• The Korean Journal of Applied Statistics
• /
• v.32 no.4
• /
• pp.593-605
• /
• 2019

Significant literature exists on the area under the ROC curve (AUC) and the volume under the ROC surface (VUS) which are statistical measures of the discriminant power of classification models. Whereas the partial AUC is restricted on the false positive rate, the two-way partial AUC is restricted on both the false positive rate and true positive rate, which could be more efficient and accurate than partial AUC. The two-way partial AUC was suggested as more efficient and accurate than the partial AUC. Partial VUS as well as the three-way partial VUS were also developed for the ROC surface. A proposed AUC is expressed in this paper with probability and integration using two truncated distribution functions restricted on both the false positive rate and true positive rate. It is also found that this AUC has a relation with the two-way partial AUC. The three-way partial VUS for the ROC surface is also related to the VUS using truncated distribution functions. These AUC and VUS are represented and estimated in terms of Mann-Whitney statistics. Their parametric and non-parametric estimation methods are explored based on normal distributions and random samples.

• The Korean Journal of Applied Statistics
• /
• v.33 no.5
• /
• pp.541-553
• /
• 2020

The receiver operating characteristic (ROC) curve is expressed as both sensitivity and specificity; in addition, some optimal thresholds using the ROC curve are also represented with both sensitivity and specificity. In addition to the sensitivity and specificity, the expected usefulness function is considered as disease prevalence and usefulness. In particular, partial the area under the ROC curve (AUC) on a certain range should be compared when the AUCs of the crossing ROC curves have similar values. In this study, partial AUCs representing high sensitivity and specificity are proposed by using sensitivity and specificity lines, respectively. Assume various distribution functions with ROC curves that are crossing and AUCs that have the same value. We propose a method to improve the discriminant power of the classification models while comparing the partial AUCs obtained using sensitivity and specificity lines.

• BMB Reports
• /
• v.46 no.1
• /
• pp.41-46
• /
• 2013

Identifying genes indispensable for an organism's life and their characteristics is one of the central questions in current biological research, and hence it would be helpful to develop computational approaches towards the prediction of essential genes. The performance of a predictor is usually measured by the area under the receiver operating characteristic curve (AUC). We propose a novel method by implementing genetic algorithms to maximize the partial AUC that is restricted to a specific interval of lower false positive rate (FPR), the region relevant to follow-up experimental validation. Our predictor uses various features based on sequence information, protein-protein interaction network topology, and gene expression profiles. A feature selection wrapper was developed to alleviate the over-fitting problem and to weigh each feature's relevance to prediction. We evaluated our method using the proteome of budding yeast. Our implementation of genetic algorithms maximizing the partial AUC below 0.05 or 0.10 of FPR outperformed other popular classification methods.

• The Korean Journal of Applied Statistics
• /
• v.32 no.2
• /
• pp.187-198
• /
• 2019

Extensive literature exists on how to estimate optimal thresholds based on various accuracy measures using receiver operating characteristic (ROC) and cumulative accuracy profile (CAP) curves. This paper now proposes an alternative measure to represented the specific partial area under the ROC and CAP curves. The relationship between ROC and CAP functions is examined using differential equations of the new defined partial area under curves. In addition, the relationship with the optimal thresholds under conditions of various accuracy measures for the ROC and CAP functions is also derived. We assume there are two kinds of distribution functions composing the mixed distribution as various normal distributions before finding the optimal thresholds. Corresponding type 1 and 2 errors are also explored and discussed under various conditions for accuracy measures.

• Journal of Pharmaceutical Investigation
• /
• v.37 no.5
• /
• pp.297-303
• /
• 2007

The purposes of this study were to clarify the involvement of P-glycoprotein (P-gp) in the efflux of levosulpiride in knockout mice that lack the mdr1a1b gene and to evaluate the relationship between the genetic polymorphisms in MDR1 gene (exon 21) and levosulpiride disposition in healthy Korean subjects. After oral administration ($10\;{\mu}g/g$) of levosulpiride to mdr1a/1b(-/-) and wild-type mice, plasma and brain samples were obtained at 45 min. We also investigated the genotype for MDR1 (exon 21) gene in humans using a polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) method. A single oral dose of 25 mg levosulpiride was administered to 58 healthy subjects, who were based on the MDR1 genotype for the G2677T SNP. Blood samples were taken up to 36 hr after dosing. The concentrations of levosulpiride in mouse plasma and brain were statistically significant difference between the two animal groups (P<0.05). In addition, the average brain-to-plasma concentration ratio (Kp) of levosulpiride was 3.4-fold (P<0.01) higher in the mdr1a/1b(-/-) mice compared with the wild-type mice. We also found that the values of $AUC_{0-{\infty}$, partial AUC ($AUC_{0-4h}$) and $C_{max}$ were significantly different between homozygous 2677TT subjects and the subjects with at least one wild-type allele (GG and GT subjects, P=0.012 for $AUC_{0-{\infty}$; P=0.008 for $AUC_{0-4h}$; P=0.038 for $C_{max}$). The results confirm that levosulpiride is a P-gp substrate in vivo, and clearly demonstrate the effect of SNP 2677G>T in exon 21 of the MDR1 gene on levosulpiride disposition.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.5 no.3
• /
• pp.263-267
• /
• 2005

Building classifiers for financial real-world classification problems is often plagued by severely overlapping and highly skewed class distribution. New performance measures such as receiver operating characteristic (ROC) curve and area under ROC curve (AUC) have been recently introduced in evaluating and building classifiers for those kind of problems. They are, however, in-effective to evaluation of classifier's discrimination performance in a particular class of the classification problems that interests lie in only a local operating range of the classifier, In this paper, a new method is proposed that enables us to directly improve classifier's discrimination performance at a desired local operating range by defining and optimizing a partial area under ROC curve or domain-specific curve, which is difficult to achieve with conventional classification accuracy based learning methods. The effectiveness of the proposed approach is demonstrated in terms of fraud detection capability in a real-world fraud detection problem compared with the MSE-based approach.

• YAKHAK HOEJI
• /
• v.52 no.4
• /
• pp.299-305
• /
• 2008

Gabapentin, 1-(aminomethyl) cyclohexaneacetic acid, is a amino acid derivative, and is clinically effective in the treatment of neuropathic pain and partial seizures of epilepsy as a complementary therapy. The purpose of the present study was to evaluate the bioequivalence of two gabapentin tablets, $Neurontin^{R}$ tablet 800 mg (Pfizer Pharmaceuticals Co., Ltd.) and Gabapenin tablet 800 mg (Hanmi Pharm. Co., Ltd.), according to the guidelines of the Korea Food and Drug Administration (KFDA). The release of gabapentin from the two gabapentin formulations in vitro was tested using KP VIII Apparatus II method with 0.06 M HCI dissolution media. Twenty six healthy male subjects, $23.85{\pm}2.24$ years in age and $69.40{\pm}11.11$ kg in body weight, were divided into two groups and a randomized $2{\times}2$ crossover study was employed. After a single tablet containing 800 mg as gabapentin was orally administered, blood samples were taken at predetermined time intervals and the concentrations of gabapentin in serum were determined using HPLC with fluorescence detector. The dissolution profiles of two formulations were similar in the tested dissolution media. The pharmacokinetic parameters such as $AUC_{t}$, $C_{max}$ and $T_{max}$ were calculated, and ANOVA test was utilized for the statistical analysis of the parameters using logarithmically transformed $AUC_{t}$, $C_{max}$ and untransformed $T_{max}$. The results showed that the differences between two formulations based on the reference drug, $Neurontin^{R}$, were 1.28%, 0.63% and 0.62% for $AUC_{t}$, $C_{max}$ and $T_{max}$, respectively. There were no sequence effects between two formulations in these parameters. The 90% confidence intervals using logarithmically transformed data were within the acceptance range of log 0.8 to log 1.25 (e.g., $log0.9097{\sim}log1.1598$ and $log0.8919{\sim}log1.1262$ for $AUC_{t}$ and $C_{max}$, respectively). Thus, the criteria of the KFDA bioequivalence guideline were satisfied, indicating Gabapenin tablet 800 mg was bioequivalent to $Neurontin^{R}$ tablet 800 mg.

• YAKHAK HOEJI
• /
• v.52 no.3
• /
• pp.195-200
• /
• 2008

Gabapentin, [1-(aminomethyl) cyclohexaneacetic acid], a structural analog of $\gamma$-aminobutyric acid (GABA), is being developed for the treatment of epilepsy. Unlike GABA, gabapentin crosses the blood-brain barrier after systemic administration. Gabapentin is an effective antiepileptic drug in patients with partial and secondarily generalized seizures who are uncontrolled with use of existing anticonvulsant drug therapy. The purpose of the present study was to evaluate the bioequivalence of two gabapentin 400 mg capsules, $Neurontin^{(R)}$ capsule 400 mg (Pfizer Inc.) and Gabatin capsule 400 mg (Korean Drug Co. Ltd), according to the guidelines of the Korea Food and Drug Administration (KFDA). The release of gabapentin from the two gabapentin formulations in vitro was tested using KP VIII Apparatus II method with various dissolution media (pH 1.2, 4.0, 6.8 buffer solution and water). Twenty six healthy male subjects, 23.58$\pm$1.50 years in age and 66.74$\pm$8.31 kg in body weight, were divided into two groups and a randomized 2$\times$2 cross-over study was employed. After one capsule containing 400 mg as gabapentin were orally administered, blood was taken at predetermined time intervals and the concentrations of gabapentin in serum were determined using HPLC with fluorescence detector. The dissolution profiles of two formulations were similar at all dissolution media. In addition, the pharmacokinetic parameters such as $AUC_t$, $C_{max}$ and $T_{max}$ were calculated and ANOVA test was utilized for the statistical analysis of the parameters using logarithmically transformed $AUC_t$, $C_{max}$ and untransformed $T_{max}$. The results showed that the differences between two formulations based on the reference drug, $Neurontin^{(R)}$ capsule 400 mg, were 2.04, -3.68 and 16.79% for $AUC_t$, $C_{max}$ and $T_{max}$, respectively. There were no sequence effects between two formulations in these parameters. The 90% confidence intervals using logarithmically transformed data were within the acceptance range of log 0.8 to log 1.25 (e.g., log 0.91$\sim$log 1.16 and log 0.87$\sim$log 1.11 for $AUC_t$ and $C_{max}$, respectively). Thus, the criteria of the KFDA bioequivalence guideline were satisfied, indicating Gabatin capsule 400 mg was bioequivalent to $Neurontin^{(R)}$ capsule 400 mg.

• Korean Journal of Community Nutrition
• /
• v.26 no.4
• /
• pp.280-295
• /
• 2021

Objectives: This study aimed to examine the characteristics of patients according to their nutritional status as assessed by five nutritional screening tools: Patient-Generated Subjective Global Assessment (PG-SGA), NUTRISCORE, Nutritional Risk Index (NRI), Prognostic Nutritional Index (PNI), and Controlling Nutritional Status (CONUT) and to compare the agreement, sensitivity, and specificity of these tools. Methods: A total of 952 gastric cancer patients who underwent gastrectomy and chemotherapy from January 2009 to December 2012 at the Samsung Medical Center were included. We categorized patients into malnourished and normal according to the five nutritional screening tools 1 month after surgery and compared their characteristics. We also calculated the Spearman partial correlation, Cohen's Kappa coefficient, the area under the curve (AUC), sensitivity, and specificity of each pair of screening tools. Results: We observed 86.24% malnutrition based on the PG-SGA and 85.82% based on the NUTRISCORE among gastric cancer patients in our study. When we applied NRI or CONUT, however, the malnutrition levels were less than 30%. Patients with malnutrition as assessed by the PG-SGA, NUTRISCORE, or NRI had lower intakes of energy and protein compared to normal patients. When NRI, PNI, or CONUT were used to identify malnutrition, lower levels of albumin, hemoglobin, total lymphocyte count, total cholesterol, and longer postoperative hospital stays were observed among patients with malnutrition compared to those without malnutrition. We found relatively high agreement between PG-SGA and NUTRISCORE; sensitivity was 90.86% and AUC was 0.78. When we compared NRI and PNI, sensitivity was 99.64% and AUC was 0.97. AUC ranged from 0.50 to 0.67 for comparisons between CONUT and each of the other nutritional screening tools. Conclusions: Our study suggests that PG-SGA and NRI have a relatively high agreement with the NUTRISCORE and PNI, respectively. Further cohort studies are needed to examine whether the nutritional status assessed by PG-SGA, NUTRISCORE, NRI, PNI, and CONUT predicts the gastric cancer prognosis.

• Pediatric Gastroenterology, Hepatology & Nutrition
• /
• v.26 no.4
• /
• pp.193-200
• /
• 2023

Purpose: Crohn's disease (CD) is a chronic, idiopathic bowel disorder that can progress to partial or complete bowel obstruction. At present, there are no reliable diagnostic tests that can readily distinguish between acute inflammatory, purely fibrotic and mixed inflammatory and fibrotic. Our aim is to study the utility of contrast enhanced ultrasound (CEUS) in combination with shear wave elastography (SWE) to differentiate fibrotic from inflammatory strictures in children with obstructive CD of the terminal ileum. Methods: Twenty-five (19 male) children between 2016-2021 with CD of the terminal ileum were recruited into the study. Among these patients, 22 had CEUS kinetic measurements of tissue perfusion, including wash-in slope (dB/sec), peak intensity (dB), time to peak intensity (sec), area under the curve (AUC) (dB sec), and SWE. In total, 11 patients required surgery due to bowel obstruction. Histopathologic analysis was performed by a pathologist who was blinded to the CEUS and SWE test results. Results: Patients that underwent surgical resection had significantly higher mean area under the curve on CEUS compared to patients responsive to medical therapy (p=0.03). The AUC also correlated with the degree of hypertrophy and the percent fibrosis of the muscularis propria, as determined by histopathologic grading (p<0.01). There was no difference in the mean elastography measurements between these two patient groups. Conclusion: CEUS is a useful radiological technique that can help identify pediatric patients with medically refractory obstructive fibrotic strictures of the terminal ileum that should be considered for early surgical resection.