Real Beam Data Research Articles

Students training on virtual medical linear accelerator

At the Czech Technical University in Prague, we have installed virtual software environment simulating a radiotherapy treatment room with a medical linear accelerator (software VERT and VERT Physics by Vertual Ltd., United Kingdom). It is possible to simulate patient treatment and physics QA (quality assurance) measurements in virtual reality. This is beneficial for our students and teachers because there is no radiation present while doing the exercises and it is accessible at any time, unlike medical linear accelerators at hospitals. For our department, we have chosen the TrueBeam model by Varian Medical Systems, USA. The linear accelerator can be operated by the user in the same way as at the hospital, including the hardware pendant. There is a selection of patient treatment plans, CT images and RT structures, all of which can be visualised inside the treatment room, also during beam on, simulating the situation inside the patient while irradiating the tumour. Any DICOM files, even from another manufacturer’s linac (linear accelerator) or treatment planning system, can be imported into VERT. Apart from this, there is also an imaging interface and cone-beam CT interface showing options for patient setup. Another part of the software is a physics module with various phantoms and detectors for dosimetry as well as geometry tests. For example, a large water tank can be used for relative dosimetry of photon beams (percent depth doses and profiles), a smaller water phantom can be used for reference dosimetry or chamber cross calibration according to the TRS 398 (1) or IPEM (2) protocol. The results of measurements are based on look-up tables coming from real beam data, so dependence on the depth of measurement, field size, temperature and pressure and other parameters can be demonstrated. Moreover, it is possible to introduce many types of errors into the linac, such as jaw position, laser, or collimator miscalibration, and show how these errors propagate to patient irradiation and how they can be detected by QA tools. This paper is a short report on how this system has been used at our university and how it might be used in modern learning, pointing out its advantages and drawbacks. At the 16th Workshop on European Collaboration in Higher Education on Radiological and Nuclear Engineering and Radiation Protection (CHERNE) it was proposed that this virtual system could be used for an international student training course on medical physics in radiotherapy within the CHERNE network. This course might contain theoretical lectures with demonstrations on the virtual linac, practical exercises on the virtual linac and visits to hospitals (a photon radiotherapy centre, the Thomayer University Hospital in particular, and a proton therapy centre).

Reconstruction of transverse phase space from transverse feedback data for real time extraction of vital LHC machine parameters

The LHC transverse feedback system (ADT) provides bunch by bunch, turn by turn, normalized and digitized beam position signals from four pick-ups per plane and for each beam. Together with already existing powerful computer-based observation systems, this data can be used to reconstruct in real-time the transverse phase space coordinates of the centre-of-charges, for each individual bunch. Such information is extremely valuable for machine operation, or transverse instability diagnostics.This paper aims on discussing and evaluating methods of combining four position signals for such analysis in the presence of noise and with active transverse feedback. Comparisons are made based on the extraction of vital parameters like the fractional tune or transverse activity. Analytical and numerical results are further benchmarked against real beam data.

Learning to decompose the modes in few-mode fibers with deep convolutional neural network.

We introduce a deep-learning technique to perform complete mode decomposition for few-mode optical fibers for the first time. Our goal is to learn a fast and accurate mapping from near-field beam patterns to the complete mode coefficients, including both modal amplitudes and phases. We train the convolutional neural network with simulated beam patterns and evaluate the network on both the simulated beam data and the real beam data. In simulated beam data testing, the correlation between the reconstructed and the ideal beam patterns can achieve 0.9993 and 0.995 for 3-mode case and 5-mode case, respectively. While in the real 3-mode beam data testing, the average correlation is 0.9912 and the mode decomposition can be potentially performed at 33 Hz frequency on a graphic processing unit, indicating real-time processing ability. The quantitative evaluations demonstrate the superiority of our deep learning-based approach.

Scale factor correction for Gaussian beam truncation in second moment beam radius measurements

Charged-couple devices (CCD) and complementary metal oxide semiconductor (CMOS) image sensors, in conjunction with the second moment radius analysis method, are effective tools for determining the radius of a laser beam. However, the second moment method heavily weights sensor noise, which must be dealt with using a thresholding algorithm and a software aperture. While these noise reduction methods lower the random error due to noise, they simultaneously generate systematic error by truncating the Gaussian beam’s edges. A scale factor that is invariant to beam ellipticity and corrects for the truncation of the Gaussian beam due to thresholding and the software aperture has been derived. In particular, simulations showed an order of magnitude reduction in measured beam radius error when using the scale factor—irrespective of beam ellipticity—and further testing with real beam data demonstrated that radii corrected by the scale factor are independent of the noise reduction parameters. Thus, through use of the scale factor, the accuracy of beam radius measurements made with a CCD or CMOS sensor and the second moment are significantly improved.

Numerical modeling of spatial coherence using the elementary function method

The elementary function method is an approximate method for propagation calculations in spatially, partially coherent light in two dimensions. In this paper, we present the numerical application of this method to a 248 nm UV excimer laser source. We present experimental results of the measurement of the degree of spatial coherence and the beam profile of this source. The elementary function method is then applied to the real beam data and used to simulate the effects of imaging an opaque edge with a source of varying degrees of spatial coherence. The effect of spatial coherence on beam homogenization is also presented.

Superresolution of Coherent Sources in Real-Beam Data

In this work we study the unique problems associated with resolving the direction of arrival (DOA) of coherent signals separated by less than an antenna beamwidth when the data are collected in the beamspace domain with, for example, electronically or holographically scanned antennas. We also propose a technique that is able to resolve these coherent signals. The technique is based on interpolation of the data measured by an element-space virtual array. Although the data are collected in the beamspace domain, the coherence structure can be broken by interpolating multiple shifted element-space virtual arrays. The efficacy of this technique depends on a fundamental tradeoff that arises due to a nonuniform signal-to-noise ratio (SNR) profile across the elements of the virtual array. This profile is due to the structure imposed by the specific beam pattern of the antenna. In addition to describing our technique and studying the SNR profile tradeoff, we also incorporate a strategy for improving performance through a subswath technique that improves covergence of covariance estimates.

A new optimised quadrupole pick-up design using magnetic coupling

The idea of quadrupole pick-ups, sensitive to beam size, originated several decades back. Such pick-ups measure the quantity σ x 2− σ y 2, where σ x and σ y are the horizontal and vertical r.m.s. beam sizes. Thus, a quadrupole pick-up is a candidate for non-invasive study of processes such as coherent beam width oscillations due to injection mismatch. Up to now, quadrupole pick-ups have been produced essentially by enhancing the electronics of normal position pick-ups to produce the so-called quadrupole signal, with little or no effort being put into the design of the pick-up itself. In developing a quadrupole pick-up for the CERN PS, however, it has been found necessary to optimise the pick-up design. The result is a somewhat unconventional pick-up, where magnetic coupling is employed to suppress the otherwise very strong, and undesired, common mode-signal. In this paper, the basic design idea and the final optimised design are presented, together with simulations, test bench measurements and real beam data.