[en] In this thesis, we solve inverse radiation transport problems by an Artificial Neural Network(ANN) approach. ANNs have many interesting properties such as nonlinear, parallel, and distributed processing. Some of the promising applications of ANNs are optimization, image and signal processing, system control, etc. In some optimization problems, Hopfield Neural Network(HNN) which has one-layered and fully interconnected neurons with feed-back topology showed that it worked well with acceptable fault tolerance and efficiency. The identification of radioactive source in a medium with a limited number of external detectors is treated as an inverse radiation transport problem in this work. This kind of inverse problem is usually ill-posed and severely under-determined; however, its applications are very useful in many fields including medical diagnosis and nondestructive assay of nuclear materials. Therefore, it is desired to develop efficient and robust solution algorithms. Firstly, we study a representative ANN model which has learning ability and fault tolerance, i.e., feed-forward neural network. It has an error backpropagation learning algorithm processed by reducing error in learning patterns that are usually results of test or calculation. Although it has enough fault tolerance and efficiency, a major obstacle is 'curse of dimensionality'--required number of learning patterns and learning time increase exponentially proportional to the problem size. Therefore, in this thesis, this type of ANN is used as benchmarking the reliability of the solution. Secondly, another approach for solving inverse problems, a modified version of HNN is proposed. When diagonal elements of the interconnection matrix are not zero, HNN may become unstable. However, most problems including this identification problem contain non-zero diagonal elements when programmed on neural networks. According to Soulie et al., discrete random iterations could produce the stable minimum state of an associative memory. We modify the conventional HNN into a new HNN which has random delayed updating intervals in order to alleviate the above unstable phenomenon. So we shall call it 'Delayed Hopfield-Like Neural Network(DHNN).' We tested DHNN under various noisy environments to verify its efficiency and noise robustness, and compared with previous works in this category. It proved that our approach is more robust and efficient than previous ones in a number of tests. Although the probability of successful identification decreases as the noise level increases, the successful identification rate is still acceptable. Further, we investigated and applied DHNN to some potential applications, such as medical imaging (tumor or cancer detection) and nuclear waste assay, and showed that these problems could be solved by our approach. As a concluding remark, we may say that DHNN can solve ill-posed problems with reasonable efficiency and robustness, including the problem of identification of the radioactive source in a medium

[en] It is important to estimate the radioactivities induced by the patient-specific devices installed near the patient skin because it was directly expose the patient during the treatment. This study evaluated the radioactivities induced by the brass collimator and PMMA range compensator in the proton beam, and identified the radionuclide produced by the nuclear interactions of the proton beam with the patient specific devices. When the radiation dose was evaluated at the highest energy region, the exposure of patients by the radioactivities in the patient-specific devices was usually very low when compared to the dose limit in Korea. PMMA, which is mainly composed of oxygen and carbon, is used to fabricate the range compensators. The aperture of the used brass, which is mainly composed of zinc and copper, determines the boundary between the tumor and normal tissue in the body of the patient. Furthermore, these materials are widely used for phantoms and apertures. The radioactivities induced in brass and PMMA after the irradiation of proton beam was evaluated. The radionuclides in brass and PMMA produced by protons and neutrons were the major cause of radiation exposure for patients in proton therapy. In both case of PMMA and brass, 1 hour later their activities reach about 2% of the initial values. The amount of radiation dose is small due to the short treatment time, and the radiation exposure of patients is negligible when compared to the total dose to the patient

[en] Since the 1970s, aircraft crash accidents have been considered as one of the severest external events that should be evaluated for license application of nuclear reactors. After the 9.11 terrorist attacks, many countries have performed safety assessment against intentional or targeted aircraft crashes into nuclear related facilities. In some countries, assessment against targeted aircraft crash was enforced by regulation and considered an important task for license approval. Safety assessment against aircraft crash is a technically difficult task and many countries manage R&D programs to improve its reliability. In this paper, regulations of many countries regarding safety assessment against aircraft crash are summarized, separating regulations for accident aircraft crash and those for targeted aircraft crash. Research performed in various countries on safety assessment of nuclear facility against aircraft crash are summarized, with a focus on spent nuclear fuel dry storage facilities

[en] The Monte Carlo simulation code CEARPPU has been developed and updated to provide pulse pile-up correction spectra for high counting rate cases. For neutron activation analysis, CEARPPU correction spectra were used in library least-squares method to give better isotopic activity results than the convention library least-squares fitting with uncorrected spectra

No abstract available

No abstract available

[en] AIM: To evaluate the pattern and site of involvement in neuro-Behcet's disease (NBD). MATERIALS AND METHODS: Twenty-one patients with NBD were evaluated. Using 1.5T magnetic resonance imaging (MRI), T1-weighted axial and sagittal images, gadolinium enhanced axial and coronal images and T2-weighted axial images were obtained. RESULTS: The brainstem, basal ganglia, cerebral white matter, internal capsule, thalamus and spinal cord were involved in eighteen, nine, nine, seven, six and two patients, respectively. In nine patients with cerebral white matter involvement, four had subcortical involvement and three had periventricular involvement, in addition to two patients with focal deep white matter lesions. Among the brainstem lesions, pons involvement was seen in fourteen patients, all had ventrally located lesions, and nine had tegmental involvement. Midbrain involvement was seen in fourteen patients; the cerebral peduncle was involved in 11 of these. Five patients had brainstem atrophy: two cases were demonstrated at initial MRI, the other three cases were seen on follow-up MRI. Pyramidal signs, the most common neurological signs, were demonstrated in fourteen patients. Follow-up MRI was obtained 10 days to 20 months after the initial MRI in eight cases; all showed changes in size, shape and site of involvement. After gadolinium enhancement, thirteen patients demonstrated mottled non-confluent enhancement in the brainstem (eight patients), posterior limb of the internal capsule (three patients), pachymeninges (two patients) and spinal cord (two patients). CONCLUSION: NBD manifests a reversible course, but chronic NBD may result in brainstem atrophy. Characteristic involvement along the corticospinal tract is well correlated with neurological signs. Lee, S. H. et al. (2001)

[en] The spent fuel rods are replaced by beam elements with effective material properties calculated from the nproperties of cladding and fuel pellets. However, a well agreed procedure for the model degeneration has not been developed yet and the validity of the effective material properties have not been discussed in depth. In some approach, the stiffness of the pellets is ignored while their masses are lumped into cladding. In others, The average values of Young's modulus are used while no considerations are made on the plastic properties of cladding. In this work, a procedure for material property calibration for spent nuclear fuel rods is proposed based on the static analyses and optimization and the validity of the properties are checked using the dynamic impact simulations Using a sophisticated calibration procedure, the effective material properties of simplified beam model of spent fuel rods can be successfully determined.

[en] We have calculated the scattering cross-section of the e⁺e^-→γγ process in the lowest order of QED. In addition to the single electron exchange diagram we have considered the possible contribution of heavy excited electrons. A symbolic manipulation program, REDUCE, is used for algebraic calculation of the cross section. Illustrations are given or QED tests for possible compositeness of leptons. Also we have studied the QED cuttoff parameters for the excited electrons search. (Author)

[en] Pulse pile-up distortion is a common problem for radiation spectroscopy measurements involving high counting rates. The Monte Carlo pile-up to true approach (MCPUT) is proposed and benchmarked in this article for correcting pile-up distorted pulse-height spectra to true spectra. In previous work, a Monte Carlo approach was used for predicting the pile-up distorted pulse-height spectra for high counting-rate measurements ('the forward calculation'). The present work improves the previous simulation by employing a better ADC dead-time model. Based on this improved 'forward calculation', the MCPUT approach introduces an iterative procedure for correcting pile-up distortions. Experiments with an Fe-55 source and a Si(Li) detector are used for benchmarking purposes. The MCPUT corrected spectrum for the high counting-rate measurement shows excellent agreement with the measured true spectrum at low counting rate with reduced chi-square as the quantitative measure. The approach is also efficient, as accurate calculations are possible in a few minutes