• Korean Journal of Breeding Science
• /
• v.43 no.1
• /
• pp.68-80
• /
• 2011

This study was conducted to investigate characteristics and variations of rice panicle traits for breeding temperate japonica varieties adapted to tropical environment. Eleven panicle traits were investigated from nine Korean rice varieties cultivated in Korea and wet season of the Philippines. Tested cultivars were composed of six temperate japonica varieties, three Tongil-type varieties, and one indica variety bred in the Philippines. The number of spikelets on secondary rachis branches (SRBs) was the most variable trait in both environments, while the mean number of spikelets on a primary rachis branch (PRB) was the least variable. Compared with PRB-related traits, SRB traits showed higher correlation with the number of spikelets per panicle. Compared with the plants grown in Korea, the number of spikelets on SRBs, the number of SRBs, spikelets, and rachis branches per panicle were decreased more than other traits in the Philippines. According to path analysis, the number of spikelets on SRBs per panicle affects the number of spikelets per panicle more than the number of spikelets on PRBs per panicle. Climatic factors such as growth duration, cumulative mean temperature, and integrated solar radiation were highly correlated with the relative rate of number of spikelets per panicle. To breed temperate japonica rice varieties adapted to tropical environment, it would be important to select lines which maintain proper growth duration and spikelets on SRBs in target region.

• Korean Journal of Digital Imaging in Medicine
• /
• v.16 no.1
• /
• pp.13-19
• /
• 2014

Purpose: FLAIR image is beneficial for the diagnosis of various bran diseases including ischemic CVS, brain tumors and infections. However the border between the legion of brain metastasis and surrounding edema may not be clear. Therefore, this study aims to investigate the practical benefits of delayed imaging by comparing the image from a patient with brain metastasis before a contrast enhancement and the image 10 minutes after a contrast enhancement. Materials and methods: Of the 92 people who underwent MRI brain metastases in suspected patients 13 people in three patients there is no video to target the 37 people confirmed cases, and motion artifacts brain metastases in our hospital June-December 2013, 18 people measurement position except for the three incorrect patient (male: 11 people, female: 7 people, average age: 60 years) in the target, test equipment, 3.0T MR System (ACHIEVA Release, Philips, I was 8ChannelSENSE Head Coil use Best, and the Netherlands). TR 11000 ms, TE 125 ms, TI2800 ms, Slice Thickness 5 mm, gap 5 mm, is a Slice number 21, the parameters of the 3D FFE, T2 FLAIR variable that was used to test, TR 8.1 ms, TE 3.7 ms, Slice number 240 I set to. The experiment was conducted by acquiring the FLAIR prior to contrast enhancement (heretofore referred to as Pre FLAIR), and acquiring the 3D FFE CE five minutes after the contrast enhancement, and recomposing the images in an axial plane of S/T 3mm, G 0mm (heretofore referred to as MPR TRA CE). Using the FLAIR 10 minutes after the contrast enhancement (heretofore referred to as Post FLAIR) and Pi-View, a retrospective study was conducted. Using MRIcro on the image of a patient confirmed for his diagnosis, the images before and after the contrast media, as well as the CNR and SNR of the MPR TRA CE images of the lesion and the site absent of lesion were compared and analyzed using a one-way analysis of variance. Results: CNR for Pre FLAIR and Post FLAIR were 34.35 and 60.13, respectively, with MPR TRA CE at 23.77 showing no significant difference (p<0.050). Post-experiment analysis shows a difference between Pre FLAIR and Post FLAIR in terms of CNR (p<0.050), but no difference in CNR between Post FLAIR and MPR TRA CE (p>0.050), indicating that the contrast media had an effect only on Pre FLAIR and Post FLAIR. The SNR for the normal site Pre FLAIR was 106.43, and for the lesion site 140.79. Post FLAIR for the normal site was 107.79, and for the lesion site 167.91. MPR TRA CE for the normal site was 140.23 and for the lesion site 183.19, showing significant difference (p<0.050), and post-experiment analysis shows that there was a difference in SNR only on the lesion sites for Pre FLAIR and Post FLAIR (p<0.050). There was no difference in SNR between the normal site and lesion site for Post FLAIR and MPR TRA CE, indicating no effect from the contrast media (p>0.050). Conclusions: This experiment shows that Post FLAIR has a higher contrast than Pre FLAIR, and a higher SNR for lesions, It was not not statistically significant and MPR TRA CE but CNR came out high. Inspection of post-contrast which is used in a high magnetic field is frequently used images of 3D T1 but, since the signal of the contrast medium and the blood flow is included, this method can be diagnostic accuracy is reduced, it is believed that when used in combination with Post FLAIR, and that can provide video information added to the diagnosis of brain metastases.

• Journal of Intelligence and Information Systems
• /
• v.21 no.3
• /
• pp.101-116
• /
• 2015

Recently, due to the development of smart devices and social media, vast amounts of information with the various forms were accumulated. Particularly, considerable research efforts are being directed towards analyzing unstructured big data to resolve various social problems. Accordingly, focus of data-driven decision-making is being moved from structured data analysis to unstructured one. Also, in the field of recommendation system, which is the typical area of data-driven decision-making, the need of using unstructured data has been steadily increased to improve system performance. Approaches to improve the performance of recommendation systems can be found in two aspects- improving algorithms and acquiring useful data with high quality. Traditionally, most efforts to improve the performance of recommendation system were made by the former approach, while the latter approach has not attracted much attention relatively. In this sense, efforts to utilize unstructured data from variable sources are very timely and necessary. Particularly, as the interests of users are directly connected with their needs, identifying the interests of the user through unstructured big data analysis can be a crew for improving performance of recommendation systems. In this sense, this study proposes the methodology of improving recommendation system by measuring interests of the user. Specially, this study proposes the method to quantify interests of the user by analyzing user's internet usage patterns, and to predict user's repurchase based upon the discovered preferences. There are two important modules in this study. The first module predicts repurchase probability of each category through analyzing users' purchase history. We include the first module to our research scope for comparing the accuracy of traditional purchase-based prediction model to our new model presented in the second module. This procedure extracts purchase history of users. The core part of our methodology is in the second module. This module extracts users' interests by analyzing news articles the users have read. The second module constructs a correspondence matrix between topics and news articles by performing topic modeling on real world news articles. And then, the module analyzes users' news access patterns and then constructs a correspondence matrix between articles and users. After that, by merging the results of the previous processes in the second module, we can obtain a correspondence matrix between users and topics. This matrix describes users' interests in a structured manner. Finally, by using the matrix, the second module builds a model for predicting repurchase probability of each category. In this paper, we also provide experimental results of our performance evaluation. The outline of data used our experiments is as follows. We acquired web transaction data of 5,000 panels from a company that is specialized to analyzing ranks of internet sites. At first we extracted 15,000 URLs of news articles published from July 2012 to June 2013 from the original data and we crawled main contents of the news articles. After that we selected 2,615 users who have read at least one of the extracted news articles. Among the 2,615 users, we discovered that the number of target users who purchase at least one items from our target shopping mall 'G' is 359. In the experiments, we analyzed purchase history and news access records of the 359 internet users. From the performance evaluation, we found that our prediction model using both users' interests and purchase history outperforms a prediction model using only users' purchase history from a view point of misclassification ratio. In detail, our model outperformed the traditional one in appliance, beauty, computer, culture, digital, fashion, and sports categories when artificial neural network based models were used. Similarly, our model outperformed the traditional one in beauty, computer, digital, fashion, food, and furniture categories when decision tree based models were used although the improvement is very small.

• Geophysics and Geophysical Exploration
• /
• v.8 no.1
• /
• pp.7-17
• /
• 2005

The determination of seismic velocities in refractors for near-surface seismic refraction investigations is an ill-posed problem. Small variations in the computed time parameters can result in quite large lateral variations in the derived velocities, which are often artefacts of the inversion algorithms. Such artefacts are usually not recognized or corrected with forward modelling. Therefore, if detailed refractor models are sought with model based inversion, then detailed starting models are required. The usual source of artefacts in seismic velocities is irregular refractors. Under most circumstances, the variable migration of the generalized reciprocal method (GRM) is able to accommodate irregular interfaces and generate detailed starting models of the refractor. However, where the very-near-surface environment of the Earth is also irregular, the efficacy of the GRM is reduced, and weathering corrections can be necessary. Standard methods for correcting for surface irregularities are usually not practical where the very-near-surface irregularities are of limited lateral extent. In such circumstances, the GRM smoothing statics method (SSM) is a simple and robust approach, which can facilitate more-accurate estimates of refractor velocities. The GRM SSM generates a smoothing 'statics' correction by subtracting an average of the time-depths computed with a range of XY values from the time-depths computed with a zero XY value (where the XY value is the separation between the receivers used to compute the time-depth). The time-depths to the deeper target refractors do not vary greatly with varying XY values, and therefore an average is much the same as the optimum value. However, the time-depths for the very-near-surface irregularities migrate laterally with increasing XY values and they are substantially reduced with the averaging process. As a result, the time-depth profile averaged over a range of XY values is effectively corrected for the near-surface irregularities. In addition, the time-depths computed with a Bero XY value are the sum of both the near-surface effects and the time-depths to the target refractor. Therefore, their subtraction generates an approximate 'statics' correction, which in turn, is subtracted from the traveltimes The GRM SSM is essentially a smoothing procedure, rather than a deterministic weathering correction approach, and it is most effective with near-surface irregularities of quite limited lateral extent. Model and case studies demonstrate that the GRM SSM substantially improves the reliability in determining detailed seismic velocities in irregular refractors.

• Tuberculosis and Respiratory Diseases
• /
• v.44 no.4
• /
• pp.871-888
• /
• 1997

Background : It is now thought that the earliest manifestation of idiopathic pulmonary fibrosis is alveolitis, that is, an accumulation of inflammatory and immune effector cells within alveolar walls and spaces. Inflammatory cells including alveolar macrophages and resident normal pulmonary tissue cells participate through the release of many variable mediators such as inflammatory growth factors and cytokines, which contribute to tissue damage and finally cause chronic pulmonary inflammation and fibrosis. This study was performed to investigate the source and distribution pattern of transforming growth factor-${\beta}_1$(TGF-${\beta}_1$), platelet derived growth factor(PDGF), basic fibroblast growth factor(bFGF), interleukin 1(IL-1), interleukin 6(IL-6), tumor necrosis factor-$\alpha$ (TNF-$\alpha$) and the role of these mediators on bleomycin(BLM)-induced pulmonary injury and fibrosis in rats. Method : Wistar rats were divided into three groups(control group, BLM treated group, BLM and vitamine E treated group). Animals were sacrificed periodically at 1, 2, 3, 4, 5, 7, 14, 21, 28 days after saline or BLM administration. The effects were compared to the results of bronchoalveolar lavage fluid analysis, light microscopic findings, immunohistochemical stains for six different mediators(TGF-${\beta}_1$, PDGF, bFGF, IL-1, IL-6 and TNF-$\alpha$) and mRNA in situ hybridization for TGF-${\beta}_1$. Results : IL-1 and IL-6 are maximally expressed at postbleomycin 1~7th day which are mainly produced by neutrophils and bronchiolar epithelium. It is thought that they induce recruitment of inflammatory cells at the injury site. The expression of IL-1 and IL-6 at the bronchiolar epithelium within 7th day is an indirect evidence of contribution of bronchiolar epithelial cells to promote and maintain the inflammatory and immune responses adjacent to the airways. TNF-$\alpha$ is mainly produced by neutrophils and bronchiolar epithelial cells during 1~5th day, alveolar macrophages during 7~28th day. At the earlier period, TNF-$\alpha$ causes recruitment of inflammatory cells at the injury site and later stimulates pulmonary fibrosis. The main secreting cells of TGF-${\beta}_1$ are alveolar macrophages and bronchiolar epithelium and the target is pulmonary fibroblasts and extracellular matrix. TGF-${\beta}_1$ and PDGF stimulate proliferation of pulmonary fibroblasts and TGF-${\beta}_1$ and bFGF incite the fibroblasts to produce extracellular matrix. The vitamine E and BLM treated group shows few positive cells(p<0.05). Conclusion : After endothelial and epithelial injury, the neutrophils and bronchiolar epithelium secrete IL-1, IL-6, TNF-$\alpha$ which induce infiltration of many neutrophils. It is thought that variable enzymes and $O_2$ radicals released by these neutrophils cause destruction of normal lung architecture and progression of pulmonary fibrosis. At the 7~28th day, TGF-${\beta}_1$, PDGF, bFGF, TNF-$\alpha$ secreted by alveolar macrophages sting pulmonary fibroblasts into proliferating with increased production of extracellular matrix and finally, they make progression of pulmonary fibrosis. TNF-$\alpha$ compares quite important with TGF-${\beta}_1$ to cause pulmonary fibrosis. Vitamine E seems to decrease the extent of BLM induced pulmonary fibrosis.

• Journal of Global Scholars of Marketing Science
• /
• v.20 no.2
• /
• pp.135-148
• /
• 2010

Arab boycotts of Danish products, Australian boycotts of French products and Chinese consumer aversion toward Japanese products are all examples of how adverse actions at the country level might impact consumers' behavior. The animosity literature has examined how consumers react to the adverse actions of other countries, and how such animosity impacts consumers' attitudes and preferences for products from the transgressing country. For example, Chinese consumers are less likely to buy Japanese products because of Japanese atrocities during World War II and the unjust economic dealings of the Japanese (Klein, Ettenson and Morris 1998). The marketing literature, however, has not examined how consumers react to adverse actions committed by their own country against other countries, and whether such actions affect their attitudes towards purchasing products that originated from the adversely affected country. The social psychology literature argues that consumers will experience a feeling called collective guilt, in response to such adverse actions. Collective guilt stems from the distress experienced by group members when they accept that their group is responsible for actions that have harmed another group (Branscombe, Slugoski, and Kappenn 2004). Examples include Americans feeling guilty about the atrocities committed by the U.S. military at Abu Ghraib prison (Iyer, Schamder and Lickel 2007), and the Dutch about their occupation of Indonesia in the past (Doosje et al. 1998). The primary aim of this study is to examine consumers' perceptions of adverse actions by members of one's own country against another country and whether such perceptions affected their attitudes towards products originating from the country transgressed against. More specifically, one objective of this study is to examine the perceptual antecedents of collective guilt, an emotional reaction to adverse actions performed by members of one's country against another country. Another objective is to examine the impact of collective guilt on consumers' perceptions of, and preference for, products originating from the country transgressed against by the consumers' own country. If collective guilt emerges as a significant predictor, companies originating from countries that have been transgressed against might be able to capitalize on such unfortunate events. This research utilizes the animosity model introduced by Klein, Ettenson and Morris (1998) and later expanded on by Klein (2002). Klein finds that U.S. consumers harbor animosity toward the Japanese. This animosity is experienced in response to events that occurred during World War II (i.e., the bombing of Pearl Harbor) and more recently the perceived economic threat from Japan. Thus this study argues that the events of Word War II (i.e., bombing of Hiroshima and Nagasaki) might lead U.S. consumers to experience collective guilt. A series of three hypotheses were introduced. The first hypothesis deals with the antecedents of collective guilt. Previous research argues that collective guilt is experienced when consumers perceive that the harm following a transgression is illegitimate and that the country from which the transgressors originate should be responsible for the adverse actions. (Wohl, Branscombe, and Klar 2006). Therefore the following hypothesis was offered: H1a. Higher levels of perceived illegitimacy for the harm committed will result in higher levels of collective guilt. H1b. Higher levels of responsibility will be positively associated with higher levels of collective guilt. The second and third hypotheses deal with the impact of collective guilt on the preferences for Japanese products. Klein (2002) found that higher levels of animosity toward Japan resulted in a lower preference for a Japanese product relative to a South Korean product but not a lower preference for a Japanese product relative to a U.S. product. These results therefore indicate that the experience of collective guilt will lead to a higher preference for a Japanese product if consumers are contemplating a choice that inv olves a decision to buy Japanese versus South Korean product but not if the choice involves a decision to buy a Japanese versus a U.S. product. H2. Collective guilt will be positively related to the preference for a Japanese product over a South Korean product, but will not be related to the preference for a Japanese product over a U.S. product. H3. Collective guilt will be positively related to the preference for a Japanese product over a South Korean product, holding constant product judgments and animosity. An experiment was conducted to test the hypotheses. The illegitimacy of the harm and responsibility were manipulated by exposing respondents to a description of adverse events occurring during World War II. Data were collected using an online consumer panel in the United States. Subjects were randomly assigned to either the low levels of responsibility and illegitimacy condition (n=259) or the high levels of responsibility and illigitemacy (n=268) condition. Latent Variable Structural Equation Modeling (LVSEM) was used to test the hypothesized relationships. The first hypothesis is supported as both the illegitimacy of the harm and responsibility assigned to the Americans for the harm committed against the Japanese during WWII have a positive impact on collective guilt. The second hypothesis is also supported as collective guilt is positively related to preference for a Japanese product over a South Korean product but is not related to preference for a Japanese product over a U.S. product. Finally there is support for the third hypothesis, since collective guilt is positively related to the preference for a Japanese product over a South Korean product while controlling for the effect of product judgments about Japanese products and animosity. The results of these studies lead to several conclusions. First, the illegitimacy of harm and responsibility can be manipulated and that they are antecedents of collective guilt. Second, collective guilt has an impact on a consumers' decision when they face a choice set that includes a product from the country that was the target of the adverse action and a product from another foreign country. This impact however disappears from a consumers' decision when they face a choice set that includes a product from the country that was the target of the adverse action and a domestic product. This result suggests that collective guilt might be a viable factor for company originating from the country transgressed against if its competitors are foreign but not if they are local.

• Journal of Intelligence and Information Systems
• /
• v.18 no.3
• /
• pp.185-202
• /
• 2012

.Since the value of information has been realized in the information society, the usage and collection of information has become important. A facial expression that contains thousands of information as an artistic painting can be described in thousands of words. Followed by the idea, there has recently been a number of attempts to provide customers and companies with an intelligent service, which enables the perception of human emotions through one's facial expressions. For example, MIT Media Lab, the leading organization in this research area, has developed the human emotion prediction model, and has applied their studies to the commercial business. In the academic area, a number of the conventional methods such as Multiple Regression Analysis (MRA) or Artificial Neural Networks (ANN) have been applied to predict human emotion in prior studies. However, MRA is generally criticized because of its low prediction accuracy. This is inevitable since MRA can only explain the linear relationship between the dependent variables and the independent variable. To mitigate the limitations of MRA, some studies like Jung and Kim (2012) have used ANN as the alternative, and they reported that ANN generated more accurate prediction than the statistical methods like MRA. However, it has also been criticized due to over fitting and the difficulty of the network design (e.g. setting the number of the layers and the number of the nodes in the hidden layers). Under this background, we propose a novel model using Support Vector Regression (SVR) in order to increase the prediction accuracy. SVR is an extensive version of Support Vector Machine (SVM) designated to solve the regression problems. The model produced by SVR only depends on a subset of the training data, because the cost function for building the model ignores any training data that is close (within a threshold ${\varepsilon}$) to the model prediction. Using SVR, we tried to build a model that can measure the level of arousal and valence from the facial features. To validate the usefulness of the proposed model, we collected the data of facial reactions when providing appropriate visual stimulating contents, and extracted the features from the data. Next, the steps of the preprocessing were taken to choose statistically significant variables. In total, 297 cases were used for the experiment. As the comparative models, we also applied MRA and ANN to the same data set. For SVR, we adopted '${\varepsilon}$-insensitive loss function', and 'grid search' technique to find the optimal values of the parameters like C, d, ${\sigma}^2$, and ${\varepsilon}$. In the case of ANN, we adopted a standard three-layer backpropagation network, which has a single hidden layer. The learning rate and momentum rate of ANN were set to 10%, and we used sigmoid function as the transfer function of hidden and output nodes. We performed the experiments repeatedly by varying the number of nodes in the hidden layer to n/2, n, 3n/2, and 2n, where n is the number of the input variables. The stopping condition for ANN was set to 50,000 learning events. And, we used MAE (Mean Absolute Error) as the measure for performance comparison. From the experiment, we found that SVR achieved the highest prediction accuracy for the hold-out data set compared to MRA and ANN. Regardless of the target variables (the level of arousal, or the level of positive / negative valence), SVR showed the best performance for the hold-out data set. ANN also outperformed MRA, however, it showed the considerably lower prediction accuracy than SVR for both target variables. The findings of our research are expected to be useful to the researchers or practitioners who are willing to build the models for recognizing human emotions.

• Journal of Bio-Environment Control
• /
• v.22 no.4
• /
• pp.341-348
• /
• 2013

Maintenance of adequate soil tension or content during the period of crop growth is necessary to support optimum plant growth and yields. A better understanding of soil tension and content for precision irrigation would allow optimal soil water condition to crops and minimize the adverse effects of water stress on crop growth and development. This research reports on a comparison of soil water tension and content variations in differently textured soils over time under drip irrigation using two different water management methods, i.e. pulse time and required water irrigation methods. The pulse time-based irrigation was performed by turning the solenoid valve on and off for preset times to allow the wetting front to disperse in root zone before additional water was applied. The required water estimation method was a new water control logic designed by Rural Development Administration that applies the amount of water required based on a conversion of the measured water tension into water content. The use of the pulse time irrigation method under drip irrigation at a high tension of -20 kPa and high temperatures over $30^{\circ}C$ was not successful at maintaining moisture tensions within an appropriate range of 5 kPa because the preset irrigation times used for water control could not compensate for the change in evapotranspiration during day and night. The response time and pattern of water contents for all of the tested soils measured with capacitance-based sensor probes were faster and more direct than those of water tensions measured with porous and ceramic cup-based tensiometers when water was applied, indicating water content would be a better control variable for automatic irrigation. The required water estimation-based irrigation method provided relatively stable control of moisture tension, even though somewhat lower tension values were obtained as compared to the target tension of -20 kPa, indicating that growers could expect to be effective in controlling low tensions ranging from -10 to -20 kPa with the required water estimation system.

• Korean Chemical Engineering Research
• /
• v.56 no.4
• /
• pp.479-495
• /
• 2018

This study was carried out to research the multimedia fate modeling, concentration distribution and impact assessment of polycyclic aromatic hydrocarbons (PAHs) emitted from automobiles, which are known as carcinogenic and mutation chemicals. The amount of emissions of PAHs was determined based on the census data of automobiles at a target S-city and emission factors of PAHs, where multimedia fugacity modeling was conducted by the restriction of PAHs transfer between air-soil at the impervious area. PAHs' Concentrations and their distributions at several environmental media were predicted by multimedia fugacity model (level III). The residual amounts and the distributions of PAHs through mass transfer of PAHs between environment media were used to assess the health risk of PAHs at unsteady state (level IV), where the sensitivity analyses of the model parameter of each variable were conducted based on Monte Carlo simulation. The experimental result at S-city showed that Fluoranthene among PAHs substances are the highest residual concentrations (60%, 53%, 32% and 34%) at all mediums (atmospheric, water, soil, sediment), respectively, where most of the PAHs were highly accumulated in the sediment media (more than 80%). A result of PAHs concentration changes in S-city over the past 34 years identified that PAHs emissions from all environmental media increased from 1983 to 2005 and decreased until 2016, where the emission of heavy-duty vehicle including truck revealed the largest contribution to the automotive emissions of PAHs at all environment media. The PAHs concentrations in soil and water for the last 34 years showed the less value than the legal standards of PAHs, but the PAHs in air exceeded the air quality standards from 1996 to 2016. The result of this study is expected to contribute the effective management and monitoring of toxic chemicals of PAHs at various environment media of Metropolitan city.

• Radiation Oncology Journal
• /
• v.20 no.1
• /
• pp.1-16
• /
• 2002

Purpose : Three dimensional conformal radiotherapy planning is being used widely for the treatment of patients with brain tumor. However, it takes much time to develop an optimal treatment plan, therefore, it is difficult to apply this technique to all patients. To increase the efficiency of this technique, we need to develop standard radiotherapy plant for each site of the brain. Therefore we developed several 3 dimensional conformal radiotherapy plans (3D plans) for tumors at each site of brain, compared them with each other, and with 2 dimensional radiotherapy plans. Finally model plans for each site of the brain were decide. Materials and Methods : Imaginary tumors, with sizes commonly observed in the clinic, were designed for each site of the brain and drawn on CT images. The planning target volumes (PTVs) were as follows; temporal $tumor-5.7\times8.2\times7.6\;cm$, suprasellar $tumor-3\times4\times4.1\;cm$, thalamic $tumor-3.1\times5.9\times3.7\;cm$, frontoparietal $tumor-5.5\times7\times5.5\;cm$, and occipitoparietal $tumor-5\times5.5\times5\;cm$. Plans using paralled opposed 2 portals and/or 3 portals including fronto-vertex and 2 lateral fields were developed manually as the conventional 2D plans, and 3D noncoplanar conformal plans were developed using beam's eye view and the automatic block drawing tool. Total tumor dose was 54 Gy for a suprasellar tumor, 59.4 Gy and 72 Gy for the other tumors. All dose plans (including 2D plans) were calculated using 3D plan software. Developed plans were compared with each other using dose-volume histograms (DVH), normal tissue complication probabilities (NTCP) and variable dose statistic values (minimum, maximum and mean dose, D5, V83, V85 and V95). Finally a best radiotherapy plan for each site of brain was selected. Results : 1) Temporal tumor; NTCPs and DVHs of the normal tissue of all 3D plans were superior to 2D plans and this trend was more definite when total dose was escalated to 72 Gy (NTCPs of normal brain 2D $plans:27\%,\;8\%\rightarrow\;3D\;plans:1\%,\;1\%$). Various dose statistic values did not show any consistent trend. A 3D plan using 3 noncoplanar portals was selected as a model radiotherapy plan. 2) Suprasellar tumor; NTCPs of all 3D plans and 2D plans did not show significant difference because the total dose of this tumor was only 54 Gy. DVHs of normal brain and brainstem were significantly different for different plans. D5, V85, V95 and mean values showed some consistent trend that was compatible with DVH. All 3D plans were superior to 2D plans even when 3 portals (fronto-vertex and 2 lateral fields) were used for 2D plans. A 3D plan using 7 portals was worse than plans using fewer portals. A 3D plan using 5 noncoplanar portals was selected as a model plan. 3) Thalamic tumor; NTCPs of all 3D plans were lower than the 2D plans when the total dose was elevated to 72 Gy. DVHs of normal tissues showed similar results. V83, V85, V95 showed some consistent differences between plans but not between 3D plans. 3D plans using 5 noncoplanar portals were selected as a model plan. 4) Parietal (fronto- and occipito-) tumors; all NTCPs of the normal brain in 3D plans were lower than in 2D plans. DVH also showed the same results. V83, V85, V95 showed consistent trends with NTCP and DVH. 3D plans using 5 portals for frontoparietal tumor and 6 portals for occipitoparietal tumor were selected as model plans. Conclusion : NTCP and DVH showed reasonable differences between plans and were through to be useful for comparing plans. All 3D plans were superior to 2D plans. Best 3D plans were selected for tumors in each site of brain using NTCP, DVH and finally by the planner's decision.