Optimization of Brain Tumor MR Image Classification Accuracy Using Optimal Threshold, PCA and Training ANFIS with Different Repetitions.

BACKGROUND: One of the leading causes of death is brain tumors. Accurate tumor classification leads to appropriate decision making and providing the most efficient treatment to the patients. This study aims to optimize brain tumor MR images classification accuracy using optimal threshold, PCA and training Adaptive Neuro Fuzzy Inference System (ANFIS) with different repetitions. MATERIAL AND METHODS: The procedure used in this study consists of five steps: (1) T1, T2 weighted images collection, (2) tumor separation with different threshold levels, (3) feature extraction, (4) presence and absence of feature reduction applying principal component analysis (PCA) and (5) ANFIS classification with 0, 20 and 200 training repetitions. RESULTS: ANFIS accuracy was 40%, 80% and 97% for all features and 97%, 98.5% and 100% for the 6 selected features by PCA in 0, 20 and 200 training repetitions, respectively. CONCLUSION: The findings of the present study demonstrated that accuracy can be raised up to 100% by using an optimized threshold method, PCA and increasing training repetitions.

Evaluation of the Effect of Source Geometry on the Output of Miniature X-ray Tube for Electronic Brachytherapy through Simulation.

OBJECTIVE: The use of miniature X-ray source in electronic brachytherapy is on the rise so there is an urgent need to acquire more knowledge on X-ray spectrum production and distribution by a dose. The aim of this research was to investigate the influence of target thickness and geometry at the source of miniature X-ray tube on tube output. METHOD: Five sources were simulated based on problems each with a specific geometric structure and conditions using MCNPX code. Tallies proportional to the output were used to calculate the results for the influence of source geometry on output. RESULTS: The results of this work include the size of the optimal thickness of 5 miniature sources, energy spectrum of the sources per 50 kev and also the axial and transverse dose of simulated sources were calculated based on these thicknesses. The miniature source geometric was affected on the output x-ray tube. CONCLUSION: The result of this study demonstrates that hemispherical-conical, hemispherical and truncated-conical miniature sources were determined as the most suitable tools.

An Analytical-empirical Calculation of Linear Attenuation Coefficient of Megavoltage Photon Beams.

BACKGROUND: In this study, a method for linear attenuation coefficient calculation was introduced. METHODS: Linear attenuation coefficient was calculated with a new method that base on the physics of interaction of photon with matter, mathematical calculation and x-ray spectrum consideration. The calculation was done for Cerrobend as a common radiotherapy modifier and Mercury. RESULTS: The values of calculated linear attenuation coefficient with this new method are in acceptable range. Also, the linear attenuation coefficient decreases slightly as the thickness of attenuating filter (Cerrobend or mercury) increased, so the procedure of linear attenuation coefficient variation is in agreement with other documents. The results showed that the attenuation ability of mercury was about 1.44 times more than Cerrobend. CONCLUSION: The method that was introduced in this study for linear attenuation coefficient calculation is general enough to treat beam modifiers with any shape or material by using the same formalism; however, calculating was made only for mercury and Cerrobend attenuator. On the other hand, it seems that this method is suitable for high energy shields or protector designing.