• International Journal of Computer Science & Network Security
• /
• 제22권12호
• /
• pp.98-106
• /
• 2022

A condition of the retina known as hypertensive retinopathy (HR) is connected to high blood pressure. The severity and persistence of hypertension are directly correlated with the incidence of HR. To avoid blindness, it is essential to recognize and assess HR as soon as possible. Few computer-aided systems are currently available that can diagnose HR issues. On the other hand, those systems focused on gathering characteristics from a variety of retinopathy-related HR lesions and categorizing them using conventional machine-learning algorithms. Consequently, for limited applications, significant and complicated image processing methods are necessary. As seen in recent similar systems, the preciseness of classification is likewise lacking. To address these issues, a new CAD HR-diagnosis system employing the advanced Deep Dense CNN Learning (DD-CNN) technology is being developed to early identify HR. The HR-diagnosis system utilized a convolutional neural network that was previously trained as a feature extractor. The statistical investigation of more than 1400 retinography images is undertaken to assess the accuracy of the implemented system using several performance metrics such as specificity (SP), sensitivity (SE), area under the receiver operating curve (AUC), and accuracy (ACC). On average, we achieved a SE of 97%, ACC of 98%, SP of 99%, and AUC of 0.98. These results indicate that the proposed DD-CNN classifier is used to diagnose hypertensive retinopathy.

• Nuclear Engineering and Technology
• /
• 제55권1호
• /
• pp.100-108
• /
• 2023

In this study, a neural network model inspired by a one-dimensional convolution U-net is developed to automatically accelerate edge localized mode (ELM) detection from big diagnostic data of fusion devices and increase the detection accuracy regardless of the hyperparameter setting. This model recognizes the input signal patterns and overcomes the problems of existing detection algorithms, such as the prominence algorithm and those of differential methods with high sensitivity for the threshold and signal intensity. To train the model, 10 sets of discharge radiation data from the KSTAR are used and sliced into 11091 inputs of length 12 ms, of which 20% are used for validation. According to the receiver operating characteristic curves, our model shows a positive prediction rate and a true prediction rate of approximately 90% each, which is comparable to the best detection performance afforded by other algorithms using their optimized hyperparameters. The accurate and automatic ELM-burst detection methodology used in our model can be beneficial for determining plasma properties, such as the ELM frequency from big data measured in multiple experiments using machines from the KSTAR device and ITER. Additionally, it is applicable to feature detection in the time-series data of other engineering fields.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2021년도 추계학술발표대회
• /
• pp.805-807
• /
• 2021

In-hospital cardiac arrest is a significant problem for medical systems. Although the traditional early warning systems have been widely applied, they still contain many drawbacks, such as the high false warning rate and low sensitivity. This paper proposed a strategy that involves a deep learning approach based on a novel interpretable deep tabular data learning architecture, named TabNet, for the Rapid Response System. This study has been processed and validated on a dataset collected from two hospitals of Chonnam National University, Korea, in over 10 years. The learning metrics used for the experiment are the area under the receiver operating characteristic curve score (AUROC) and the area under the precision-recall curve score (AUPRC). The experiment on a large real-time dataset shows that our method improves compared to other machine learning-based approaches.

• 한국음향학회지
• /
• 제26권6호
• /
• pp.227-234
• /
• 2007

본 논문은 CP(Cyclic Prefix)를 사용한 SC-FDE(Single Carrier with Frequency Domain Equalization) 기반의 전송 방식에 대해 기술하였다. SC-FBE 방식은 OFDM(Orthogonal Frequency Division Multiplexing)과 기능 구성이 유사하며, OFDM과 유사하게 주파수 영역에서 블록 단위로 효율적인 등화 수행이 가능하다. 특별히 단일 반송파 방식은 OFDM에 비해 비선형 왜곡에 상대적으로 덜 민감한 장점을 가지고 있다. 본 논문에서는 결정 지향 방식의 SC-FDE 수신부를 설계하고, 시뮬레이션을 통해 성능 결과를 제시하였다.

• Investigative Magnetic Resonance Imaging
• /
• 제21권3호
• /
• pp.154-161
• /
• 2017

Purpose: To evaluate the diagnostic performance of diffusion-weighted steady-state free precession (DW-SSFP) in comparison to diffusion-weighted echo-planar imaging (DW-EPI) for differentiating the neoplastic and benign osteoporotic vertebral compression fractures. Materials and Methods: The subjects were 40 patients with recent vertebral compression fractures but no history of vertebroplasty, spine operation, or chemotherapy. They had received 3-Tesla (T) spine magnetic resonance imaging (MRI), including both DW-SSFP and DW-EPI sequences. The 40 patients included 20 with neoplastic vertebral fracture and 20 with benign osteoporotic vertebral fracture. In each fracture lesion, we obtained the signal intensity normalized by the signal intensity of normal bone marrow (SI norm) on DW-SSFP and the apparent diffusion coefficient (ADC) on DW-EPI. The correlation between the SI norm and the ADC in each lesion was analyzed using linear regression. The optimal cut-off values for the diagnosis of neoplastic fracture were determined in each sequence using Youden's J statistics and receiver operating characteristic curve analyses. Results: In the neoplastic fracture, the median SI norm on DW-SSFP was higher and the median ADC on DW-EPI was lower than the benign osteoporotic fracture (5.24 vs. 1.30, P = 0.032, and 0.86 vs. 1.48, P = 0.041, respectively). Inverse linear correlations were evident between SI norm and ADC in both neoplastic and benign osteoporotic fractures (r = -0.45 and -0.61, respectively). The optimal cut-off values for diagnosis of neoplastic fracture were SI norm of 3.0 in DW-SSFP with the sensitivity and specificity of 90.4% (95% confidence interval [CI]: 81.0-99.0) and 95.3% (95% CI: 90.0-100.0), respectively, and ADC of 1.3 in DW-EPI with the sensitivity and specificity of 90.5% (95% CI: 80.0-100.0) and 70.4% (95% CI: 60.0-80.0), respectively. Conclusion: In 3-T MRI, DW-SSFP has comparable sensitivity and specificity to DW-EPI in differentiating the neoplastic vertebral fracture from the benign osteoporotic vertebral fracture.

• 한국콘텐츠학회논문지
• /
• 제13권12호
• /
• pp.829-839
• /
• 2013

본 연구는 로지스틱 회귀분석과 의사결정나무 분석을 활용하여 일 대도시 주민의 우울에 영향을 주는 요인을 예측하고 비교하고자 시도된 서술적 조사연구이다. 연구대상은 20세에서 65세 미만의 일 대도시 주민 462명이었다. 자료 수집은 2011년 10월 7일부터 10월 21일까지이었으며, 자료 분석은 SPSS 18.0 프로그램을 이용하여 빈도, 백분율, 평균과 표준편차 및 ${\chi}^2$-test, t-test, 로지스틱 회귀분석, roc curve, 의사결정나무 분석으로 분석하였다. 본 연구 결과, 로지스틱 회귀분석과 의사결정나무 분석에서 공통적으로 나타난 우울 예측요인은 사회부적응, 주관적 신체증상 및 가족 지지이었다. 로지스틱 회귀분석에서 특이도 93.8%, 민감도 42.5%이었고, 본 연구의 모형 적합도를 roc curve 검증 한 결과 AUC=.84으로 본 연구 모형은 적합(p=<.001)하다고 할 수 있다. 우울예측에 대한 의사결정나무 분석은 분류에 대한 예측 정확도에서 특이도 98.3%, 민감도 20.8%이었고, 전체 분류 정확도는 로지스틱 회귀분석은 82.0%, 의사결정나무 분석은 80.5% 이었다. 본 연구 결과 민감성과 분류 정확도와 더 높게 나타난 로지스틱 회귀분석 방법이 지역 주민의 우울 예측 모형을 구축하는데 더 유용한 자료로 사용될 수 있으리라 사료된다.

• Asian Pacific Journal of Cancer Prevention
• /
• 제16권14호
• /
• pp.5889-5897
• /
• 2015

Background: The performance of combined testing visual inspection with acetic acid (VIA) and cervical cytology tests might differ from one setting to another. The average estimate of the testing accuracy across studies is informative, but no meta-analysis has been carried out to assess this combined method. Objective: The objective of this study was to estimate the average sensitivity and specificity of the combined VIA and cervical cytology tests for the detection of cervical precancerous lesions. Materials and Methods: We conducted a systematic review and a meta-analysis, according to the Cochrane Handbook for Systematic Review of Diagnostic Test Accuracy. We considered two cases. In the either-positive result case, a positive result implies positivity in at least one of the tests. A negative result implies negativity in both tests. In the both-positive case, a positive result implies having both tests positive. Eligible studies were identified using Pubmed, Embase, Website of Science, CINHAL and COCRANE databases. True positive, false positive, false negative and true negative values were extracted. Estimates of sensitivity and specificity, positive and negative likelihood (LR) and diagnostic odds ratios (DOR) were pooled using a hierarchical random effect model. Hierarchical summary receiver operating characteristics (HSROC) were generated and heterogeneity was verified through covariates potentially influencing the diagnostic odds ratio. Findings: Nine studies fulfilled inclusion criteria and were included in the analysis. Pooled estimates of the sensitivities of the combined tests in either-positive and both-positive cases were 0.87 (95% CI: 0.83-0.90) and 0.38 (95% CI: 0.29-0.48), respectively. Corresponding specificities were 0.79 (95% CI: 0.63-0.89) and 0.98 (95% CI: 0.96-0.99) respectively. The DORs of the combined tests in either-positive or both-positive result cases were 27.7 (95% CI: 12.5-61.5) and 52 (95% CI: 22.1-122.2), respectively. When including only articles without partial verification bias and also a high-grade cervical intraepithelial neoplasia as a threshold of the disease, DOR of combined test in both-positive result cases remained the highest. However, DORs decreased to 12.1 (95% CI: 6.05-24.1) and 13.8 (95% CI: 7.92-23.9) in studies without partial verification bias for the combined tests in the either-positive and both-positive result cases, respectively. The screener, the place of study and the size of the population significantly influenced the DOR of combined tests in the both-positive result case in restriction analyses that considered only articles with CIN2+ as disease threshold. Conclusions: The combined test in the either-positive result case has a high sensitivity, but a low specificity. These results suggest that the combined test should be considered in developing countries as a primary screening test if facilities exist to confirm, through colposcopy and biopsy, a positive result.

• Asian Pacific Journal of Cancer Prevention
• /
• 제16권10호
• /
• pp.4223-4228
• /
• 2015

Objective: To investigate the diagnostic value of the R-way colposcopic evaluation system (R-way system) in cervical cancer screening. Materials and Methods: Between August 2013 and August 2014, a total of 1,059 cases referred to colposcopy in Peking University First Hospital were studied using both the R-way system and conventional colposcopy. Our study evaluated and compared the diagnostic ability of the two methods in detecting high-grade lesions and cervical cancer (hereinafter called CIN2+). Evaluation indicators including sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), Youden index and the area under the curve (AUC) of the receiver operating characteristic (ROC) were calculated. Results: The R-way system had a slightly lower specificity (94.5%) than conventional colposcopy (96.0%) for CIN2+ detection (P=0.181). However, the sensitivity (77.8%) was significantly higher than with the conventional colposcopic method (46.6%) (${\chi}^2=64.351$, P<0.001). In addition, the AUC of the ROC for CIN2+ detection using the R-way system (0.839) was larger than that with conventional colposcopy (0.731) (Z=4.348, P<0.001). If preliminary result had been drawn from cervical exfoliated cytology before colposcopy referral, combination of the R-way system with cytology could increase the sensitivity to 93.9% for CIN2+ detection (excluding ASCUS\LSIL), confirmed by multipoint biopsy or ECC. Conclusions: The diagnostic value of the R-way evaluation system is higher than that of conventional colposcopic evaluation in cervical cancer screening. Moreover, taking the ease of use and standardized quality control management into account, the R-way system is highly preferable.

• 가정의학회지
• /
• 제39권6호
• /
• pp.333-339
• /
• 2018

Background: There is a distinction in alcohol consumption behavior between adults and college students. This study aims to verify the usability and the optimal cutoff point of Alcohol Use Disorders Identification Test-Korean revised version (AUDIT-KR) for screening alcohol use disorder in college students when the diagnostic and statistical manual of mental disorders (DSM), 5th edition diagnostic criteria is applied. Methods: A total of 922 college students living in Daejeon were enrolled and divided into two groups based on how many items they corresponded to among DSM-5 alcohol use disorder diagnostic criteria: those who corresponded to ${\geq}2$ of the 11 items were classified into the patient group (107 males, 89 females) while the others into the control group (311 males, 415 females). The participants were evaluated using AUDIT-KR to find the optimal cutoff point for screening alcohol use disorder, sensitivity, and specificity. Results: The mean${\pm}$standard deviation scores in the AUDIT-KR were $12.76{\pm}7.27$, $10.72{\pm}4.62$ for males and females, respectively, in the patient group. In contrast, in the control group the scores were $6.26{\pm}5.23$ and $3.95{\pm}3.59$ in males and females, respectively. The area under the receiver operating characteristic curve (95% confidence interval) regarding alcohol use disorder screening by AUDIT-KR was 0.768 (0.715-0.821) and 0.883 (0.848-0.919) for males and females, respectively. The optimal cutoff point of alcohol use disorder for males was >9, sensitivity 64.49%, and specificity 76.85%. The optimal cutoff point for females was >6, sensitivity 82.02%, and specificity 80.48%. Conclusion: This study suggested that AUDIT-KR can be used as a screening tool for alcohol use disorder in groups of college students when DSM-5 diagnosis criteria are applied.

• The Korean Journal of Pain
• /
• 제34권2호
• /
• pp.229-233
• /
• 2021

Background: Iliotibial band friction syndrome (ITBFS) is a common disorder of the lateral knee. Previous research has reported that the iliotibial band (ITB) thickness (ITBT) is correlated with ITBFS, and ITBT has been considered to be a key morphologic parameter of ITBFS. However, the thickness is different from inflammatory hypertrophy. Thus, we made the ITB cross-sectional area (ITBCSA) a new morphological parameter to assess ITBFS. Methods: Forty-three patients with ITBFS group and from 43 normal group who underwent T1W magnetic resonance imaging were enrolled. The ITBCSA was measured as the cross-sectional area of the ITB that was most hypertrophied in the magnetic resonance axial images. The ITBT was measured as the thickest site of ITB. Results: The mean ITBCSA was 25.24 ± 6.59 ㎟ in the normal group and 38.75 ± 9.11 ㎟ in the ITBFS group. The mean ITBT was 1.94 ± 0.41 mm in the normal group and 2.62 ± 0.46 mm in the ITBFS group. Patients in ITBFS group had significantly higher ITBCSA (P < 0.001) and ITBT (P < 0.001) than the normal group. A receiver operator characteristic curve analysis demonstrated that the best cut-off value of the ITBT was 2.29 mm, with 76.7% sensitivity, 79.1% specificity, and area under the curve (AUC) 0.88. The optimal cut-off score of the ITBCSA was 30.66 ㎟, with 79.1% sensitivity, 79.1% specificity, and AUC 0.87. Conclusions: ITBCSA is a new and sensitive morphological parameter for diagnosing ITBFS, and may even be more accurate than ITBT.