• Journal of the korean academy of Pediatric Dentistry
• /
• v.48 no.4
• /
• pp.405-413
• /
• 2021

The purpose of this in vivo study was to assess the clinical screening performance of a quantitative light-induced fluorescence (QLF) device in detecting proximal caries in primary molars. Fluorescence loss, red autofluorescence and a simplified QLF score for proximal caries (QS-proximal) were evaluated for their validity in detecting proximal caries in primary molars compared to bitewing radiography. Three hundred and forty-four primary molar surfaces were included in the study. Carious lesions were scored according to lesion severity assessed by visual-tactile and radiographic examinations. The QLF images were analyzed for two quantitative parameters, fluorescence loss and red autofluorescence, as well as for QS-proximal. For both quantitative parameters and QS-proximal, the sensitivity, specificity and area under receiver operating curve (AUROC) were calculated as a function of the radiographic scoring index at enamel and dentin caries levels. Both quantitative parameters showed fair AUROC values for detecting dentine level caries (△F = 0.794, △R = 0.750). QS-proximal showed higher AUROC values (0.757 - 0.769) than that of visual-tactile scores (0.653) in detecting dentine level caries. The QLF device showed fair screening performance in detecting proximal caries in primary molars compared to bitewing radiography.

• Journal of Trauma and Injury
• /
• v.19 no.1
• /
• pp.35-40
• /
• 2006

Purpose: Injuries are the third leading cause of death in Korea. Isolated chest injury is not uncommon and shows high mortality and morbidity. Several scoring systems are used for triage and stratification for trauma patients, but no standard system is accepted. We aimed to analyze the accuracy of identification of isolated chest injury by using several scoring systems. Methods: We reviewed a total of 75 patients admitted with isolated chest injury between January 2005 and October 2005. Medical records were reviewed by using the Injury Severity Score (ISS), the Revised Trauma Score (RTS), and the Trauma and Injury Severity Score (TRISS). The scoring systems were compared by using statistics methods. Results: The overall predictive accuracy of the TRISS was 12.5%, 12.0% greater than those of the RTS and the ISS. By using the area under the receiver operating characteristic (AUROC) curve, the TRISS showed an excellent discriminative power (AUROC 0.931) compared to the ISS (AUROC 0.926) and the RTS (AUROC 0.872). Conclusion: Compared with the RTS and the ISS, the TRISS is an easily applied tool with excellent prognostic abilities for isolated chest trauma patients. However, the TRISS, the ISS, and the RTS showed high specificity and low sensitivity, so another scoring system is required for triage and stratification of isolated chest injury patients.

• Journal of Digital Convergence
• /
• v.15 no.1
• /
• pp.199-208
• /
• 2017

This study investigates the components and rating system of reliable technology credit information for a technology finance donor who is a consumer of the information and aims to create an effective and optimal technology credit appraisal system to enlarge technology finance supply. Firstly, we calculate the optimal TCAR which becomes the maximum AUROC through the combination of ratio change, verify the substitution possibility between TAR and CR through the existing CR and system gap simulation, and propose a rating system by which financial institutes can utilize the TCAR as a credit rating. As a result, 70% : 30% is the most suitable as the weighted combination ratio of credit rating : technology rating. As a result of this study, we confirmed the possibility that the technical credit rating information could be substituted by the credit rating or the technology appraisal rating. Furthermore, it also suggests that sophisticated risk management is possible through using technology credit rating that are combined with credit and technology appraisal rating.

• Communications for Statistical Applications and Methods
• /
• v.16 no.2
• /
• pp.239-347
• /
• 2009

This paper presents Kolmogorov-Smirnov, mean difference, AUROC and AR, four well known statistics that have been widely used for evaluating the discriminatory power of credit rating models. Criteria for these statistics are determined by the value of mean difference under the assumption of normality and equal standard deviation. Alternative criteria are proposed through the simulations according to various sample sizes, type II error rates, and the ratio of bads, also we suggest the meaning of statistic on the basis of discriminatory power. Finally we make a comparative study of the currently used guidelines and simulated results.

• Korean Management Science Review
• /
• v.30 no.2
• /
• pp.117-132
• /
• 2013

The objective of this paper is to provide an improved technology appraisal model, which considers a variety of macroeconomic variables such as consumer price index and producer price index. The improved model was built using cross correlation analysis and logistic regression analysis. The AUROC analysis showed that goodness-of-fit of the proposed model turned out to be improved than that of the existing model. The model proposed in the paper would be helpful for making a reasonable investments and financing decision, lessening the default rates by systematic risk management, and enhancing the technology commercialization capabilities.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.7
• /
• pp.49-55
• /
• 2016

In digital forensic images, there is a serious problem that is distributed with various image types. For the problem solution, this paper proposes a classification algorithm of the forensic image types. The proposed algorithm extracts the 21-dim. feature vector with the contrast and energy from GLCM (Gray Level Co-occurrence Matrix), and the entropy of each image type. The classification test of the forensic images is performed with an exhaustive combination of the image types. Through the experiments, TP (True Positive) and FN (False Negative) is detected respectively. While it is confirmed that performed class evaluation of the proposed algorithm is rated as 'Excellent(A)' because of the AUROC (Area Under Receiver Operating Characteristic Curve) is 0.9980 by the sensitivity and the 1-specificity. Also, the minimum average decision error is 0.1349. Also, at the minimum average decision error is 0.0179, the whole forensic image types which are involved then, our classification effectiveness is high.

• Journal of Internet Computing and Services
• /
• v.24 no.1
• /
• pp.49-59
• /
• 2023

Recently, attention to the pandemic situation represented by COVID-19 emerged problems caused by unexpected shortage of medical personnel. In this paper, we present a method for diagnosing the presence or absence of lesional sign on PA chest X-ray images as computer vision solution to support diagnosis tasks. Method for visual anomaly detection based on feature modeling can be also applied to X-ray images. With extracting feature vectors from PA chest X-ray images and divide to patch unit, region-specific abnormality can be detected. As preliminary experiment, we created simulation data set containing multiple objects and present results of the comparative experiments in this paper. We present method to improve both efficiency and performance of the process through hard masking of patch features to aligned images. By summing up regional specificity and global anomaly detection results, it shows improved performance by 0.069 AUROC compared to previous studies. By aggregating region-specific and global anomaly detection results, it shows improved performance by 0.069 AUROC compared to our last study.

• Korean Journal of Radiology
• /
• v.23 no.1
• /
• pp.139-149
• /
• 2022

Objective: To compare the effects of bone suppression imaging using deep learning (BSp-DL) based on a generative adversarial network (GAN) and bone subtraction imaging using a dual energy technique (BSt-DE) on radiologists' performance for pulmonary nodule detection on chest radiographs (CXRs). Materials and Methods: A total of 111 adults, including 49 patients with 83 pulmonary nodules, who underwent both CXR using the dual energy technique and chest CT, were enrolled. Using CT as a reference, two independent radiologists evaluated CXR images for the presence or absence of pulmonary nodules in three reading sessions (standard CXR, BSt-DE CXR, and BSp-DL CXR). Person-wise and nodule-wise performances were assessed using receiver-operating characteristic (ROC) and alternative free-response ROC (AFROC) curve analyses, respectively. Subgroup analyses based on nodule size, location, and the presence of overlapping bones were performed. Results: BSt-DE with an area under the AFROC curve (AUAFROC) of 0.996 and 0.976 for readers 1 and 2, respectively, and BSp-DL with AUAFROC of 0.981 and 0.958, respectively, showed better nodule-wise performance than standard CXR (AUAFROC of 0.907 and 0.808, respectively; p ≤ 0.005). In the person-wise analysis, BSp-DL with an area under the ROC curve (AUROC) of 0.984 and 0.931 for readers 1 and 2, respectively, showed better performance than standard CXR (AUROC of 0.915 and 0.798, respectively; p ≤ 0.011) and comparable performance to BSt-DE (AUROC of 0.988 and 0.974; p ≥ 0.064). BSt-DE and BSp-DL were superior to standard CXR for detecting nodules overlapping with bones (p < 0.017) or in the upper/middle lung zone (p < 0.017). BSt-DE was superior (p < 0.017) to BSp-DL in detecting peripheral and sub-centimeter nodules. Conclusion: BSp-DL (GAN-based bone suppression) showed comparable performance to BSt-DE and can improve radiologists' performance in detecting pulmonary nodules on CXRs. Nevertheless, for better delineation of small and peripheral nodules, further technical improvements are required.

• Journal of Trauma and Injury
• /
• v.34 no.1
• /
• pp.3-12
• /
• 2021

Purpose: To evaluate the severity of trauma, many scoring systems and predictive models have been presented. The quick Sequential Organ Failure Assessment (qSOFA) is a simple scoring system based on vital signs, and we expect it to be easier to apply to trauma patients than other trauma assessment tools. Methods: This study was a cross-sectional study of trauma patients who visited the emergency department of Jeju National University Hospital. We excluded patients under the age of 18 years and unknown outcomes. We calculated the qSOFA, the Modified Early Warning Score (mEWS), Revised Trauma Score (RTS), and Injury Severity Score (ISS) based on patients' initial vital signs and assessments performed in the emergency department (ED). The primary outcome was mortality within 14 days of trauma. We analyzed qSOFA scores using multivariate logistic regression analysis and compared the predictive accuracy of these scoring systems using the area under the receiver operating characteristic curve (AUROC). Results: In total, 27,764 patients were analyzed. In the multivariate logistic regression analysis of the qSOFA, the adjusted odds ratios with 95% confidence interval (CI) for mortality relative to a qSOFA score of 0 were 27.82 (13.63-56.79) for a qSOFA score of 1, 373.31 (183.47-759.57) for a qSOFA score of 2, and 494.07 (143.75-1698.15) for a qSOFA score of 3. In the receiver operating characteristic (ROC) curve analysis for the qSOFA, mEWS, ISS, and RTS in predicting the outcomes, for mortality, the AUROC for the qSOFA (AUROC [95% CI]; 0.912 [0.871-0.952]) was significantly greater than those for the ISS (0.700 [0.608-0.793]) and RTS (0.160 [0.108-0.211]). Conclusions: The qSOFA was useful for predicting the prognosis of trauma patients evaluated in the ED.

• Korean Journal of Radiology
• /
• v.23 no.5
• /
• pp.505-516
• /
• 2022

Objective: To evaluate whether artificial intelligence (AI) for detecting breast cancer on mammography can improve the performance and time efficiency of radiologists reading mammograms. Materials and Methods: A commercial deep learning-based software for mammography was validated using external data collected from 200 patients, 100 each with and without breast cancer (40 with benign lesions and 60 without lesions) from one hospital. Ten readers, including five breast specialist radiologists (BSRs) and five general radiologists (GRs), assessed all mammography images using a seven-point scale to rate the likelihood of malignancy in two sessions, with and without the aid of the AI-based software, and the reading time was automatically recorded using a web-based reporting system. Two reading sessions were conducted with a two-month washout period in between. Differences in the area under the receiver operating characteristic curve (AUROC), sensitivity, specificity, and reading time between reading with and without AI were analyzed, accounting for data clustering by readers when indicated. Results: The AUROC of the AI alone, BSR (average across five readers), and GR (average across five readers) groups was 0.915 (95% confidence interval, 0.876-0.954), 0.813 (0.756-0.870), and 0.684 (0.616-0.752), respectively. With AI assistance, the AUROC significantly increased to 0.884 (0.840-0.928) and 0.833 (0.779-0.887) in the BSR and GR groups, respectively (p = 0.007 and p < 0.001, respectively). Sensitivity was improved by AI assistance in both groups (74.6% vs. 88.6% in BSR, p < 0.001; 52.1% vs. 79.4% in GR, p < 0.001), but the specificity did not differ significantly (66.6% vs. 66.4% in BSR, p = 0.238; 70.8% vs. 70.0% in GR, p = 0.689). The average reading time pooled across readers was significantly decreased by AI assistance for BSRs (82.73 vs. 73.04 seconds, p < 0.001) but increased in GRs (35.44 vs. 42.52 seconds, p < 0.001). Conclusion: AI-based software improved the performance of radiologists regardless of their experience and affected the reading time.