The result is that patients only receive a high radiation dose where they need it and healthy tissue is preserved.
The problem with IMRT so far, however, is that it becomes increasingly difficult to verify that patients receive the prescribed dose of radiation. "IMRT prescriptions are based on very complex computer simulations, so it is important to validate these simulations by verifying exactly how much radiation is reaching the patient and where it is landing," says Aleksandar Jaksic, INVORAD project coordinator at Ireland's Tyndall National Institute.
INVORAD developed two sensors, a silicon diode and a p-channel metal-oxide semiconductor field-effect transistor (MOSFET), to do just that. "Several features, such as miniature size, response to types of radiation involved in radiotherapy, compatibility with microprocessors that enables real-time read-out and low cost, make these semiconductor sensors eminently suitable for the intended application," says Jaksic.
The diode sensor system is arranged in a series of modules containing 1069 individual diodes that can pick up incoming radiation.
"These diodes need to be very small and while there are commercial packaged diodes out there we needed diodes in bare die form with some novel properties so we developed the diodes ourselves, here at the Tyndall Institute," says Jaksic.
The arrays are extremely accurate and can track radiation at micro-Gray resolution over millimetres of spatial resolution.
These are then linked to a read-out unit and a PC with dedicated software. The read-out unit is based on ASIC (Application Specific Integrated Circuit) and mic roprocessor technologies, and its function is to communicate with, and retrieve data from, the sensor arrays. The PC and software provide system control, connectivity to other parts of an overall radiotherapy system, such as record and verify packages, and patient-specific data storage.
INVORAD also developed a cylindrical 'body phantom'. The 'phantom' is given the prescribed dose and the diode sensors pick up the dose actually delivered. "The two modular 2D diode arrays are placed in orthogonal positions inside the phantom, so we have data in 3D over time," says Jaksic.
If the 'phantom' treatment matches the prescription of the simulator, the patient is given treatment. If not, the treatment plan needs to be corrected. "We created modifications on the diodes and diode arrays, improving their specifications for this project. In fact, every element of the project we worked on received some sort of improvement on current systems," says Jaksic.
Some types of MOSFETs can also detect radiation. In the INVORAD MOSFET-based system these are used in-vivo, mounted in medical catheters in the form of linear arrays, entering the patient through a cavity.
"We're currently testing that device in patients with our clinical partner, the Clatterbridge Centre for Oncology, one of the largest oncology centres in the UK. Of the two devices, the diode system is the most commercially viable. However, the MOSFET system is working and we'll have the results of patients trials in the next few months," says Jaksic.
"We need to further optimise some parameters of the diode sensor system, but from the work we've done so far we know how to solve these remaining issues." Jaksic believes it is worth the wait. "Unlike most projects, this device will go straight to market and our commercial partner, ScandiDos in Uppsala, Sweden, is a start-up created for the manufacture and marketing of the device."
Jaksic is particularly pleased because the new sens or systems will improve treatment verification for a large number of cancer patients.
"The prevailing opinion is that IMRT improves treatment outcomes," says Jaksic. "Crucially, IMRT reduces the side-effects patients often suffer from radiotherapy and improves accuracy of dose delivery, and these are the most important impacts in the treatment of cancer."