Transfer Learning based DNN-SVM Hybrid Model for Breast Cancer Classification
-
- Journal of the Korea Society of Computer and Information
- /
- v.28 no.11
- /
- pp.1-11
- /
- 2023
Breast cancer is the disease that affects women the most worldwide. Due to the development of computer technology, the efficiency of machine learning has increased, and thus plays an important role in cancer detection and diagnosis. Deep learning is a field of machine learning technology based on an artificial neural network, and its performance has been rapidly improved in recent years, and its application range is expanding. In this paper, we propose a DNN-SVM hybrid model that combines the structure of a deep neural network (DNN) based on transfer learning and a support vector machine (SVM) for breast cancer classification. The transfer learning-based proposed model is effective for small training data, has a fast learning speed, and can improve model performance by combining all the advantages of a single model, that is, DNN and SVM. To evaluate the performance of the proposed DNN-SVM Hybrid model, the performance test results with WOBC and WDBC breast cancer data provided by the UCI machine learning repository showed that the proposed model is superior to single models such as logistic regression, DNN, and SVM, and ensemble models such as random forest in various performance measures.
A determination method of aromatic amino acids such as trytophan (Trp), tyrosine (Tyr), and phenylalanine (Phe) using luminol-
Since 11 September 2001, warnings of risk in the nexus of terrorism and nuclear weapons and materials which poses one of the gravest threats to the international community have continued. The purpose of this study is to analyze the aim, principles, characteristics, activities, impediments to progress and developmental recommendation of the Global Initiative to Combat Nuclear Terrorism(GICNT). In addition, it suggests implications of the GICNT for the ROK policy. International community will need a comprehensive strategy with four key elements to accomplish the GICNT: (1) securing and reducing nuclear stockpiles around the world, (2) countering terrorist nuclear plots, (3) preventing and deterring state transfers of nuclear weapons or materials to terrorists, (4) interdicting nuclear smuggling. Moreover, other steps should be taken to build the needed sense of urgency, including: (1) analysis and assessment through joint threat briefing for real nuclear threat possibility, (2) nuclear terrorism exercises, (3) fast-paced nuclear security reviews, (4) realistic testing of nuclear security performance to defeat insider or outsider threats, (5) preparing shared database of threats and incidents. As for the ROK, main concerns are transfer of North Korea's nuclear weapons, materials and technology to international terror groups and attacks on nuclear facilities and uses of nuclear devices. As the 5th nuclear country, the ROK has strengthened systems of physical protection and nuclear counterterrorism based on the international conventions. In order to comprehensive and effective prevention of nuclear terrorism, the ROK has to strengthen nuclear detection instruments and mobile radiation monitoring system in airports, ports, road networks, and national critical infrastructures. Furthermore, it has to draw up effective crisis management manual and prepare nuclear counterterrorism exercises and operational postures. The fundamental key to the prevention, detection and response to nuclear terrorism which leads to catastrophic impacts is to establish not only domestic law, institution and systems, but also strengthen international cooperation.
Species identification of animal tissues in meat products is an important issue to protect the consumer from illegal and/or undesirable adulteration; for economic, religious and health reasons. In this reason, accurate analytical methods are needed for the labeling of meat products with requiring simple and fast procedure. Recently, applications of PCR in food analysis have been increased because of their simplicity, specificity and sensitivity. Therefore, in this study, a multiplex PCR assay was developed for the simultaneous identification of eight species of cow, pig, chicken, duck, goat, sheep, horse and turkey from raw meats. The primers were designed in different regions of mitochondrial 16S RNA after alignment of the available sequences in the GenBank database. Two multiplex primer sets were designed as Set 1 (cow, pig, chicken, duck) and Set 2 (goat, sheep, horse, turkey), respectively. Total 274 samples from cow (n = 55), pig (n = 30), chicken (n=30), and duck (n = 30), goat (n = 40), sheep (n = 33), horse (n = 41), and turkey (n = 15) were tested. The primers generated specific fragments of 94, 192, 279, 477 bp (pig, chicken, cow, duck), 670, 271, 152, 469 bp (goat, sheep, horse, turkey) lengths for eight species, respectively. The animal species specificity was 100% in all eight samples in the multiplex PCR assay. The detection limit of the multiplex PCR assay showed from 100 fg to 1 pg of template DNA from extracted from raw meats. When applying multiplex PCR assays to sample from pork/beef and pork/chicken, beef/chicken tested raw mixed meats and heat-treated (
Owing to the risk of fetal loss associated with prenatal diagnostic procedures (amniocentesis, chorionic villus sampling), noninvasive prenatal diagnosis (NIPD) is ultimate goal of prenatal diagnosis. The discovery of circulating cell-free fetal DNA (cffDNA) in maternal plasma in 1997 has opened up new probabilities for NIPD by Dr. Lo et al. The last decade has seen great development in NIPD. Fetal sex and fetal RhD status determination by cffDNA analysis is already in clinical use in certain countries. For routine use, this test is limited by the amount of cell-free maternal DNA in blood sample, the lack of universal fetal markers, and appropriate reference materials. To improve the accuracy of detection of fetal specific sequences in maternal plasma, internal positive controls to confirm to presence of fetal DNA should be analyzed. We have developed strategies for noninvasive determination of fetal gender, and fetal RhD genotyping using cffDNA in maternal plasma, using real-time quantitative polymerase chain reaction (RT-PCR) including RASSF1A epigenetic fetal DNA marker (gender-independent) as internal positive controls, which is to be first successful study of this kind in Korea. In our study, accurate detection of fetal gender through gestational age, and fetal RhD genotyping in RhD-negative pregnant women was achieved. In this assay, we show that the assay is sensitive, easy, fast, and reliable. These developments improve the reliability of the applications of circulating fetal DNA when used in clinical practice to manage sex-linked disorders (e.g., hemophilia, Duchenne muscular dystrophy), congenital adrenal hyperplasia (CAH), RhD incompatibility, and the other noninvasive pregnant diagnostic tests on the coming soon. The study was the first successful case in Korea using cffDNA in maternal plasma, which has created a new avenue for clinical applications of NIPD.
Background : Since the advent of AIDS, tuberculosis has become a major public health problem in the western society. Therefore, it is essential that pulmonary tuberculosis be rapidly diagnosed. Light microscopic detection of acid-fast organisms in sputum has traditionally been used for rapidly diagnosing tuberculosis. However positive smears are only observed in about one-half to three-quarters of cases. Studies using PCR for diagnosing pulmonary tuberculosis disclosed several shortcomings suggesting an inability to distinguish between active and treated or inactive tuberculosis. In this study, the clinical significance of a PCR-based rapid technique for detecting Mycobacterium tuberculosis DNA in peripheral blood was investigated. Materials and Methods : From July 1, 1998 through to August 30, 1999, 59 patients with presumed tuberculosis, who had no previous history of anti-tuberculosis medication use within one year prior to this study were recruited and followed up for more than 3 months. AFB stain and culture in the sputum and/or pleural fluids and biopsies when needed were performed. Blood samples from each of the 59 patients were obtained in order to identify Mycobacterium Tuberculosis DNA by a PCR test. Results : 1) Forty five out of 59 patients had a final diagnosis of tuberculosis ; Twenty eight were confirmed as having active pulmonary tuberculosis by culture or biopsy. Four were clinically diagnosed with pulmonary tuberculosis. The other 13 patients were diagnosed as having tuberculous pleurisy (9) and extrapulmonary tuberculosis (4). 2) Fourteen patients showed a positive blood PCR test. The PCR assay correctly identified active tuberculosis in 13 out of 14 patients. The overall sensitivity and specificity of this blood peR assay for diagnosing tuberculosis were 29% and 93%, respectively. The positive predictive value was 93%, the negative predictive value was 29% and the diagnostic accuracy was 44%.3) Six out of 14(43%) patients with blood PCR positive tuberculosis were immunologically compromised hosts. 4) A simple chest radiograph in blood PCR positive tuberculosis patients showed variable and inconsistent findings. Conclusion : A peripheral blood PCR assay for Mycobacterium tuberculosis is not recommended as a screening method for diagnosing active tuberculosis. However, it was suggested that the blood PCR assay could contribute to an early diagnostic rate due to its high positive predictive value.
The UET(ultrasound excited thermography) for the ,eat-time diagnostics of the object employs an infrared camera to image defects of the surface and subsurface which are locally heated using high-frequency putted ultrasonic excitation. The dissipation of high-power ultrasonic energy around the feces of the defects causes an increase In temperature. The defect's image appears as a hot spot (bright IR source) within a dark background field. The UET for nondestructive diagnostic and evaluation is based on the image analysis of the hot spot as a local response to ultrasonic excited heat deposition. In this paper the applicability of VET for fast imaging of defect is described. The ultrasonic energy is injected into the sample through a transducer in the vertical and horizontal directions respectively. The voltage applied to the transducer is measured by digital oscilloscope, and the waveform are compared. Measurements were performed on four kinds of materials: SUS fatigue crack specimen(thickness 14mm), PCB plate(1.8 mm), CFRP plate(3 mm) and Inconel 600 plate (1 mm). A high power ultrasonic energy with pulse durations of 250ms Is injected into the samples in the horizontal and vertical directions respectively The obtained experimental result reveals that the dissipation loss of the ultrasonic energy In the vertical injection is less than that in the horizontal direction. In the cafe or PCB, CFRP, the size of hot spot in the vortical injection if larger than that in horizontal direction. Duration time of the hot spot in the vertical direction is three times as long as that in the horizontal direction. In the case of Inconel 600 plate and SUS sample, the hot spot in the horizontal injection was detected faster than that in the vertical direction
The wall shear stress in the vicinity of end-to end anastomoses under steady flow conditions was measured using a flush-mounted hot-film anemometer(FMHFA) probe. The experimental measurements were in good agreement with numerical results except in flow with low Reynolds numbers. The wall shear stress increased proximal to the anastomosis in flow from the Penrose tubing (simulating an artery) to the PTFE: graft. In flow from the PTFE graft to the Penrose tubing, low wall shear stress was observed distal to the anastomosis. Abnormal distributions of wall shear stress in the vicinity of the anastomosis, resulting from the compliance mismatch between the graft and the host artery, might be an important factor of ANFH formation and the graft failure. The present study suggests a correlation between regions of the low wall shear stress and the development of anastomotic neointimal fibrous hyperplasia(ANPH) in end-to-end anastomoses. 30523 T00401030523 ^x Air pressure decay(APD) rate and ultrafiltration rate(UFR) tests were performed on new and saline rinsed dialyzers as well as those roused in patients several times. C-DAK 4000 (Cordis Dow) and CF IS-11 (Baxter Travenol) reused dialyzers obtained from the dialysis clinic were used in the present study. The new dialyzers exhibited a relatively flat APD, whereas saline rinsed and reused dialyzers showed considerable amount of decay. C-DAH dialyzers had a larger APD(11.70
The wall shear stress in the vicinity of end-to end anastomoses under steady flow conditions was measured using a flush-mounted hot-film anemometer(FMHFA) probe. The experimental measurements were in good agreement with numerical results except in flow with low Reynolds numbers. The wall shear stress increased proximal to the anastomosis in flow from the Penrose tubing (simulating an artery) to the PTFE: graft. In flow from the PTFE graft to the Penrose tubing, low wall shear stress was observed distal to the anastomosis. Abnormal distributions of wall shear stress in the vicinity of the anastomosis, resulting from the compliance mismatch between the graft and the host artery, might be an important factor of ANFH formation and the graft failure. The present study suggests a correlation between regions of the low wall shear stress and the development of anastomotic neointimal fibrous hyperplasia(ANPH) in end-to-end anastomoses. 30523 T00401030523 ^x Air pressure decay(APD) rate and ultrafiltration rate(UFR) tests were performed on new and saline rinsed dialyzers as well as those roused in patients several times. C-DAK 4000 (Cordis Dow) and CF IS-11 (Baxter Travenol) reused dialyzers obtained from the dialysis clinic were used in the present study. The new dialyzers exhibited a relatively flat APD, whereas saline rinsed and reused dialyzers showed considerable amount of decay. C-DAH dialyzers had a larger APD(11.70
In this paper, we suggest an application system architecture which provides accurate, fast and efficient automatic gasometer reading function. The system captures gasometer image using mobile device camera, transmits the image to a cloud server on top of private LTE network, and analyzes the image to extract character information of device ID and gas usage amount by selective optical character recognition based on deep learning technology. In general, there are many types of character in an image and optical character recognition technology extracts all character information in an image. But some applications need to ignore non-of-interest types of character and only have to focus on some specific types of characters. For an example of the application, automatic gasometer reading system only need to extract device ID and gas usage amount character information from gasometer images to send bill to users. Non-of-interest character strings, such as device type, manufacturer, manufacturing date, specification and etc., are not valuable information to the application. Thus, the application have to analyze point of interest region and specific types of characters to extract valuable information only. We adopted CNN (Convolutional Neural Network) based object detection and CRNN (Convolutional Recurrent Neural Network) technology for selective optical character recognition which only analyze point of interest region for selective character information extraction. We build up 3 neural networks for the application system. The first is a convolutional neural network which detects point of interest region of gas usage amount and device ID information character strings, the second is another convolutional neural network which transforms spatial information of point of interest region to spatial sequential feature vectors, and the third is bi-directional long short term memory network which converts spatial sequential information to character strings using time-series analysis mapping from feature vectors to character strings. In this research, point of interest character strings are device ID and gas usage amount. Device ID consists of 12 arabic character strings and gas usage amount consists of 4 ~ 5 arabic character strings. All system components are implemented in Amazon Web Service Cloud with Intel Zeon E5-2686 v4 CPU and NVidia TESLA V100 GPU. The system architecture adopts master-lave processing structure for efficient and fast parallel processing coping with about 700,000 requests per day. Mobile device captures gasometer image and transmits to master process in AWS cloud. Master process runs on Intel Zeon CPU and pushes reading request from mobile device to an input queue with FIFO (First In First Out) structure. Slave process consists of 3 types of deep neural networks which conduct character recognition process and runs on NVidia GPU module. Slave process is always polling the input queue to get recognition request. If there are some requests from master process in the input queue, slave process converts the image in the input queue to device ID character string, gas usage amount character string and position information of the strings, returns the information to output queue, and switch to idle mode to poll the input queue. Master process gets final information form the output queue and delivers the information to the mobile device. We used total 27,120 gasometer images for training, validation and testing of 3 types of deep neural network. 22,985 images were used for training and validation, 4,135 images were used for testing. We randomly splitted 22,985 images with 8:2 ratio for training and validation respectively for each training epoch. 4,135 test image were categorized into 5 types (Normal, noise, reflex, scale and slant). Normal data is clean image data, noise means image with noise signal, relfex means image with light reflection in gasometer region, scale means images with small object size due to long-distance capturing and slant means images which is not horizontally flat. Final character string recognition accuracies for device ID and gas usage amount of normal data are 0.960 and 0.864 respectively.