay MRIdian. These systems provide real-time soft tissue imaging during treatment delivery and enable online adaptive replanning based on the patient’s daily anatomy. MRI guidance offers the advantage of no ionizing radiation dose from imaging, superior soft tissue contrast, and the ability to gate treatment directly on tumor position. Challenges include the electron return effect at tissue-air interfaces, stringent MRI compatibility requirements, workflow complexity, and longer treatment times compared to conventional IGRT.
IGRT Workflows
Offline Correction
Offline correction involves acquiring and reviewing images but applying positional corrections only to subsequent treatment fractions, not the current one. This approach identifies systematic errors over the first few fractions and applies corrections accordingly. The No Action Level (NAL) protocol calculates the mean setup error from the initial fractions and applies a correction for the remainder of treatment. The extended NAL (eNAL) protocol continues to update corrections throughout the treatment course. While offline correction reduces systematic errors, it does not address random errors and is appropriate for lower-risk scenarios where daily correction is not essential.
Online Correction
Online correction entails acquiring images, reviewing them, and applying positional corrections immediately before each treatment fraction. This method reduces both systematic and random setup errors and is the standard workflow for most modern treatments. The process begins with patient positioning using skin marks, lasers, and surface guidance, followed by acquisition of a CBCT scan. The CBCT is registered to the planning CT using automatic algorithms with manual review. Couch shifts are then applied, including translational and rotational adjustments if a six-degree-of-freedom couch is available. Verification imaging with orthogonal kV images may be performed if needed before treatment delivery.
Intrafraction Monitoring
Intrafraction monitoring continuously tracks patient position during beam delivery and triggers a beam hold if the position deviates beyond a predefined tolerance. Technologies used for intrafraction monitoring include surface-guided radiation therapy, fluoroscopy for fiducial tracking, MRI for soft tissue tracking, and electromagnetic systems like Calypso. This monitoring is critical for stereotactic radiosurgery (SRS) and stereotactic body radiation therapy (SBRT), where even small positional shifts of 2 to 3 millimeters during a 15-minute delivery can lead to significant underdosage of the target or overdose of organs at risk. Intrafraction monitoring is also essential for DIBH treatments, where beam gating is based on respiratory position.
Matching Strategies
Bony Anatomy Matching
Bony anatomy matching aligns skeletal landmarks visible on IGRT images to those on the planning CT. This strategy assumes that the target position is stable relative to the bone. It is appropriate when the target is within or fixed to bone, such as in brain or spine treatments, or when soft tissue visualization is not feasible. However, it is insufficient when the target moves independently of the bone, as seen in prostate, lung, or liver tumors.
Soft Tissue Matching
Soft tissue matching directly aligns the target or adjacent soft tissue structures on volumetric imaging modalities like CBCT or MRI. This approach is preferred for targets such as the prostate, cervix (due to uterus and cervix motion), liver, and pancreas. Soft tissue matching is more accurate but requires adequate image quality and consistent anatomy for reliable registration.
Fiducial Matching
Fiducial matching aligns implanted markers as surrogates for the soft tissue target. This method is faster than soft tissue matching and less dependent on image quality. It assumes a stable spatial relationship between the markers and the target, which may not hold true for large tumors or after treatment-induced changes.
<image>A multi-panel comparison showing four IGRT technologies used for prostate cancer setup verification: (1) kV orthogonal radiographs with gold seed fiducials visible as bright dots on the AP and lateral images, (2) kV cone-beam CT axial slice showing the prostate, rectum, and bladder with fiducial markers, (3) surface-guided setup showing the patient's external contour compared to the reference surface, and (4) MR-linac sagittal image showing the prostate and surrounding soft tissue in real-time during treatment. Each panel is labeled with the technology name, typical imaging dose, and primary matching strategy (bony, soft tissue, fiducial, or surface).</image>
<image>A flowchart diagram of the online IGRT correction workflow. Steps include: patient setup with skin marks and SGRT -> acquire CBCT -> automatic registration to planning CT -> physician/therapist review of registration -> apply translational and rotational couch corrections -> verification imaging if needed -> begin treatment delivery -> intrafraction monitoring with SGRT or fluoroscopy -> beam hold if position exceeds tolerance -> treatment complete. Decision points and feedback loops are clearly marked.</image>
<image>A comparison of CBCT image quality (left) versus MR-linac image quality (right) for a rectal cancer case, both showing the same anatomic level. The CBCT shows the rectal tumor as a faintly visible soft tissue mass with limited contrast. The MR image clearly delineates the tumor, mesorectal fascia, and adjacent organs. Annotations highlight the superior soft tissue contrast of MRI and the implications for online adaptive planning.</image>
Key Clinical Pearls
IGRT is not a one-size-fits-all approach; the choice of imaging modality, matching strategy, and correction protocol must be carefully matched to the clinical scenario and the dominant source of geometric uncertainty. Daily CBCT with online correction represents the minimum standard for curative-intent intensity-modulated radiation therapy (IMRT) or volumetric modulated arc therapy (VMAT) in most disease sites, rendering weekly port films obsolete for high-precision treatments. The imaging dose from daily CBCT is non-trivial, accumulating to 1 to 3 Gy over a treatment course, which is particularly important to consider in pediatric patients. Surface guidance can reduce the need for daily ionizing radiation imaging. The most common IGRT error is not a failure of technology but a matching error; therefore, physicians must regularly review IGRT images and ensure that therapists are matching to the correct anatomy, focusing on the target rather than just bone. For SBRT and SRS, intrafraction monitoring is essential because a 2 to 3 mm shift during a 15-minute delivery can lead to clinically significant underdosage of the target or overdose of organs at risk. MR-guided radiation therapy represents the current frontier of IGRT by combining real-time soft tissue visualization with online adaptive planning; however, the longer treatment times and increased workflow complexity must be carefully weighed against the benefits for each clinical scenario.
References
- Jaffray DA. "Image-guided radiotherapy: from current concept to future perspectives." Nat Rev Clin Oncol. 2012;9(12):688-699.
- Dawson LA, Sharpe MB. "Image-guided radiotherapy: rationale, benefits, and limitations." Lancet Oncol. 2006;7(10):848-858.
- de Boer HC, Heijmen BJ. "eNAL: an extension of the NAL setup correction protocol for effective use of weekly follow-up measurements." Int J Radiat Oncol Biol Phys. 2007;67(5):1586-1595.
- Winkel D et al. "Adaptive radiotherapy: the Elekta Unity MR-linac concept." Clin Transl Radiat Oncol. 2019;18:54-59.
