Ensuring Clinical Readiness: Commissioning, Acceptance and Performance Testing of Closed Bore Linear Accelerator Compatible Surface Guided Radiotherapy System

Abstract

Introduction:

Surface Guided Radiation Therapy (SGRT) enhances radiation therapy by providing real-time support without additional X-ray exposure. It ensures precise patient positioning, continuous monitoring, and motion management. However, closed-bore LINACs face optical line-of-sight challenges with ceiling-mounted SGRT systems.

Objective:

This study commissions the AlignRT InBore™ SGRT system on the Ethos™ LINAC at a Mumbai tertiary care center, evaluating accuracy, precision, reproducibility, and temporal stability.

Methods:

System Commissioning: 1) Acceptance tests per Vision RT’s Form 412: Camera calibration, setup validation, thermal stability, relative shift accuracy. Phantom measurements for performance assessment. 3) Deliberate rotational motion errors to test detection capability.

Results:

Comparison with Existing Systems: 1) Consistent with SGRT performance on Halcyon™ LINAC (Nguyen et al.). 2) Reliable mechanical and imaging test results. Patient-Specific QA: 1) 51 adaptive treatment sessions. 2) High gamma pass rates confirmed clinical efficacy. 3) Essential for Ethos™ LINAC, which lacks 6D couch correction.

Conclusion:

This study demonstrates successful SGRT integration with Ethos™, improving treatment accuracy, patient comfort, and efficiency in closed-bore LINACs, advancing radiation oncology in India.

INTRODUCTION

Surface guided radiation therapy (SGRT) has significantly advanced the field of radiation therapy by providing real-time support during image-guided radiation therapy without the need for additional X-ray exposure. SGRT ensures precise patient positioning, continuous monitoring, and effective motion management, such as beam-gating during free-breathing or deep-inspiration breath-hold. It enhances the efficiency of patient setup compared to traditional tattoo markers, increases patient comfort by using open-faced masks, and reduces dependency on daily port or X-ray imaging.[1,2,3,4] However, closed bore linear accelerators face challenges with ceiling-mounted SGRT systems due to optical line-of-sight limitations, which impede motion tracking within the bore.[5]

The authors aim to discuss the commissioning of the AlignRT InBore™ (AlignRT, Vision RT, London, UK) SGRT system on the Ethos™ Linear Accelerator system (a closed bore linear accelerator by Varian, Palo Alto, USA) at a tertiary care center in Mumbai, India during the period of May–June 2023.

MATERIALS AND METHODS

All tests and quality assurance (QA) procedures were conducted in accordance with the guidelines established in AAPM Task Group 147 (QA for nonradiographic localization and positioning systems) and AAPM Task Group 302 (surface guided radiotherapy protocols). These protocols ensure the accuracy, reproducibility, and safety of SGRT systems, with specific focus on the integration of advanced imaging technologies and patient setup verification.

Phantom measurements

Phantom measurements were performed to evaluate the system’s accuracy, precision, reproducibility, and temporal stability. The phantom was translated in lateral, longitudinal, and vertical directions and rotated around two axes (pitch, roll) using a precise, in-house built positioning stage.

Acceptance tests

Acceptance tests were conducted according to Vision RT Form 412.

System calibration

The calibration protocol includes daily, monthly, and MV radiation isocenter checks. Daily and monthly QA procedures ensure that the cameras are accurately aligned with the treatment isocenter. The monthly calibration requires a vendor-supplied calibration plate, consisting of a white slab with a 32 × 32 grid of black spots (“blobs”), each approximately 1 cm in diameteperipherr and spaced 2 cm apart. Four of these spots, located at the corners of a 10-blob square, are larger (1.7 cm in diameter) and numbered for reference. The daily QA also uses a calibration plate to verify that the cameras remain correctly positioned relative to each other since the previous calibration. MV isocenter calibration is not feasible with ring gantry systems, as the Ethos™ platform supports only Kilovoltage cone beam computed tomography (CBCT) imaging during treatment.

• Calibration of align RT in-bore camera system:

  • Plate calibration test and daily QA verification test were performed

  • The root mean square value was recorded and deemed acceptable per acceptance testing protocol (ATP) (<0.2 mm).

• Setup ceiling camera system validation and analysis, and real-time data (RTD) performance:

  • Thermal stability: The ceiling camera and ring camera were positioned at the isocenter. A reference surface was captured and monitored over 20 min to verify system stability and prevent significant measurement drift

  • Relative shift accuracy: The phantom was positioned at the isocenter, a reference image was captured, and the phantom was moved in 1.0 mm steps to a total translation of 10 mm in all three translational directions

  • Ambient lighting stability: The stability of delta values was tested when room lighting was switched from setup to treatment conditions [Figure 1].

Figure 1.

(a) InBore camera: thermal stability – Plot of delta mag against elapsed time, (b) ceiling camera: thermal stability – Plot of delta mag against elapsed time

In-bore camera features

The cameras in the AlignRT in-bore system are used to generate a three-dimensional surface model of the patient by capturing images that enable precise patient positioning during setup. The system is equipped with the Horizon camera, which has the following specifications: dimensions of 480 mm (W) ×127 mm (H) ×140 mm (D) and a weight of 3.5 kg. The camera operates on a voltage supply ranging from 100V to 240V.

• Ambient light effect: Testing delta values under different lighting conditions are a common QA practice to ensure that lighting changes do not affect the system’s localization accuracy. This test evaluates the stability of delta values when transitioning room lighting from setup to treatment mode. The stability of the in-bore camera is measured to ensure consistent localization under different lighting conditions

• Camera occlusion test: It’s accurate to test whether occlusion of part of the setup field (e.g. by the gantry head) affects camera accuracy. For the Ceiling Camera, this test assesses the stability of delta values when part of the lateral pod is occluded by the gantry head. The difference in mean values is calculated between monitoring sessions conducted without and with occlusion to determine if the occlusion affects the system’s accuracy.

Real-time monitoring (real-time datavalues)

A reference capture was taken from the in-bore camera, and the couch was moved 1 cm along each axis for motion tracking. Accuracy was confirmed to be <0.5 mm.

• Spatial reproducibility and drift: A thermal stability test of the AlignRT system was conducted using a phantom. The phantom was monitored over a 20-min period, during which the overall spatial drift was calculated for both the setup and the in-bore camera systems. This assessment helps quantify any positional deviations due to thermal changes during the procedure

• Static localization accuracy: A procedure was developed to assess the SGRT system’s ability to monitor and track a reference surface. In this test, an MV isocenter phantom was used, and deliberate shifts of the treatment couch were made in vertical, lateral, and longitudinal directions by known increments. Movements ranged from 0.1 cm to 0.5 cm in 1 mm increments, from 0.5 cm to 3 cm in 0.5 mm increments, and from 3 cm to 5 cm in 1 cm increments. These shifts were captured by the SGRT system using both the in-bore and setup cameras. This test established the correlation between the SGRT system’s coordinates and the couch reference. In addition to translational movements, rotational shifts were also evaluated. The difference between the applied and detected shifts reflects the static localization accuracy of the SGRT system for translational movements. Notably, the Ethos™ platform supports only translational movements and does not feature 6D couch capabilities

• Dynamic localization accuracy: The Ethos™ system does not include a dedicated motion management interface, lacks a full-fledged respiratory motion management interface like those found in other systems such as Varian’s TrueBeam, which incorporates gated beam delivery for respiratory motion. Similarly, the AlignRT system does not feature automatic beam-hold functionality. However, dynamic localization accuracy was assessed using the Varian RPM phantom and the SGRT mannequin. A respiratory wave pattern was simulated during CT imaging, and both the in-bore and setup camera systems were tested under identical conditions to verify their ability to reproduce the same wave pattern.

Co-calibration of ceiling setup camera and in-bore ring camera system

• Correspondence between ceiling-mounted and InBore™ cameras was established during calibration using a MV cube (Vision RT Ltd., London, UK), ensuring accurate transition from virtual to treatment isocenter

• The MV Cube was aligned to sub 0.2 mm/0.1° accuracy using the ceiling-mounted camera, then automatically moved inside the bore to the treatment isocenter using Varian™-specific delta couch shifts

• InBore™ monitoring was activated, and RTDs were compared to ceiling-mounted values.

Cone beam computed tomography image quality check with ring gantry

• An advanced electron density phantom with 10 inserts was imaged with and without the ring to check image quality and HU values.

Initial patient-specific quality assurance with ring gantry

• Patient-specific QA was conducted for five patients across 51 online adaptive treatment sessions, with gamma passing rates assessed using ArcCHECK™ (Sun Nuclear Corporation, Melbourne, Florida, USA).

Quality assurance test protocol

• Daily: The calibration plate should be used to ensure that the camera positions remain undisturbed. This step verifies that the system is accurately aligned and ready for treatment

• Monthly:

  1. The alignment between the localization system isocenter and the treatment isocenter should be evaluated

  2. Motion tracking accuracy should be assessed by moving the phantom a known distance (e.g., 5 cm) and verifying that the localization system detects the shift within a tolerance of 2 mm.

• Annually:

  1. The alignment between the localization system isocenter and the treatment isocenter should be thoroughly evaluated

  2. Motion tracking accuracy should be tested by moving the phantom a known distance (e.g., 5 cm) and confirming that the localization system accurately registers the shift within 2 mm of the expected value.

End-to-end test

We performed an end-to-end test involving a patient scan using a Siemens Biograph™ machine. The acquired scan was exported to the varian contouring workstation for delineation, as there is no direct communication with the Ethos system. To enable availability in both the Eclipse and Ethos™ systems, all patients must first be registered in the ARIA system. Once registered, the patient’s treatment plan was generated on Ethos.

Due to the lack of direct integration with the Ethos™ Treatment Planning System, the approved plan was then exported to Eclipse for laser origin alignment and Delta couch insertion. After these adjustments, the plan was re-exported to Ethos™. This process allowed for a comprehensive end-to-end verification, ensuring the integration and functionality of all peripheral systems involved.

Integration of peripheral equipment

The initial treatment planning scan’s DICOM coordinates and patient body surface data are imported and subsequently exported to the Ethos™ treatment console. However, due to the lack of direct integration, the Ethos™ treatment plan must first be routed through the Eclipse TPS for essential setup adjustments, including delta couch values and laser origin. This indirect process ensures the integrity of the treatment plan by preserving the original DICOM coordinates. In addition, the imported patient body surface data allows for accurate treatment monitoring, further enhancing precision during radiation delivery.

RESULTS

• Acceptance tests were performed in accordance with Vision RT’s Form 412 requirements and demonstrated successful outcomes, achieving sub-millimeter accuracy and <1° of rotational and translational shifts. A precise transition from virtual to treatment isocenter was successfully accomplished [Table 1]

• Real-time monitoring: The results for static localization accuracy, and dynamic localization are summarized in Table 2 and Figure 2a, b

• Patient-specific QA patient-specific QA using ArcCHECK™ with a 95% gamma pass rate criterion (3%, 3 mm) was passed for all 5 adaptive patients across 51 delivered fractions [Table 3]

• CBCT image quality check: Image quality checks were conducted for various density materials and reported all values within the tolerance limit of 50 HU [Figure 3 and Table 4]

• Verification of Rotational Shifts Despite the Varian Ethos™ couch’s inability to apply rotational shifts, pitch and roll were minimized and verified using the SMARTTool™ and the MV Cube phantom. Intentional pitch and roll shifts were introduced which were successfully detected by SGRT [Table 5]

• The integration of peripheral equipment with the Ethos system is illustrated in Figure 4.

Table 1.

Acceptance tests

Camera	Test name	Maximum threshold	Error read	Pass/fail
In-bore camera	Thermal stability test	Drift LNG <1.0 mm	0.20 mm	Pass
		Drift LAT <1.0 mm	0.13 mm	Pass
		Drift VRT <1.0 mm	0.11 mm	Pass
		Drift ROLL <1.0°	0.06°	Pass
		Drift PITCH <1.0°	0.08°	Pass
		Drift YAW <1.0°	0.04°	Pass
Ceiling camera	Thermal stability test	Drift LNG <1.0 mm	0.37 mm	Pass
		Drift LAT <1.0 mm	0.52 mm	Pass
		Drift VRT <1.0 mm	0.29 mm	Pass
		Drift ROLL <1.0°	0.08°	Pass
		Drift PITCH <1.0°	0.09°	Pass
		Drift YAW <1.0°	0.09°	Pass
In-bore camera	Relative shift accuracy test
	LNG + 10 mm	LNG error <1.0 mm	0.06±0.04 mm	Pass
	VRT + 10 mm	VRT error <1.0 mm	0.29±0.01 mm	Pass
	LAT + 10 mm	Lat error <1.0 mm	0.08±0.02 mm	Pass
Ceiling camera	Relative shift accuracy test
	LNG + 10 mm	LNG error <1.0 mm	0.13±0.03 mm	Pass
	VRT + 10 mm	VRT error <1.0 mm	0.05±0.03 mm	Pass
	LAT + 10 mm	Lat error <1.0 mm	0.06±0.05 mm	Pass
In-bore camera	Ambient lighting during patient alignment and treatment	Drift LNG <0.2 mm	0.01mm	Pass
		Drift LAT <0.2 mm	0.00 mm	Pass
		Drift VRT <0.2 mm	0.01 mm	Pass
		Drift ROLL <0.2°	0.00°	Pass
		Drift PITCH <0.2°	0.00°	Pass
		Drift YAW <0.2°	0.00°	Pass
Ceiling camera	Ambient lighting during patient alignment and treatment	Drift LNG <0.2 mm	0.02 mm	Pass
		Drift LAT <0.2 mm	0.01 mm	Pass
		Drift VRT <0.2 mm	0.01 mm	Pass
		Drift ROLL <0.2°	0.00°	Pass
		Drift PITCH <0.2°	0.01°	Pass
		Drift YAW <0.2°	0.00°	Pass
Ceiling camera	Delta stability and pod occlusion test, gantry at 0°	LNG <1.0	0.14±0.09 mm	Pass
		LAT <1.0 mm	0.01±0.03 mm	Pass
		VRT <1.0 mm	0.10±0.02 mm	Pass
		ROLL <1.0°	0.11±0.01°	Pass
		PTICH <1.0°	0.00±0.03°	Pass
		YAW <1.0°	0.05±0.02°	Pass
		MAG SD <0.2 mm	0.07 mm	Pass

LNG: Longitudinal, LAT: Lateral, VRT: Vertical, MAG SD: Magnitude of standard deviation

Table 2.

Stable localisation accuracy (a) InBore camera (b) ceiling camera

InBore camera
Vertical			Longitudinal			Lateral
Translation (cm)	SGRT	Difference	Translation (cm)	SGRT	Difference	Translation (cm)	SGRT	Difference
0.10	0.09	0.01	0.10	0.10	0.00	0.10	0.10	0.00
0.20	0.20	0.00	0.20	0.20	0.00	0.20	0.20	0.00
0.30	0.30	0.00	0.30	0.31	−0.01	0.30	0.30	0.00
0.40	0.40	0.00	0.40	0.41	−0.01	0.40	0.41	−0.01
0.50	0.50	0.00	0.50	0.51	−0.01	0.50	0.50	0.00
1.00	1.0	0.00	1.00	1.01	−0.01	1.00	1.01	−0.01
1.50	1.51	−0.01	1.50	1.51	−0.01	1.50	1.52	−0.02
2.00	2.01	−0.01	2.00	2.03	−0.03	2.00	2.01	−0.01
2.50	2.52	−0.02	2.50	2.54	−0.04	2.50	2.51	−0.01
3.00	3.53	−0.53	3.00	3.04	−0.04	3.00	3.01	−0.01
4.00	4.03	−0.03	4.00	4.05	−0.05	4.00	4.01	−0.01
5.00	5.04	−0.04	5.00	5.05	−0.05	5.00	5.03	−0.03
Mean±SD	0.003±0.1446		Mean±SD	0.0025±0.0181		Mean±SD	0.0125±0.0862
Ceiling camera
−0.10	−0.10	0.00	−0.10	−0.10	0.00	−0.10	−0.10	0.00
−0.20	−0.20	0.00	−0.20	−0.20	0.00	−0.20	−0.20	0.00
−0.30	−0.29	0.01	−0.30	−0.30	0.00	−0.30	−0.31	0.01
−0.40	−0.39	0.01	−0.40	−0.40	0.00	−0.40	−0.41	0.01
−0.50	−0.50	0	−0.50	−0.50	0.00	−0.50	−0.51	0.01
−1.00	−0.99	0.01	−1.00	−1.01	0.01	−1.00	−1.01	0.01
−1.50	−1.50	0.00	−1.50	−1.50	0.00	−1.50	−1.50	0.00
−2.00	−2.00	0.00	−2.00	−2.02	0.02	−2.00	−2.01	0.01
−2.50	−2.50	0.00	−2.50	−2.52	0.02	−2.50	−2.52	0.02
−3.00	−3.01	0.01	−3.00	−3.00	0.03	−3.00	−3.02	0.02
−4.00	−4.02	0.02	−4.00	−4.01	0.01	−4.00	−4.01	0.01
−5.00	−5.04	0.04	−5.00	−5.00	0.03	−5.00	−5.00	0.00
Mean±SD	0.003±0.0137		Mean±SD	0.0025±0115		Mean±SD	0.0125±0.0068

SD: Standard deviation, SGRT: Surface guided radiation therapy

Figure 2.

(a) Dynamic Localisation, (b) Dynamic Localisation

Table 3.

Patient specific quality assurance conducted using ArcCHECK™

	Number of fractions	Average gamma pass %
Patient 1	20	96.6
Patient 2	11	98.5
Patient 3	10	98.2
Patient 4	6	96.4
Patient 5	5	97

Figure 3.

Cone beam computed tomography image quality checks: without ring (L), with ring (R)

Table 4.

CBCT image quality checks

Material	Without ring HU	With ring HU	Difference
Solid water	1.29	2.17	−0.88
True water	−843.4	−849.94	6.54
Liver	−84.67	−55.98	−28.69
General adipose	−90.68	−89.48	−1.2
Aluminium	1598.96	1616.59	−17.63
Brain	−6.39	−4.94	−1.45
Cortical bone	759.53	795.07	−35.54
Breast	−16.49	−17.11	0.62
CaCo3	676.8	711.8	−35
Air 1	−877.98	−876.5	−1.48
Air 2	−912.85	−918.47	5.62

HU: Hounsfield units, CBCT: Cone beam CT

Table 5.

Verification of rotational shifts

Roll reading			Pitch reading
Set value (°)	In bore measured value	Difference	Set value (°)	In bore measured value	Difference
0	0	0	0.1	0.1	0
0.6	0.6	0	1.1	1.2	0.1
0.9	0.8	−0.1	2.2	2.2	0
2	1.9	−0.1	3.5	3.5	0
3	2.9	−0.1	4.1	4	−0.1
3.8	3.7	−0.1	5	5	0
5	4.8	−0.2

Figure 4.

Integration of peripheral equipment workflow

DISCUSSION

The Varian Ethos™ is an advanced adaptive radiotherapy system that integrates artificial intelligence and machine learning to enhance cancer treatment. It is designed to create personalized treatment plans daily, adapting to changes in the patient’s anatomy. This system uses multi-modality imaging, such as CT, positron emission tomography, MR, and CBCT, to ensure precise treatment delivery. AlignRT^® InBore™ is a SGRT solution that complements Varian Ethos™ by providing accurate patient positioning and monitoring inside bore-based linear accelerators. Since this technology is new and offers different features compared to open bore compatible SGRT solutions, it is essential to commission the system carefully to ensure precise accuracy.

The authors would like to highlight the innovative nature of this study, as they were unable to find prior research incorporating both technologies. Nguyen et al. have performed commissioning and performance testing of the SGRT system on the Halcyon™ linear accelerator, which is quite similar to the Ethos™ in most hardware aspects.[5]

This study aligns well with Nguyen et al.’s findings,[5] particularly in the assessment of mechanical and imaging tests, demonstrating consistent results. Moreso, all the tests performed met the ATP as outlined in the Vision RT’s Form 412. In addition to the required tests, we deliberately introduced rotational motion errors, such as pitch and roll, to assess the SGRT system’s ability to detect these inaccuracies. This evaluation is particularly relevant because the Ethos™ linear accelerator lacks a 6D couch correction capability. Integrating the SGRT system could therefore help minimize these positional errors, enhancing the overall accuracy of the treatment.

CONCLUSION

In conclusion, the commissioning and acceptance testing of the AlignRT InBore™ SGRT system on the Ethos™ Linear Accelerator demonstrate robust performance in ensuring precise patient positioning within a closed bore environment. The comprehensive methodology, validated through phantom measurements, acceptance tests, and patient-specific QA, underscores its efficacy in clinical practice. These findings support the integration of SGRT technology to enhance efficiency, patient comfort, and treatment accuracy in radiotherapy settings utilizing closed bore LINACs. This study represents the first documentation of such comprehensive testing and implementation within an Indian healthcare setting, contributing uniquely to the field of radiation oncology in the region.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.