• Asian Pacific Journal of Cancer Prevention
• /
• 제16권8호
• /
• pp.3293-3299
• /
• 2015

Background: Arm protein lost in epithelial cancers, on chromosome X (ALEX) is a novel subgroup within the armadillo (ARM) family, which has one or two ARM repeat domains as opposed to more than six-thirteen repeats in the classical Armadillo family members. Materials and Methods: In the study, we explore the biological functions of ALEX1 in breast cancer cells. Overexpression of ALEX1 and silencing of ALEX1 were performed with SK-BR3 and MCF-7 cell lines. Cell proliferation and colony formation assays, along with flow cytometry, were carried out to evaluate the roles of ALEX1. Results: ALEX1 overexpression in SK-BR3 breast cancer cells inhibited proliferation and induced apoptosis. Furthermore, depletion of ALEX1 in MCF-7 breast cancer cells increased proliferation and inhibited apoptosis. Additional analyses demonstrated that the overexpression of ALEX1 activated the intrinsic apoptosis cascades through up-regulating the expression of Bax, cytosol cytochrome c, active caspase-9 and active caspase-3 and down-regulating the levels of Bcl-2 and mitochondria cytochrome c. Simultaneouly, silencing of ALEX1 inhibited intrinsic apoptosis cascades through down-regulating the expression of Bax, cytosol cytochrome c, active caspase-9, and active caspase-3 and up-regulating the level of Bcl-2 and mitochondria cytochrome c. Conclusions: Our data suggest that ALEX1 as a crucial tumor suppressor gene has been involved in cell proliferation and apoptosis in breast cancer, which may serve as a novel candidate therapeutic target.

• Medical Lasers
• /
• 제8권1호
• /
• pp.24-27
• /
• 2019

Seborrheic keratoses (SKs) have been treated with non-ablative longpulsed (LP) lasers, including LP 532-nm neodymium (Nd): yttrium aluminum garnet (YAG), LP 695-nm ruby, LP 755-nm alexandrite (Alex), and LP 1,064-nm Nd:YAG lasers, with a pulse durations of 1-300 msec. Dual-wavelength LP 755-nm Alex/1,064-nm Nd:YAG laser systems have been used to remove hair follicles and treat various vascular and pigmented disorders by sequentially delivering two pulses of different wavelengths with interpulse intervals in the millisecond range. This paper reports the case of a female patient with multiple, discrete, light brown SKs on the dorsum of both hands that were treated effectively with one session of dual-wavelength LP 1,064-nm Nd:YAG/755-nm Alex laser treatment. The treatment settings for the LP Nd:YAG laser were comprised of a wavelength of 1,064 nm, fluence of 50 J/cm², pulse duration of 5 msec, and beam size of 3 mm. The settings for the LP Alex laser were comprised of a wavelength of 755 nm, fluence of 50 J/cm², pulse duration of 5 msec, and beam size of 3 mm. A hybrid mode was used to automatically deliver LP Nd:YAG and LP Alex laser pulses in succession at interpulse intervals of 20 msec. Six weeks after treatment, the patient exhibited remarkable improvement of the light brown seborrheic keratoses and was satisfied with the results.

• Journal of Information Processing Systems
• /
• 제14권1호
• /
• pp.205-217
• /
• 2018

Deep learning using convolutional neural networks (CNNs) is being studied in various fields of image recognition and these studies show excellent performance. In this paper, we compare the performance of CNN architectures, KCR-AlexNet and KCR-GoogLeNet. The experimental data used in this paper is obtained from PHD08, a large-scale Korean character database. It has 2,187 samples of each Korean character with 2,350 Korean character classes for a total of 5,139,450 data samples. In the training results, KCR-AlexNet showed an accuracy of over 98% for the top-1 test and KCR-GoogLeNet showed an accuracy of over 99% for the top-1 test after the final training iteration. We made an additional Korean character dataset with fonts that were not in PHD08 to compare the classification success rate with commercial optical character recognition (OCR) programs and ensure the objectivity of the experiment. While the commercial OCR programs showed 66.95% to 83.16% classification success rates, KCR-AlexNet and KCR-GoogLeNet showed average classification success rates of 90.12% and 89.14%, respectively, which are higher than the commercial OCR programs' rates. Considering the time factor, KCR-AlexNet was faster than KCR-GoogLeNet when they were trained using PHD08; otherwise, KCR-GoogLeNet had a faster classification speed.

• 천문학회보
• /
• 제42권2호
• /
• pp.60.1-60.1
• /
• 2017

In this study we apply Convolution Neural Network(CNN) to solar flare occurrence prediction with various parameter options using the 00:00 UT MDI images from 1996 to 2010 (total 4962 images). We assume that only X, M and C class flares correspond to "flare occurrence" and the others to "non-flare". We have attempted to look for the best options for the models with two CNN pre-trained models (AlexNet and GoogLeNet), by modifying training images and changing hyper parameters. Our major results from this study are as follows. First, the flare occurrence predictions are relatively good with about 80 % accuracies. Second, both flare prediction models based on AlexNet and GoogLeNet have similar results but AlexNet is faster than GoogLeNet. Third, modifying the training images to reduce the projection effect is not effective. Fourth, skill scores of our flare occurrence model are mostly better than those of the previous models.

• 스마트미디어저널
• /
• 제12권7호
• /
• pp.52-58
• /
• 2023

Alzheimer's disease is one type of dementia, the symptoms can be treated by detecting the disease at its early stages. Recently, many computer-aided diagnosis using magnetic resonance image(MRI) have shown a good results in the classification of AD. Taken these MRI images and feed to Free surfer software to extra the features. In consideration, using T1-weighted images and classifying using the convolution neural network (CNN) model are proposed. In this paper, taking the subjects from ADNI of subcortical and cortical features of 190 subjects. Consider the study to reduce the complexity of the model by using the single layer in the Res-Net, VGG, and Alex Net. Multi-class classification is used to classify four different stages, CN, EMCI, LMCI, AD. The following experiment shows for respective classification Res-Net, VGG, and Alex Net with the best accuracy with VGG at 96%, Res-Net, GoogLeNet and Alex Net at 91%, 93% and 89% respectively.

• 인터넷정보학회논문지
• /
• 제18권5호
• /
• pp.61-67
• /
• 2017

본 논문에서는 딥러닝의 회선신경망을 이용한 실제 소셜 네트워크 상의 이미지 분류가 얼마나 효과적인지 알아보기 위한 실험을 수행하고, 그 결과와 그를 통해 알게 된 교훈에 대해 소개한다. 이를 위해 ImageNet Large Scale Visual Recognition Challenge(ILSVRC)의 2012년 대회와 2015년 대회에서 각각 우승을 차지한 AlexNet 모델과 ResNet 모델을 이용하였다. 평가를 위한 테스트 셋으로 인스타그램에서 수집한 이미지를 사용하였으며, 12개의 카테고리, 총 240개의 이미지로 구성되어 있다. 또한, Inception V3모델을 이용하여 fine-tuning을 실시하고, 그 결과를 비교하였다. AlexNet과 ResNet, Inception V3, fine-tuned Inception V3 이 네 가지 모델에 대한 Top-1 error rate들은 각각 49.58%, 40.42%, 30.42% 그리고 5.00%로 나타났으며, Top-5 error rate들은 각각 35.42%, 25.00%, 20.83% 그리고 0.00%로 나타났다.

• 전기전자학회논문지
• /
• 제25권4호
• /
• pp.664-671
• /
• 2021

본 논문은 합성곱 신경망에 데이터 재사용 방법을 효과적으로 적용하여 연산 횟수와 메모리 접근 횟수를 줄일 수 있는 GPGPU구조를 제안한다. 합성곱은 kernel과 입력 데이터를 이용한 2차원 연산으로 kernel이 slide하는 방법으로 연산이 이루어 진다. 이때, 합성곱 연산이 완료될 때 까지 kernel을 캐시메모리로 부터 전달 받는 것이 아니고 내부 레지스터를 이용하는 재사용 방법을 제안한다. SIMT방법으로 명령어가 실행되는 GPGPU의 원리 이용하여 데이터 재사용의 효과를 높이기 위해 합성곱에 직렬 연산 방식을 적용하였다. 본 논문에서는 레지스터기반 데이터 재사용을 위하여 kernel을 4×4로 고정하고 이를 효과적으로 지원하기 위한 warp 크기와 레지스터 뱅크를 갖는 GPGPU를 설계하였다. 설계된 GPGPU의 합성곱 신경망에 대한 성능을 검증하기 위해 FPGA로 구현한 뒤 LeNet을 실행시키고 TensorFlow를 이용한 비교 방법으로 AlexNet에 대한 성능을 측정하였다. 측정결과 AlexNet기준 1회 학습 속도는 0.468초이며 추론 속도는 0.135초이다.

• 대한원격탐사학회지
• /
• 제36권2_2호
• /
• pp.293-307
• /
• 2020

매년 황해와 동중국해에서는 대형 부유조류인 녹조와 갈조가 대량으로 발생하고 있다. 이러한 대형 부유조류는 연안의 양식 시설물이나 해변으로 유입되며, 제거하는데 막대한 경제적 손실을 발생시킨다. 현재는 연안으로 유입되는 대형 부유조류를 탐지하기 위해 원격탐사 방법이 활발하게 사용되고 있다. 그러나 대형 부유조류는 해양의 다양한 대상들과 중첩되는 파장이 존재하기에 이를 정확하게 탐지하는데 한계가 있다. 더욱이 녹조와 갈조는 유사한 스펙트럼 특성을 보이기 때문에 원격탐사 자료를 이용한 구분을 더욱 어렵게 만든다. 따라서 본 연구에서는 위성 영상에 딥러닝 기법을 적용하여 녹조와 갈조를 효과적으로 구분하고자 하였다. 이를 위한 네트워크를 결정하기 위해 최적의 학습 조건을 찾아 AlexNet 신경망을 전이 학습하였으며, 학습과 검증을 위해 Gaofen-1 WFV 영상을 이용하여 데이터셋을 구성하였다. 최적의 학습 조건으로 학습된 네트워크를 이용하여 실험 데이터에 대한 결과를 확인하였다. 그 결과 실험 데이터에 대한 정확도는 88.89%를 보였으며, 녹조와 갈조에 대해 각각 66.67%와 100%의 정밀도로 구분이 가능하였다. 이는 전이 학습된 AlexNet 신경망이 녹조와 갈조의 미세한 차이를 구분할 수 있는 것으로 해석된다. 본 연구를 통해 해양의 다양한 대상으로부터 녹조와 갈조를 효과적으로 분류하고 각각 구분할 수 있을 것으로 기대된다.

• 한국인구학
• /
• 제17권1호
• /
• pp.115-126
• /
• 1994

• 한국미생물·생명공학회지
• /
• 제47권4호
• /
• pp.621-629
• /
• 2019

In this study we attempted to screen bacteria and fungi that generate electricity while treating wastewater using optimized double-chamber microbial fuel cell (MFC) system parameters. Optimization was carried out for five best exoelectrogenic isolates (two bacteria and three fungi) at pH values of 6.0, 7.5, 8.5, and 9.5, and temperatures of 30, 35, 40, and 45℃; the generated power densities were measured using a digital multimeter (DT9205A). The isolates were identified using molecular characterization, followed by the phylogenetic analysis of isolates with known exoelectrogenic microorganisms. The bacterium, Proteus species, N6 (KX548358.1) and fungus, Candida parapsilosis, S10 (KX548360) produced the highest power densities of 1.59 and 1.55 W/m² (at a pH of 8.5 and temperatures of 35 and 40℃) within 24 h, respectively. Other fungi-Clavispora lusitaniae, S9 (KX548359.1) at 40℃, Clavispora lusitaniae, S14 (KX548361.1) at 35℃-and bacterium-Providencia species, N4 (KX548357.1) at 40℃-produced power densities of 1.51, 1.46, and 1.44 W/m², respectively within 24 h. The MFCs achieved higher power densities at a pH of 8.5, temperature of 40℃ within 24 h. The bacterial isolates have a close evolutionary relationship with other known exoelectrogenic microorganisms. These findings helped us determine the optimal pH, temperature, evolutionary relationship, and exoelectrogenic fungal species other than bacteria that enhance MFC performance.