Proton radiation therapy is clinically used to deliver highly conformal dose distributions to treatment volumes. However, existing uncertainties in proton range calculations has prevented it from reaching its full potential.These uncertainties mainly arise from target motion, patient setup and dose calculations, which can cause the proton Bragg peak to miss the treatment volume. Such uncertainties are currently managed by applying treatment margins around the target volume, which negatively affects the conformity of the dose distribution and exposes additional healthy tissues/organs along the beam path.
Proton-based imaging systems can eliminate range uncertainties by directly generating a 3D distribution of proton stopping power ratios (SPRs).These SPR values can be incorporated into the treatment planning to accurately calculate proton ranges within tissues. In addition to treatment planning, a proton imager installed inside the proton therapy gantry room can provide 3D imaging at the treatment isocenter. Thus, proton systems can be useful for image-guided proton therapy as well as aid with quick and precise positioning of patients for proton therapy.
Our laboratory is currently developing a first-of-its-kind proton-integrating imaging system design based on a 3D scintillator detector, as seen in the figure. Unlike existing 2D scintillator detector systems, our approach can record a wide distribution of energy-attenuated proton beams exiting the patient from a single proton beam exposure. This design can potentially save time and reduce dose requirements for proton imaging. Our proton imaging system is capable of generating proton radiographs and tomograms.
While modern radiotherapy treatment fields have complex 3D features, the standard measurement tools are point detectors and 2D detector arrays. As a result, verification procedures can be very time consuming, even when measuring the dose at only a few points. This is especially problematic in proton radiation therapy, because of the complex shapes of proton dose distributions. Ideally, 3D dose distributions would be measured prior to treatment for all patients. Unfortunately, this is impractical because of time constraints and the limited availability of suitable detectors.
Large-volume scintillation detectors have the potential for fast, high-resolution, 3D dose measurement of complex radiotherapy fields. Scintillators convert the energy from x-rays and proton beams into visible light, which can then be recorded with a camera and converted into radiation dose. By recording the scintillation light in a large detector from multiple angles, we can reconstruct the 3D shape of the dose distribution.
Liquid scintillator-based detectors have several attractive features for dosimetric measurements. They exhibit high resolution, dose rate independence and a linear dose response. A large volume of liquid scintillator can serve as the measurement medium in addition to the detector, thus eliminating perturbations to the radiation field by the detector. Additionally, liquid scintillator measurements are immediate and do not require readout after the dose delivery like other 3D dosimeters. The light emission from the LS occurs within nanoseconds, making it possible to perform repeated measurements very quickly.
We have developed LET-dependence and optical artifact correction techniques, 3D reconstruction algorithms and are working on quality assurance procedures for scanned proton beams. Our long-term goal is to reduce radiation treatment errors and improve dose verification accuracy by developing a new fast, reusable 3D detector for patient treatment verification.
The projection image of a proton beam using our liquid scintillator detector. Characteristics of the proton beam, such as the Bragg peak, can be measured quantitatively.
The technology involved in delivering radiation therapy to patients is enormously complex and prone to human error. Therefore, safety checks were introduced to minimize errors and were generally very successful in catching errors. However, a small number of high profile misadministrations resulting in severe patient injuries demonstrates that errors can still occur. In vivo dosimetry, or measuring the dose received by patients as they are receiving radiation treatment, is an effective method to catch any errors by verifying delivery at the point of treatment.
In vivo dosimetry is not practiced routinely because of the high cost of detectors suitable for in vivo dosimetry and the technical expertise required to use them. However, plastic scintillation detectors are relatively inexpensive and can be reused extensively (some detectors such as MOSFETS experience radiation induced damage and must be replaced regularly). Plastic scintillation detectors are also free from response altering effects due to factors other than the deposited dose – for example: radiation energy, dose rate and detector orientation – making them straightforward and accurate detectors for in vivo dosimetry.
We are currently developing in vivo dosimetry systems based on plastic scintillation detectors to establish the usefulness and ease of implementation of such systems. Our long term goal is to see increasingly common use of plastic scintillation detectors for routine clinical in vivo dosimetry, providing a higher standard of care to patients.
A CT image of the plastic scintillation detectors (PSD) in vivo used for dose verification in a prostate radiation treatment.
Real time measured in vivo dose measured with a pair of plastic scintillation detectors (PSD), compared to dose calculated by the treatment planning system.
During brachytherapy treatments, one or multiple radioactive seeds are inserted within the tumor to eradicate it. Even if this treatment is very conformal, it is still important to limit the dose of radiation delivered to the healthy tissues. Plastic scintillation detectors show a great promise to measure in vivo dose delivered in real-time to the tumor and its surroundings. Their sub-millimeter size, fast response (~ns), water-equivalency and small dependence to temperature give them an advantage over other detectors. Work was previously done by our group for Ir-192 high-dose rate brachytherapy, leading to multiple publications.
An on-going project focuses on extending the application of PSDs to low-dose rate brachytherapy. Radioactive sources of low energy are used (Iodine-125 and Palladium-103), which possess polyenergetic emission spectra with average energies of 28 keV and 21 keV, respectively. The overall objective of this research project is to account for the energy-dependence and scintillation quenching at these energies and develop a clinically reliable and accurate plastic scintillation detector system.
Accurate dose delivery of radiation therapy can improve treatment outcome and reduce side-effects. Unfortunately, respiratory motion can create spatial uncertainty that can lead to insufficient dose to the tumor or/and overdose to normal tissue. 4D (time resolved volumetric) imaging techniques acquire images of a patient during a complete breathing cycle. As a result, one can account for the respiratory motion in the planning process of radiation treatment.
The major advantage of MR imaging over other medical imaging modalities is its superior soft tissue contrast, which makes segmentation between tumor and normal tissue much easier. Therefore, development of 4D MRI techniques can be extremely beneficial to radiation therapy treatment planning.
We are currently interested in the development of 4D MRI protocols as well as new algorithms to sort the acquired MR images into a 4D volume.
Unlike x-rays beams, which pass through patients, proton beams have a finite range. They also deliver a burst of dose just before stopping. By appropriately selecting the initial energy, proton beams can be tuned to deliver the final burst, called the Bragg peak, in the tumor allowing for a highly conformal dose distribution. Unfortunately, there is a large uncertainty in the beam range in patients due to the tissue heterogeneities. Therefore, proton treatment plans must include large margins around the tumor. The plans must also avoid placing the Bragg peak directly in front of critical organs. Therefore, in order to maximize the benefits provided by the Bragg peak, we need to reduce the range uncertainty.
We are investigating ways to verify the beam range in-vivo by producing images from secondary radiation emitted from tissue during treatment. In collaboration with Dr. Jerimy Polf, Assistant Professor of Medical Physics at the University of Maryland, we are designing a multistage Compton scatter camera, gamma-ray detection system. This system works by reconstructing the likely origins of thousands of gammas that each scatter in multiple times in the detector. The proton beam range can then be verified from images produced of the gamma emission origins.
Both the hardware and the imaging algorithms for the detection system are active areas of research. We are working to improve the imaging algorithms through the application of Monte Carlo methods, machine learning and high performance computing. Our long term goal is to enable real-time monitoring of the beam range from the clinic’s treatment control room.