Machine Learning in Magnetic Resonance Images of Glioblastoma: A Review.

BACKGROUND: The purpose of this work was to identify which Glioblastoma (GBM) problems can be handled by Magnetic Resonance Imaging (MRI) and Machine Learning (ML) techniques. Results, limitations, and trends through a review of the scientific literature in the last 5 years were performed. Google Scholar, PubMed, Elsevier databases, and forward and backward citations were used for searching articles applying ML techniques in GBM. The 50 most relevant papers fulfilling the selection criteria were deeply analyzed. The PRISMA statement was followed to structure our report. METHODS: A partial taxonomy of the GBM problems tackled with ML methods was formulated with 15 subcategories grouped into four categories: extraction of characteristics from tumoral regions, differentiation, characterization, and problems based on genetics. RESULTS: The dominant techniques in solving these problems are: Radiomics for feature extraction, Least Absolute Shrinkage and Selection Operator for feature selection, Support Vector Machines and Random Forest for classification, and Convolutional Neural Networks for characterization. A noticeable trend is that the application of Deep Learning on GBM problems is growing exponentially. The main limitations of ML methods are their interpretability and generalization. CONCLUSION: The diagnosis, treatment, and characterization of GBM have advanced with the aid of ML methods and MRI data, and this improvement is expected to continue. ML methods are effective in solving GBM-related problems with different precisions, Overall Survival being the hardest problem to solve with accuracies ranging from 57%-71%, and GBM differentiation the one with the highest accuracy ranging from 80%-97%.

Characterization of the radiation beam of a tomotherapy equipment with MCNP.

This work aims to model and characterize the radiation beam of one Accuray tomotherapy equipment using the Monte Carlo Code MCNP5 (Monte Carlo N-Particle). This tomotherapy equipment is used for delivering high doses of radiation in tumor regions to kill cancer cells and shrink the tumor during radiation therapy of cancer patients, however, the radiation can damage surrounding areas and nearby organs at risk (OAR) if the radiation field is not well delimited. In particular, intensity-modulated radiotherapy treatments (IMRT) with tomotherapy equipment offer great benefits to patients allowing treatment of tumor regions without affecting surrounding areas and OAR. Nowadays, it is well known that a correct simulation of transport of radiation in tomotherapy equipment facilitates considerably the estimation of ideal doses in the tumor, surrounding regions, and OAR. For that reason, in this work, we simulated the geometry of the 6 MV ACCURAY Tomotherapy equipment of the CECAN using the MCNP5. The model includes a TomoLINAC consisting of an electron source that emits Gaussian distribution particles with an average energy of 5.7 MeV and width of 0.3 MeV. The emitted particles impact the tungsten target and pass through primary collimators and jaws that define the irradiation field in the isocenter. To validate the geometry and radiation transport in the TomoLINAC the curves of depth dose percentage (PDD) estimated by simulation and the curves measured experimentally were tuned. In the same way, the simulated transverse and longitudinal profiles were compared with the experimental results. In addition, a comparison between the qualities of the radiation beam characterized with MCNP and measured experimentally in CECAN showed a deviation of 1%. For the simulations, cylindrical detectors located inside a water phantom were considered and it was employed the tally *F8. A good agreement was observed between the PDD's curves obtained from the simulation and those measured experimentally for a field of 5 × 10 cm2 in the isocenter and SSD (distance from the source to the surface) of 85 cm. Also, the comparison between the simulated and experimental transverse profiles obtained at 1.5 cm, 10 cm and 15 cm depth with a radiation field of 5 × 40 cm2 showed very good agreement. The longitudinal profiles were estimated with the same depths as the transverse ones, but for each of them, the openings of the jaws were 5.0 cm, 2.5 cm and 1.0 cm in the longitudinal direction, which corresponds to the direction in which the patient's table moves. The comparison between the simulated and experimental longitudinal profiles showed good concordance too. Once the radiation beam of the ACCURAY tomotherapy equipment had been characterized, experimental dose measurements were made using a Cheese phantom and two A1SL ionization chambers. These results obtained experimentally were compared with those estimated with MCNP for a field of 5 × 40 cm2 at the isocenter and SAD of 85 cm and, it was concluded that both results were similar considering the regions of uncertainty. Finally, we must highlight that the modeling and characterization of the radiation beam of CECAN's ACCURAY tomotherapy equipment can be a key tool for dose estimations in different cancer treatment plans and future research.

Fluorescent organic particle doped polymer-based gel dosimeter for neutron detection.

The purpose of this work is to develop a material capable of detecting neutrons produced by photodisintegration in a linear accelerator for its medical use. In this study, we have developed a gel-like material doped with fluorescent organic particles. PPO at 1 wt% is used as primary dopant and POPOP as secondary one at 0.03 wt%. A set of four samples is produced, with boric acid concentrations of 0, 400, 800 and 1200 ppm. The viscoelastic properties of the material are characterized with rheological measurements, finding a gel-like behavior, i.e., a material that can keep its original shape if no stresses are applied, but can also be deformed by applying a moderate shear rate. Furthermore, the material was irradiated with gamma, electron, and neutron emission sources from 137Cs, 22Na, 60Co, 210Po, 90Sr and 241AmBe, and its response was measured in two different experimental settings, in two different institutions, for comparative purposes. From these measurements, one can clearly establish that the new material detects neutrons, electrons, and gammas within the MeV regions and below. Thus, our findings show that the developed material and its properties make it a promising technology for its use in a neutron detector.

Structural Properties and Luminescence Characteristics of Eu3+ Doped Lithium Triborate (LiB3O5) Phosphors.

Lithium triborate LiB3O5:Eu has been prepared by precipitation assisted high-temperature solid state (PAS) reaction method using lithium hydroxide (LiOH), boric acid (H3BO3) and europium chloride (EuCl3). The as prepared powders were calcined at various temperatures. As prepared and calcined powders were characterized by X-ray diffraction (XRD), Scanning Electron Microscopy (SEM), Energy Dispersive X-ray Spectrometry (EDS), UV-Vis spectrometry and Photoluminescence (PL). The XRD profiles of the powder products revealed the presence of crystalline phase at 650 °C. Increasing bandgap with increasing dopant concentration was studied by UV-Vis spectra. The effect of dopant concentration on photoluminescence properties of LiB3O5 was also studied. The optimized Eu3+ dopant concentration for potential luminescent LiB3O5 phosphor was 0.25 mol%.

Thermoluminescence from Cu Doped Lithium Tetraborate Irradiated with X-ray and γ Using 137Cs Radioactive Source.

Lithium tetraborate (Li2B4O7) pellets prepared by using water/solution assisted method were synthesized and characterized. Copper was used as doping material in order to enhance the Li2B4O7 thermoluminescent properties. For synthesis heating temperature parameters were defined at 750 °C for 2 hr, followed by 150 °C for another 2 hr. The materials were produced at five different Cu concentrations: 0.02, 0.04, 0.06, 0.08, and 0.1 wt%. The luminescent and morphological characterizations were performed by X-ray diffraction (XRD), Scanning electron microscope (SEM), Photoluminescence (PL), and Ultraviolet-visible spectroscopy (UV-Vis). XRD and SEM analysis of intrinsic and doped materials confirmed the obtained Li2B4O7 structure and show its morphology. XRD patterns of the Li2B4O7 matched a tetragonal crystal structure. Crystals of Li2B4O7 of an average size of 50 nm were obtained. The presence of the copper dopant was confirmed in crystals of Li2B4O7:Cu by SEM-EDS (energy dispersive spectroscopy X-ray). The emission spectrum of Cu doped Li2B4O7 showed a prominent peak at 367 nm, while the main UV-Vis absorption was observed from 240 nm to 300 nm due to Cu+ ion 3d10 → 3d9 4s transitions. The thermoluminescent (TL) response was studied for both γ radiation and X-ray. A 661.7 keV γ radiation using a 137Cs source at doses of 50, 100, 200, 300, 400 and 500 mGy was applied to Li2B4O7:Cu (0.1 wt%) pellets. An X-ray source was used at doses of 600, 800 and 1000 mGy to irradiate pellets of Li2B4O7:Cu (0.02, 0.04, 0.06, 0.08 and 0.1 wt%). A linear TL response was observed for both X-ray and γ radiation. The kinetic parameters were calculated using the peak shape method for the Li2B4O7:Cu (0.1 wt%).