Real-time respiration monitoring using the radiotherapy treatment beam and four-dimensional computed tomography (4DCT)—a conceptual study

Multi-channel respiratory signal detection system for 4D-CT in radiotherapy by measuring the back pressure

This study proposes a novel respiratory signal detection system for 4D-CT in radiotherapy by measuring back pressure changes at multiple positions on CT couch. The 12-channel pressure sensor is fixed on CT couch to obtain patient's back pressure signal. The 12-channel signal is transmitted to a PC at a sampling rate of 50 Hz after a signal conditioning circuit and an analog-digital converter. The amplitude of pressure changes is characterized to select the optimal channel. This system is validated by comparing with the respiratory signal collected synchronously with a real-time position management (RPM) system on 10 healthy volunteers. The correlation coefficient between the signals is 0.82 ± 0.09 (standard deviation) and the time shift is 0.32 ± 0.15 second. We conclude that the back pressure signal acquired by the proposed system has the potential to replace the clinical RPM system for respiratory signal detection in 4D-CT data acquisition.

Automated identification and reduction of artifacts in cine four-dimensional computed tomography (4DCT) images using respiratory motion model

PURPOSE

Four-dimensional computed tomography (4DCT) images are often marred by artifacts that substantially degrade image quality and confound image interpretation. Human observation remains the standard method of 4DCT artifact evaluation, which is time-consuming and subjective. We develop a method to automatically identify and reduce artifacts in cine 4DCT images.

METHODS

We proposed an algorithm that consisted of two main stages: deformable image registration and respiratory motion simulation. Specifically, each 4DCT phase image was registered to the breath-holding CT image using the block-matching method, with erroneous spatial matches removed by the least median of squares filter and the full displacement vector field generated by the moving least squares interpolation. The lung's respiratory motion trajectory was then recovered from the displacement vector field using the parameterized polynomial function, with fitting parameters estimated by combinatorial optimization. In this way, artifacts were located according to deviations between image points and their motion trajectories and further corrected based on position prediction.

RESULTS

The mean spatial error (standard deviation) was 1.00 (0.85) mm after registration as opposed to 6.96 (4.61) mm before registration. In addition, we took human observation conducted by medical experts as the gold standard. The average sensitivity, specificity, and accuracy of the proposed method in artifact identification were 0.97, 0.84, and 0.89, respectively.

CONCLUSIONS

The proposed method identified and reduced artifacts accurately and automatically, providing an alternative way to analyze 4DCT image quality and to correct problematic images for radiation therapy.

Model-guided respiratory organ motion prediction of the liver from 2D ultrasound

With the availability of new and more accurate tumour treatment modalities such as high-intensity focused ultrasound or proton therapy, accurate target location prediction has become a key issue. Various approaches for diverse application scenarios have been proposed over the last decade. Whereas external surrogate markers such as a breathing belt work to some extent, knowledge about the internal motion of the organs inherently provides more accurate results. In this paper, we combine a population-based statistical motion model and information from 2d ultrasound sequences in order to predict the respiratory motion of the right liver lobe. For this, the motion model is fitted to a 3d exhalation breath-hold scan of the liver acquired before prediction. Anatomical landmarks tracked in the ultrasound images together with the model are then used to reconstruct the complete organ position over time. The prediction is both spatial and temporal, can be computed in real-time and is evaluated on ground truth over long time scales (5.5 min). The method is quantitatively validated on eight volunteers where the ultrasound images are synchronously acquired with 4D-MRI, which provides ground-truth motion. With an average spatial prediction accuracy of 2.4 mm, we can predict tumour locations within clinically acceptable margins.

A GPU-based framework for modeling real-time 3D lung tumor conformal dosimetry with subject-specific lung tumor motion

In this paper, we present a graphics processing unit (GPU)-based simulation framework to calculate the delivered dose to a 3D moving lung tumor and its surrounding normal tissues, which are undergoing subject-specific lung deformations. The GPU-based simulation framework models the motion of the 3D volumetric lung tumor and its surrounding tissues, simulates the dose delivery using the dose extracted from a treatment plan using Pinnacle Treatment Planning System, Phillips, for one of the 3DCTs of the 4DCT and predicts the amount and location of radiation doses deposited inside the lung. The 4DCT lung datasets were registered with each other using a modified optical flow algorithm. The motion of the tumor and the motion of the surrounding tissues were simulated by measuring the changes in lung volume during the radiotherapy treatment using spirometry. The real-time dose delivered to the tumor for each beam is generated by summing the dose delivered to the target volume at each increase in lung volume during the beam delivery time period. The simulation results showed the real-time capability of the framework at 20 discrete tumor motion steps per breath, which is higher than the number of 4DCT steps (approximately 12) reconstructed during multiple breathing cycles.

Modeling simulation and visualization of conformal 3D lung tumor dosimetry

Lung tumors move during breathing depending on the patient's patho-physiological condition and orientation, thereby compromising the accurate deposition of the radiation dose during radiotherapy. In this paper, we present and validate a computer-based simulation framework to calculate the delivered dose to a 3D moving tumor and its surrounding normal tissues. The computer-based simulation framework models a 3D volumetric lung tumor and its surrounding tissues, simulates the tumor motion during a simulated dose delivery both as a self-reproducible motion and a random motion using the dose extracted from a treatment plan, and predicts the amount and location of radiation doses deposited. A radiation treatment plan of a small lung tumor (1-3 cm diameter) was developed in a commercial planning system (iPlan software, BrainLab, Munich, Germany) to simulate the radiation dose delivered. The dose for each radiation field was extracted from the software. The tumor motion was simulated for varying values of its rate, amplitude and direction within a single breath as well as from one breath to another. Such variations represent the variations in tumor motion induced by breathing variations. During the simulation of dose delivery, the dose on the target was summed to generate the real-time dose to the tumor for each beam independently. The simulation results show that the dose accumulated on the tumor varies significantly with both the tumor size and the tumor's motion rate, amplitude and direction. For a given tumor motion rate, amplitude and direction, the smaller the tumor size the smaller is the percentage of the radiation dose accumulated. The simulation results are validated by comparing the center plane of the 3D tumor with 2D film dosimetry measurements using a programmable 4D motion phantom moving in a self-reproducible pattern. The results also show the real-time capability of the framework at 40 discrete tumor motion steps per breath, which is higher than the number of four-dimensional computed tomography (CT) steps (approximately 20) during a single breath. The real-time capability enables the framework to be coupled with real-time tumor monitoring systems such as implanted fiducials for computing the dose delivered in real time during the treatment.

Real-time motion-adaptive-optimization (MAO) in TomoTherapy

IMRT delivery follows a planned leaf sequence, which is optimized before treatment delivery. However, it is hard to model real-time variations, such as respiration, in the planning procedure. In this paper, we propose a negative feedback system of IMRT delivery that incorporates real-time optimization to account for intra-fraction motion. Specifically, we developed a feasible workflow of real-time motion-adaptive-optimization (MAO) for TomoTherapy delivery. TomoTherapy delivery is characterized by thousands of projections with a fast projection rate and ultra-fast binary leaf motion. The technique of MAO-guided delivery calculates (i) the motion-encoded dose that has been delivered up to any given projection during the delivery and (ii) the future dose that will be delivered based on the estimated motion probability and future fluence map. These two pieces of information are then used to optimize the leaf open time of the upcoming projection right before its delivery. It consists of several real-time procedures, including 'motion detection and prediction', 'delivered dose accumulation', 'future dose estimation' and 'projection optimization'. Real-time MAO requires that all procedures are executed in time less than the duration of a projection. We implemented and tested this technique using a TomoTherapy research system. The MAO calculation took about 100 ms per projection. We calculated and compared MAO-guided delivery with two other types of delivery, motion-without-compensation delivery (MD) and static delivery (SD), using simulated 1D cases, real TomoTherapy plans and the motion traces from clinical lung and prostate patients. The results showed that the proposed technique effectively compensated for motion errors of all test cases. Dose distributions and DVHs of MAO-guided delivery approached those of SD, for regular and irregular respiration with a peak-to-peak amplitude of 3 cm, and for medium and large prostate motions. The results conceptually proved that the proposed method is applicable for real-time motion compensation in TomoTherapy delivery. Extension of the method to real-time adaptive radiation therapy (ART) that compensates for all kinds of delivery errors was proposed. Further validation and clinical implementation is underway.

Real-time simulation of 4D lung tumor radiotherapy using a breathing model

In this paper, we present a real-time simulation and visualization framework that models a deformable surface lung model with tumor, simulates the tumor motion and predicts the amount of radiation doses that would be deposited in the moving lung tumor during the actual delivery of radiation. The model takes as input a subject-specific 4D Computed Tomography (4D CT) of lungs and computes a deformable lung surface model by estimating the deformation properties of the surface model using an inverse dynamics approach. Once computed, the deformable model is used to simulate and visualize lung tumor motion that would occur during radiation therapy accounting for variations in the breathing pattern. A radiation treatment plan for the lung tumor is developed using one of the 4D CT phases. During the simulation of radiation delivery, the dose on the lung tumor is computed for each beam independently.

Computational Challenges for Image-Guided Radiation Therapy: Framework and Current Research

It is arguable that the imaging and delivery hardware necessary for delivering real-time adaptive image-guided radiotherapy is available on high-end linear accelerators. Robust and computationally efficient software is the limiting factor in achieving highly accurate and precise radiotherapy to the constantly changing anatomy of a cancer patient. The limitations are not caused by the availability of algorithms but rather issues of reliability, integration, and calculation time. However, each of the software components is an active area of research and development at academic and commercial centers. This article describes the software solutions in 4 broad areas: deformable image registration, adaptive replanning, real-time image guidance, and dose calculation and accumulation. Given the pace of technological advancement, the integration of these software solutions to develop real-time adaptive image-guided radiotherapy and the associated challenges they bring will be implemented to varying degrees by all major manufacturers over the coming years.

Geometric uncertainty of 2D projection imaging in monitoring 3D tumor motion

The purpose of this study was to investigate the accuracy of two-dimensional (2D) projection imaging methods in three-dimensional (3D) tumor motion monitoring. Many commercial linear accelerator types have projection imaging capabilities, and tumor motion monitoring is useful for motion inclusive, respiratory gated or tumor tracking strategies. Since 2D projection imaging is limited in its ability to resolve the motion along the imaging beam axis, there is unresolved motion when monitoring 3D tumor motion. From the 3D tumor motion data of 160 treatment fractions for 46 thoracic and abdominal cancer patients, the unresolved motion due to the geometric limitation of 2D projection imaging was calculated as displacement in the imaging beam axis for different beam angles and time intervals. The geometric uncertainty to monitor 3D motion caused by the unresolved motion of 2D imaging was quantified using the root-mean-square (rms) metric. Geometric uncertainty showed interfractional and intrafractional variation. Patient-to-patient variation was much more significant than variation for different time intervals. For the patient cohort studied, as the time intervals increase, the rms, minimum and maximum values of the rms uncertainty show decreasing tendencies for the lung patients but increasing for the liver and retroperitoneal patients, which could be attributed to patient relaxation. Geometric uncertainty was smaller for coplanar treatments than non-coplanar treatments, as superior-inferior (SI) tumor motion, the predominant motion from patient respiration, could be always resolved for coplanar treatments. Overall rms of the rms uncertainty was 0.13 cm for all treatment fractions and 0.18 cm for the treatment fractions whose average breathing peak-trough ranges were more than 0.5 cm. The geometric uncertainty for 2D imaging varies depending on the tumor site, tumor motion range, time interval and beam angle as well as between patients, between fractions and within a fraction.