• Korean Journal of Environment and Ecology
• /
• v.38 no.3
• /
• pp.277-292
• /
• 2024

In order to manage large-scale forest fires concentrated in Gangwon-do and Gyeongsangbuk-do with severe topographical heterogeneity, a decision-making process through efficient and rapid damage assessment using satellite images is essential. Accordingly, this study targets a large-scale forest fire that ignited in Gangneung and the Donghae, Gangwon-do on March 5, 2022, and was extinguished around 19:00 on March 8, to estimate the fire severity using dNBR and derive environmental factors that affect the grade. As environmental factors, we quantified the regular vegetation index representing vegetation or fuel type, the forest index that classifies tree species, the regular moisture index representing moisture content, and DEM in relation to topography, and then analyzed the correlation with the fire severity. In terms of fire severity, the widest range was 'Unbured' at 52.4%, followed by low severity at 42.9%, medium-low severity at 4.3%, and medium-high severity at 0.4%. Environmental factors showed a negative correlation with dNDVI and dNDWI, and a positive correlation with slope. Regarding vegetation, the differences between coniferous, broad-leaved, and other groups in dNDVI, dNIWI, and slope, which were analyzed to affect the fire severity, were analyzed to be significant with p-value < 2.2e-16. In particular, the difference between coniferous and broad-leaved forests was clear, and it was confirmed that coniferous forest suffered more damage than broad-leaved forest due to the higher fire severity in the Gangwon-do region, including Pinus densiflora, which are dominant species, as well as P. koraiensis, P. rigida and P. thunbergii.

• The Journal of Korean Society for Radiation Therapy
• /
• v.34
• /
• pp.21-30
• /
• 2022

Purpose: The purpose of this study is to evaluate the effectiveness of oral positioning stent, the BinkieRT^TM in radiation treatment for head and neck cancer patients in terms of tongue positions reproducibility, tongue doses and material properties. Materials and Methods: 24 cases using BinkieRT^TM during radiation treatments were enrolled. The tongue was contoured on planning CT and CBCT images taken every 3 days during treatment, and then the DSC and center of tongue shift values were analyzed to evaluate the reproducibility of the tongue. The tongue dose was compared in terms of dose distribution when using BinkieRT^TM and different type of oral stents (mouthpiece, paraffin wax). Randomly selected respective 10 patients were measured tongue doses of initial treatment plan for nasal cavity and unilateral parotid cancer. Finally, In terms of material evaluation, HU and relative electron density were identified in RTPS. Results: As a result of DSC analysis, it was 0.8 ± 0.07, skewness -0.8, kurtosis 0.61, and 95% CI was 0.79~0.82. To analyze the deviation of the central tongue shift during the treatment period, a 95% confidence interval for shift in the LR, SI, and AP directions were indicated, and a one-sample t-test for 0, which is an ideal value in the deviation(n=144). As a result of the t-test, the mean and SD in the LR and SI directions were 0.01 ± 0.14 cm (p→.05), 0.03 ± 0.25 cm (p→.05), and -0.08 ± 0.25 cm (p ←.05) in the AP direction. In the case of unilateral parotid cancer patients, the Dmean to the tongue of patients using BinkieRT^TM was 16.92% ± 3.58% compared to the prescribed dose, and 23.99% ± 10.86% of patients with Paraffin Wax, indicating that the tongue dose was relatively lower when using BinkieRT^TM (p←.05). On the other hand, among nasal cavity cancer patients, the Dmean of tongue dose for patients who used BinkieRT^TM was 4.4% ± 5.6%, and for those who used mouthpiece, 5.9% ± 6.8%, but it was not statistically significant (p→.05). The relative electron density of Paraffin Wax, BinkieRT^TM and Putty is 0.94, 0.99, 1.26 and the mass density is 0.95, 0.99 and 1.32 (g/cc), Transmission Factor is 0.99, 0.98, 0.96 respectively. Conclusion: The result of the tongue DSC analysis over the treatment period was about 0.8 and Deviation of the center of tongue shifts were within 0.2 cm, the reproducibility was more likely excellent. In the case of unilateral head and neck cancer patients, it was found that the use of BinkieRT^TM rather than Paraffin Wax or Putty can reduce the unnecessary dose irradiated to the tongue. This study might be useful to understand of BinkieRT^TM's properties and advantages. And also it could be another considered option as oral stent to keep the reproducibility of tongue and reducing dose during head and neck radiation treatments.

• The Journal of Korean Society for Radiation Therapy
• /
• v.26 no.1
• /
• pp.43-50
• /
• 2014

In case of all patients who receive radiation therapy, a treatment plan is established and all steps of treatment are planned in the same geometrical condition. In case of head and neck cancer patients who undergo simulated treatment through computed tomography (CT), patients are fixed onto a table for planning, but laid on the top of the treatment table in the radiation therapy room. This study excogitated and fabricated an adjustable holder for head and neck cancer patients to fix patient's position and geometrical discrepancies when performing radiation therapy on head and neck cancer patients, and compared the error before and after adjusting the position of patients due to difference in weight to evaluate the correlation between patients' weight and range of error. Computed tomography system(High Advantage, GE, USA) is used for phantom to maintain the supine position to acquire the images of the therapy site for IMRT. IMRT 4MV X-rays was used by applying the LINAC(21EX, Varian, U.S.A). Treatment planning system (Pinnacle, ver. 9.1h, Philips, Madison, USA) was used. The setup accuracy was compared with each measurement was repeated five times for each weight (0, 15, and 30Kg) and CBCT was performed 30 times to find the mean and standard deviation of errors before and after the adjustment of each weight. SPSS ver.19.0(SPSS Inc., Chicago, IL,USA) statistics program was used to perform the Wilcoxon Rank test for significance evaluation and the Spearman analysis was used as the tool to analyze the significance evaluation of the correlation of weight. As a result of measuring the error values from CBCT before and after adjusting the position due to the weight difference, X,Y,Z axis was $0.4{\pm}0.8mm$, $0.8{\pm}0.4mm$, 0 for 0Kg before the adjustment. In 15Kg CBCT before and after adjusting the position due to the weight difference, X,Y,Z axis was $0.2{\pm}0.8mm$, $1.2{\pm}0.4mm$, $2.0{\pm}0.4mm$. After adjusting position was X,Y,Z axis was $0.2{\pm}0.4mm$, $0.4{\pm}0.5mm$, $0.4{\pm}0.5mm$. In 30Kg CBCT before and after adjusting the position due to the weight difference, X,Y,Z axis was $0.8{\pm}0.4mm$, $2.4{\pm}0.5mm$, $4.4{\pm}0.8mm$. After adjusting position was X,Y,Z axis was $0.6{\pm}0.5mm$, $1.0{\pm}0mm$, $0.6{\pm}0.5mm$. When the holder for the head and neck cancer was used to adjust the ab.0ove error value, the error values from CBCT were $0.2{\pm}0.8mm$ for the X axis, $0.40{\pm}0.54mm$ for Y axis, and 0 for Z axis. As a result of statistically analyzing each value before and after the adjustment the value was significant with p<0.034 at the Z axis with 15Kg of weight and with p<0.038 and p<0.041 at the Y and Z axes respectively with 30Kg of weight. There was a significant difference with p<0.008 when the analysis was performed through Kruscal-Wallis in terms of the difference in the adjusted values of the three weight groups. As it could reduce the errors, patients' reproduction could be improved for more precise and accurate radiation therapy. Development of an adjustable device for head and neck cancer patients is significant because it improves the reproduction of existing equipment by reducing the errors in patients' position.

• Journal of the Korean Society of Radiology
• /
• v.6 no.5
• /
• pp.365-371
• /
• 2012

The radiation therapy treatment technique is developed from 3D-CRT, IMRT to Tomotherapy. and these three technique was most widely using methods. We find out a comparison normal tissue doses and tumor dose of 3D-CRT, IMRT(Linac Based), and Tomotherapy on Head and Neck Cancer. We achieved radiological image used the Human model phantom (Anthropomorphic Phantom) and it was taken CT simulation (Slice Thickness : 3mm) and GTV was nasopharngeal region and PTV(including set-up margin) was GTV plus 2mm area. and transfer those images to the radiation planning system (3D-CRT - ADAC-Pinnacle3, Tomotherapy - Tomotherapy Hi-Art System). The prescription dose was 7020 cGy and measuring PTV's dose and nomal tissue (parotid gland, oral cavity, spinal cord). The PTV's doses was Tomotherapy, Linac Based - IMRT, 3D-CRT was 6923 cGy, 6901 cGy and 6718 cGy its dose value was meet TCP because its value was up to the 95% based on 7020 cGy, Nomal tissue (parotid gland, oral cavity, spinal cord) was 1966 cGy(Tomotherapy), 2405 cGy(IMRT), 2468 cGy(3D-CRT)[parotid gland], 2991 cGy(Tomotherapy), 3062 cGy(IMRT), 3684 cGy (3D-CRT)[oral cavity], 1768 cGy(Tomotherapy), 2151 cGy(IMRT), 4031 cGy(3D-CRT)[spinal cord] its value did not exceeded NTCP. All the treatment techniques are equated with tumor and nomal tissue doses. The 3D-CRT was worse than other techniques on dose distribution, but it is reasonable in terms of TCP and NTCP baseline Tomotherapy, IMRT -dose distribution was relatively superior- was hard to therapy to claustrophobic patients and patients with respiratory failure. Particularly, in case on Tomotherapy, it take MVCT before treatment so dose measurement will be unnecessary radiation exposure to patients. Conclusion, Tomotherapy was the best treatment technique and 2nd was IMRT, and 3rd 3D-CRT. But applicable differently depending on the the patient's condition even though dose not matter.

• Nuclear Medicine and Molecular Imaging
• /
• v.42 no.3
• /
• pp.192-200
• /
• 2008

To evaluate the hemodynamic changes and the predictive factors of the clinical outcome in pediatric patients with moyamoya disease, we analyzed pre/post basal/acetazolamide stress brain perfusion SPECT with automated volume of interest (VOIs) method. Methods: Total fifty six (M:F = 33:24, age $6.7{\pm}3.2$ years) pediatric patients with moyamoya disease, who underwent basal/acetazolamide stress brain perfusion SPECT within 6 before and after revascularization surgery (encephalo-duro-arterio-synangiosis (EDAS) with frontal encephalo-galeo-synangiosis (EGS) and EDAS only followed on contralateral hemisphere), and followed-up more than 6 months after post-operative SPECT, were included. A mean follow-up period after post-operative SPECT was $33{\pm}21$ months. Each patient's SPECT image was spatially normalized to Korean template with the SPM2. For the regional count normalization, the count of pons was used as a reference region. The basal/acetazolamide-stressed cerebral blood flow (CBF), the cerebral vascular reserve index (CVRI), and the extent of area with significantly decreased basal/acetazolamide- stressed rCBF than age-matched normal control were evaluated on both medial frontal, frontal, parietal, occipital lobes, and whole brain in each patient's images. The post-operative clinical outcome was assigned as good, poor according to the presence of transient ischemic attacks and/or fixed neurological deficits by pediatric neurosurgeon. Results: In a paired t-test, basal/acetazolamide-stressed rCBF and the CVRI were significantly improved after revascularization (p<0.05). The significant difference in the pre-operative basal/acetazolamide-stressed rCBF and the CVRI between the hemispheres where EDAS with frontal EGS was performed and their contralateral counterparts where EDAS only was done disappeared after operation (p<0.05). In an independent student t-test, the pre-operative basal rCBF in the medial frontal gyrus, the post-operative CVRI in the frontal lobe and the parietal lobe of the hemispheres with EDAS and frontal EGS, the post-operative CVRI, and ${\Delta}CVRI$ showed a significant difference between patients with a good and poor clinical outcome (p<0.05). In a multivariate logistic regression analysis, the ${\Delta}CVRI$ and the post-operative CVRI of medial frontal gyrus on the hemispheres where EDAS with frontal EGS was performed were the significant predictive factors for the clinical outcome (p =0.002, p =0.015), Conclusion: With probabilistic map, we could objectively evaluate pre/post-operative hemodynamic changes of pediatric patients with moyamoya disease. Specifically the post-operative CVRI and the post-operative CVRI of medial frontal gyrus where EDAS with frontal EGS was done were the significant predictive factors for further clinical outcomes.

• The Journal of Korean Society for Radiation Therapy
• /
• v.26 no.1
• /
• pp.137-147
• /
• 2014

Purpose : The purpose of this study is to verify the accuracy of dose delivery according to the patient's breathing cycle in Gated Volumetric Modulated Arc Therapy Materials and Methods : TrueBeam STxTM(Varian Medical System, Palo Alto, CA) was used in this experiment. The Computed tomography(CT) images that were acquired with RANDO Phantom(Alderson Research Laboratories Inc. Stamford. CT, USA), using Computerized treatment planning system(Eclipse 10.0, Varian, USA), were used to create VMAT plans using 10MV FFF with 1500 cGy/fx (case 1, 2, 3) and 220 cGy/fx(case 4, 5, 6) of doserate of 1200 MU/min. The regular respiratory period of 1.5, 2.5, 3.5 and 4.5 sec and the patients respiratory period of 2.2 and 3.5 sec were reproduced with the $QUASAR^{TM}$ Respiratory Motion Phantom(Modus Medical Devices Inc), and it was set up to deliver radiation at the phase mode between the ranges of 30 to 70%. The results were measured at respective respiratory conditions by a 2-Dimensional ion chamber array detector(I'mRT Matrixx, IBA Dosimetry, Germany) and a MultiCube Phantom(IBA Dosimetry, Germany), and the Gamma pass rate(3 mm, 3%) were compared by the IMRT analysis program(OmniPro I'mRT system software Version 1.7b, IBA Dosimetry, Germany) Results : The gamma pass rates of Case 1, 2, 3, 4, 5 and 6 were the results of 100.0, 97.6, 98.1, 96.3, 93.0, 94.8% at a regular respiratory period of 1.5 sec and 98.8, 99.5, 97.5, 99.5, 98.3, 99.6% at 2.5 sec, 99.6, 96.6, 97.5, 99.2, 97.8, 99.1% at 3.5 sec and 99.4, 96.3, 97.2, 99.0, 98.0, 99.3% at 4.5 sec, respectively. When a patient's respiration was reproduced, 97.7, 95.4, 96.2, 98.9, 96.2, 98.4% at average respiratory period of 2.2 sec, and 97.3, 97.5, 96.8, 100.0, 99.3, 99.8% at 3.5 sec, respectively. Conclusion : The experiment showed clinically reliable results of a Gamma pass rate of 95% or more when 2.5 sec or more of a regular breathing period and the patient's breathing were reproduced. While it showed the results of 93.0% and 94.8% at a regular breathing period of 1.5 sec of Case 5 and 6, it could be confirmed that the accurate dose delivery could be possible on the most respiratory conditions because based on the results of 100 patients's respiratory period analysis as no one sustained a respiration of 1.5 sec. But, pretreatment dose verification should be precede because we can't exclude the possibility of error occurrence due to extremely short respiratory period, also a training at the simulation and careful monitoring are necessary for a patient to maintain stable breathing. Consequently, more reliable and accurate treatments can be administered.

• Journal of Distribution Research
• /
• v.16 no.4
• /
• pp.1-27
• /
• 2011

Introduction: Sales of private labels(PU have been growing m recent years. Globally, PLs have already achieved 20% share, although between 25 and 50% share in most of the European markets(AC. Nielson, 2005). These products are aimed to have comparable quality and prices as national brand(NB) products and have been continuously eroding manufacturer's national brand market share. Stores have also started introducing premium PLs that are of higher-quality and more reasonably priced compared to NBs. Worldwide, many retailers already have a multiple-tier private label architecture. Consumers as a consequence are now able to have a more diverse brand choice in store than ever before. Since premium PLs are priced higher than regular PLs and even, in some cases, above NBs, stores can expect to generate higher profits. Brand extensions and private label have been extensively studied in the marketing field. However, less attention has been paid to the private label extension. Therefore, this research focuses on private label extension using the Multi-Attribute Attitude Model(Fishbein and Ajzen, 1975). Especially there are few studies that consider the hierarchical effect of the PL's two parent brands: store brand and the original PL. We assume that the attitude toward each of the two parent brands affects the attitude towards the extended PL. The influence from each parent brand toward extended PL will vary according to the perceived fit between each parent brand and the extended PL. This research focuses on how these two parent brands act as reference points to one another in the consumers' choice consideration. Specifically we seek to understand how store image and attitude towards original PL affect consumer perceptions of extended premium PL. How consumers perceive extended premium PLs could provide strategic suggestions for retailer managers with specific suggestions on whether it is more effective: to position extended premium PL similarly or dissimilarly to original PL especially on the quality dimension and congruency with store image. There is an extensive body of research on branding and brand extensions (e.g. Aaker and Keller, 1990) and more recently on PLs(e.g. Kumar and Steenkamp, 2007). However there are no studies to date that look at the upgrading and influence of original PLs and attitude towards store on the premium PL extension. This research wishes to make a contribution to this gap using the perceived fit difference between parent brands and extended premium PL as the context. In order to meet the above objectives, we investigate which factors heighten consumers' positive attitude toward premium PL extension. Research Model and Hypotheses: When considering the attitude towards the premium PL extension, we expect four factors to have an influence: attitude towards store; attitude towards original PL; perceived congruity between the store image and the premium PL; perceived similarity between the original PL and the premium PL. We expect that all these factors have an influence on consumer attitude towards premium PL extension. Figure 1 gives the research model and hypotheses. Method: Data were collected by an intercept survey conducted on consumers at discount stores. 403 survey responses were attained (total 59.8% female, across all age ranges). Respondents were asked to respond to a series of Questions measured on 7 point likert-type scales. The survey consisted of Questions that measured: the trust towards store and the original PL; the satisfaction towards store and the original PL; the attitudes towards store, the original PL, and the extended premium PL; the perceived similarity of the original PL and the extended premium PL; the perceived congruity between the store image and the extended premium PL. Product images with specific explanations of the features of premium PL, regular PL and NB we reused as the stimuli for the Question response. We developed scales to measure the research constructs. Cronbach's alphaw as measured each construct with the reliability for all constructs exceeding the .70 standard(Nunnally, 1978). Results: To test the hypotheses, path analysis was conducted using LISREL 8.30. The path analysis for verification of the model produced satisfactory results. The validity index shows acceptable results(${\chi}^2=427.00$(P=0.00), GFI= .90, AGFI= .87, NFI= .91, RMSEA= .062, RMR= .047). With the increasing retailer use of premium PLBs, the intention of this research was to examine how consumers use original PL and store image as reference points as to the attitude towards premium PL extension. Results(see table 1 & 2) show that the attitude of each parent brand (attitudes toward store and original pL) influences the attitude towards extended PL and their perceived fit moderates these influences. Attitude toward the extended PL was influenced by the relative level of perceived fit. Discussion of results and future direction: These results suggest that the future strategy for the PL extension needs to consider that positive parent brand attitude is more strongly associated with the attitude toward PL extensions. Specifically, to improve attitude towards PL extension, building and maintaining positive attitude towards original PL is necessary. Positioning premium PL congruently to store image is also important for positive attitude. In order to improve this research, the following alternatives should also be considered. To improve the research model's predictive power, more diverse products should be included in study. Other attributes of product should also be included such as design, brand name since we only considered trust and satisfaction as factors to build consumer attitudes.

• Radiation Oncology Journal
• /
• v.23 no.3
• /
• pp.169-175
• /
• 2005

Purpose: Although computed tomography (CT) simulators are commonly used in radiation therapy department, many Institution still use conventional CT for treatments. In this study the setup errors that occur during simulation, CT scan (diagnostic CT scanner), and treatment were evaluated for the twenty one breast cancer patients. Materials and Methods: Errors were determined by calculating the differences in isocenter location, SSD, CLD, and locations of surgical clips implanted during surgery. The anatomic structures on simulation film and DRR image were compared to determine the movement of isocenter between simulation and CT scan. The isocetner point determined from the radio-opaque wires placed on patient's surface during CT scan was moved to new position if there was anatomic mismatch between the two images Results: In 7/21 patients, anatomic structures on DRR Image were different from the simulation Image thus new isocenter points were placed for treatment planning. The standard deviations of the diagnostic CT setup errors relative to the simulator setup in lateral, longitudinal, and anterior-posterior directions were 2.3, 1.6, and 1.6 mm, respectively. The average variation and standard deviation of SSD from AP field were 1.9 mm and 2.3 mm and from tangential fields were 2.8 mm and 3.7 mm. The variation of the CLD for the 21 patients ranged from 0 to 6 mm between simulation and DRR and 0 to 5 mm between simulation and treatment. The group systematic errors analyzed based on clip locations were 1.7 mm in lateral direction, 2.1 mm in AP direction, and 1.7 mm in SI direction. Conclusion: These results represent that there was no significant differences when SSD, CLD, clips' locations and isocenter locations were considered. Therefore, it is concluded that when a diagnostic CT scanner is used to acquire an image, the set-up variation is acceptable compared to using CT simulator for the treatment of breast cancer. However, the patient has to be positioned with care during CT scan in order to reduce the setup error between simulation and CT scan.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.13 no.1
• /
• pp.86-92
• /
• 2009

Purpose: The PET/CT has a clear distinction on the lesion of the functional image by adding anatomical information. It also could reduce the examination time using CT data as the attenuation-correction. When the stomach was contracted from a fast, it could bring a misinterpretation of the cancer of the lesion with a presence of physiological $^{18}F$-FDG uptake in stomach and it occasionally would bring an additional scan to confirm. To complement this shortcoming, the method that the patients had water before the examination to extend the stomach had been attempted. However, a short excretion time of the stomach did not give sufficiently extended image of the stomach. Then the patients had additional water and had the examination again. Therefore, the noticed fact is that the stomach excretion time depends on calories, protein content, and the level of carbohydrate. In this study, we use an orange juice to evaluate the extension of the stomach and usefulness of it. Materials and Methods: PET/CT scan were obtained on total 150 of patient from February 2008 to October2008, There were 3 groups in this study and each group had 50 patients. First group drank nothing, Second group drank water and third group drank orange juice. The patients (man 25, female 25) not drinking are the age of 30~71 years old (average: 54), the patients (man: 25, female: 25) drinking water (400 cc) are the age of 28~71 years old (average: 54) and the patients (man: 25, female: 25) drinking orange juice (400 cc) are the age of 32~74 years old (average: 56). The patients were fasted in 6-8 hours before the test, the patients were not diabetic. $^{18}F$-FDG 370~555 MBq were injected intravenously. The patients were in stable position for 1 hour, than the image was obtained. The patients drank water and other patients drank orange juice before Whole body scan. The image scan started from mid-femur to skull base. The emission scan acquired for three minutes per bed and the images were reconstructed. Stomach extension analysis is measured from vertical and horizontal length. Results: Stomach Extension was described as the vertical length of the Non Drink Group was $1.20{\pm}0.50\;cm$, horizontal length was $1.4{\pm}0.53\;cm$, the vertical length of the Water Drink Group was $1.67{\pm}0.63\;cm$, horizontal length was $1.65{\pm}0.77\;cm$, the vertical length of Orange juice Drink Group was $3.48{\pm}0.77\;cm$, horizontal length was $3.66{\pm}0.77\;cm$ in coronal image. Stomach Extension was described the vertical length of the Non Drink Group was $2.03{\pm}0.62\;cm$, horizontal length was $1.69{\pm}0.68\;cm$, the vertical length of Water Drink Group was $5.34{\pm}1.62\;cm$, horizontal length was $2.45{\pm}0.72\;cm$, the vertical length of Orange juice Drink Group was $7.74{\pm}1.62\;cm$, horizontal length was $3.57{\pm}0.77\;cm$ in transverse image. The Stomach Extension has specific differences (p<0.001). The SUVs shows the Non Drink Group were measured as Liver $2.52{\pm}0.42$, Lung $0.51{\pm}0.14$, the Water Drink Group were measured as Liver $2.47{\pm}0.38$, Lung $0.50{\pm}0.14$, Orange juice Drink Group were measured as Liver $2.47{\pm}0.38$, Lung $0.50{\pm}0.14$. The SUVs did not have specific differences (p>0.759). Conclusions: There was not a large difference of SUV in three groups. When the patients drank Orange juice and water, the range extension of stomach was higher than without drinking nothing and it was possible to acquire fully extended images. Therefore, it will be possible that unnecessary additional stomach scans will be reduced by drinking orange juice before the examination so that the patients' claim from uncomfortable and long period of fast will be minimized.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.