Receiver Operating Characteristic Score Research Articles

Visual search patterns during exploration of naturalistic scenes are driven by saliency cues in individuals with cerebral visual impairment

We investigated the relative influence of image salience and image semantics during the visual search of naturalistic scenes, comparing performance in individuals with cerebral visual impairment (CVI) and controls with neurotypical development. Participants searched for a prompted target presented as either an image or text cue. Success rate and reaction time were collected, and gaze behavior was recorded with an eye tracker. A receiver operating characteristic (ROC) analysis compared the distribution of individual gaze landings based on predictions of image salience (using Graph-Based Visual Saliency) and image semantics (using Global Vectors for Word Representations combined with Linguistic Analysis of Semantic Salience) models. CVI participants were less likely and were slower in finding the target. Their visual search behavior was also associated with a larger visual search area and greater number of fixations. ROC scores were also lower in CVI compared to controls for both model predictions. Furthermore, search strategies in the CVI group were not affected by cue type, although search times and accuracy showed a significant correlation with verbal IQ scores for text-cued searches. These results suggest that visual search patterns in CVI are driven mainly by image salience and provide further characterization of higher-order processing deficits observed in this population.

UNMASKING FRAUDSTERS: Ensemble Features Selection to Enhance Random Forest Fraud Detection

Fraud detection is used in various industries, including banking institutes, finance, insurance, government agencies, etc. Recent increases in the number of fraud attempts make fraud detection crucial for safeguarding financial information that is confidential or personal. Many types of fraud problems exist, including card-not-present fraud, fake Marchant, counterfeit checks, stolen credit cards, and others. An ensemble feature selection technique based on Recursive feature elimination (RFE), Information gain (IG), and Chi-Squared (X2) in concurrence with the Random Forest algorithm, was proposed to give research findings and results on fraud detection and prevention. The objective was to choose the essential features for training the model. The Receiver Operating Characteristic (ROC) Score, Accuracy, F1 Score, and Precision are used to evaluate the model's performance. The findings show that the model can differentiate between fraudulent transactions and those that are not, with an ROC Score of 95.83% and an Accuracy of 99.6%. The F1 Score of 99.6%% and precision of 100% further sustain the model's ability to detect fraudulent transactions with the least false positives correctly. The ensemble feature selection technique reduced training time and did not compromise the model's performance, making it a valuable tool for businesses in preventing fraudulent transactions.

A simple diagnostic scoring system for COVID-19 screening

Background: The COVID‐19 pandemic has prompted the world to make various efforts to control its spread by finding ways to diagnose COVID‐19 quickly and accurately. Early identification of COVID‐19 infection is essential, especially in hospitals with limited resources. We aimed to generate two scores based upon clinical and laboratory findings in patients screen for COVID-19 infection.
 Methodology: This study used a retrospective cohort design that involved 705 adults (≥ 18 y old) admitted in Dr. Sardjito Hospital and Dr. S. Hardjolukito Hospital. The patients' data collected included demographic characteristics, anamnesis on signs and symptoms, history of contact with COVID-19 patients, history of staying or visiting an endemic area, comorbidities, and laboratory and radiologic indicators. All variables with a P < 0.25 on the bivariate test were included in a univariable logistic regression. If the P < 0.05, the variable was included in the multivariable logistic regression with a P < 0.05 considered significant. Receiver Operating Characteristic (ROC) producing an area under the curve (AUC) with 95% confidence intervals (CIs) was used to assess discrimination power.
 Results: Two scores were generated; score in Model 1 consisted of clinical signs, basic laboratory indicators, and chest X-ray, and in Model 2 consisted of clinical signs, chest X-ray, basic and advanced laboratory indicators, including C-reactive protein (CRP), lactate dehydrogenase (LDH), albumin, and D-dimer. The ROC score of Model 1 was 0.801 (0.764−0. 838), which is considered good discrimination, and of Model 2 had excellent discrimination with a ROC of 0.858 (0.826−0. 891); the differences in the ROC of the two models was statistically significant (P = 0.03). The score of Model 1 more than 5 had 85% sensitivity and 61% specificity of positive COVID-19. A score of Model 2 more than 4 had 83% sensitivity and 72% specificity for diagnosing positive COVID-19.
 Conclusions: A simple score consisting of clinical symptoms and signs, and simple laboratory indicators can be used to screen for COVID-19 infection.
 Abbreviations: ARDS: Acute respiratory distress syndrome; CRP: C-reactive protein; MLR: monocyte-to-lymphocyte ratio; NLR: Neutrophil-to-lymphocyte ratio; RT-PCR: Reverse Transcription-Polymerase Chain Reaction;
 Key words: COVID-19; Screening System; Clinical Symptoms; Laboratory Indicators
 Citation: Widyastuti Y, Sari D, Kurniawaty J, Widodo U, Fitriani RW C, Jufan AY, Sutaendy K, Jaya P, Ulfa D. A simple diagnostic scoring system for COVID-19 screening. Anaesth. pain intensive care 2022;26(6):784-791.
 DOI: 10.35975/apic.v26i6.2076

A Model for Rainfall Forecasting using Distinct Machine Learning Algorithm

As Agriculture is the pivotal point of survival, rainfall is the important source of its cultivation. Rainfall prophecy has always been a major problem as a prophecy of downfall gives awareness to people and to know in advance about rain to take necessary precautions to cover their crops from rain. A particular dataset is taken from the Kaggle community and this design predicts whether it will rain henceforth or not by using the rainfall in the dataset. Cat Boost model is executed in this design as it’s an open-sourced machine knowledge algorithm, and features great quality without parameter tuning, categorical point support, bettered delicacy, and fast prophecy. Cat Boost model is a Grade boosting toolkit and two critical algorithms classical and innovative are introduced to produce a fight in prophecy shift present in presently being prosecutions of grade boosting algorithms. Cat Boostperformed truly well giving an AUC (Area under wind) score0.8 and a ROC (Receiver operating characteristic wind) score of 89. ROC is called an assessing wind whereas AUC presents a degree or measure of separability as the model is professed enough to distinguish between classes. An Exploratory data analysis is done to examine data distribution, and outliers and provides tools for imaging and understanding the data through graphical representation.

On the Use of Receiver Operating Characteristic Tests for Evaluating Spatial Earthquake Forecasts

AbstractSpatial forecasts of triggered earthquake distributions have been ranked using receiver operating characteristic (ROC) tests. The test is a binary comparison between regions of positive and negative forecast against positive and negative presence of earthquakes. Forecasts predicting only positive changes score higher than Coulomb methods, which predict positive and negative changes. I hypothesize that removing the possibility of failures in negative forecast realms yields better ROC scores. I create a “perfect” Coulomb forecast where all earthquakes only fall into positive stress change areas and compare with an informationless all‐positive forecast. The “perfect” Coulomb forecast barely beats the informationless forecast, and adding as few as four earthquakes occurring in the negative stress regions causes the Coulomb forecast to be no better than an informationless forecast under a ROC test. ROC tests also suffer from data imbalance when applied to earthquake forecasts because there are many more negative cases than positive.

Diagnosis of occupational asthma from serial measurements of forced expiratory volume in 1 s (FEV1) using the Area Between Curves (ABC) score from the Oasys plotter

ObjectivesTo identify the changes in serial 2-hourly forced expiratory volume in 1 s (FEV1) measurements required to identify occupational asthma (OA) using the Oasys Area Between Curves (ABC) score.MethodsThe ABC...

Forecasting GOES 15 >2 MeV Electron Fluxes From Solar Wind Data and Geomagnetic Indices

AbstractThe flux of > 2 MeV electrons at geosynchronous orbit is used by space weather forecasters as a key indicator of enhanced risk of damage to spacecraft in low, medium, or geosynchronous Earth orbits. We present a methodology that uses the amount of time a single input data set (solar wind data or geomagnetic indices) exceeds a given threshold to produce deterministic and probabilistic forecasts of the >2 MeV flux at GEO exceeding 1,000 or 10,000 cm−2 s−1 sr−1 within up to 10 days. By comparing our forecasts with measured fluxes from GOES 15 between 2014 and 2016, we determine the optimum forecast thresholds for deterministic and probabilistic forecasts by maximizing the receiver‐operating characteristic (ROC) and Brier skill scores, respectively. The training data set gives peak ROC scores of 0.71 to 0.87 and peak Brier skill scores of −0.03 to 0.32. Forecasts from AL give the highest skill scores for forecasts of up to 6 days. AL, solar wind pressure, or SYM‐H give the highest skill scores over 7–10 days. Hit rates range over 56–89% with false alarm rates of 11–53%. Applied to 2012, 2013, and 2017, our best forecasts have hit rates of 56–83% and false alarm rates of 10–20%. Further tuning of the forecasts may improve these. Our hit rates are comparable to those from operational fluence forecasts, that incorporate fluence measurements, but our false alarm rates are higher. This proof‐of‐concept shows that the geosynchronous electron flux can be forecast with a degree of success without incorporating a persistence element into the forecasts.

Automatic multilabel electrocardiogram diagnosis of heart rhythm or conduction abnormalities with deep learning: a cohort study

Market-applicable concurrent electrocardiogram (ECG) diagnosis for multiple heart abnormalities that covers a wide range of arrhythmias, with better-than-human accuracy, has not yet been developed. We therefore aimed to engineer a deep learning approach for the automated multilabel diagnosis of heart rhythm or conduction abnormalities by real-time ECG analysis. We used a dataset of ECGs (standard 10 s, 12-channel format) from adult patients (aged ≥18 years), with 21 distinct rhythm classes, including most types of heart rhythm or conduction abnormalities, for the diagnosis of arrhythmias at multilabel level. The ECGs were collected from three campuses of Tongji Hospital (Huazhong University of Science and Technology, Wuhan, China) and annotated by cardiologists. We used these datasets to develop a convolutional neural network approach to generate diagnoses of arrythmias. We collected a test dataset of ECGs from a new group of patients not included in the training dataset. The test dataset was annotated by consensus of a committee of board-certified, actively practicing cardiologists. To evaluate the performance of the model we assessed the F1 score and the area under the curve (AUC) of the receiver operating characteristic (ROC) curve, as well as quantifying sensitivity and specificity. To validate our results, findings for the test dataset were compared with diagnoses made by 53 ECG physicians working in cardiology departments who had a wide range of experience in ECG interpretation (range 0 to >12 years). An external public validation dataset of 962 ECGs from other hospitals was used to study generalisability of the diagnostic model. Our training and validation dataset comprised 180 112 ECGs from 70 692 patients, collected between Jan 1, 2012, and Apr 30, 2019. The test dataset comprised 828 ECGs corresponding to 828 new patients, recorded between Sept 11, 2012, and Aug 30, 2019. At the multilabel level, our deep learning approach to diagnosing heart abnormalities resulted in an exact match in 658 (80%) of 828 ECGs, exceeding the mean performance of physicians (552 [67%] for physicians with 0-6 years of experience; 571 [69%] for physicians with 7-12 years of experience; 621 [75%] for physicians with more than 12 years of experience). Our model had an overall mean F1 score of 0·887 compared with 0·789 for physicians with 0-6 years of experience, 0·815 for physicians with 7-12 years of experience, and 0·831 for physicians with more than 12 years of experience. The model had a mean AUC ROC score of 0·983 (95% CI 0·980-0·986), sensitivity of 0·867 (0·849-0·885) and specificity of 0·995 (0·994-0·996). Promising F1 scores were also obtained from the external public database using our proposed model without any model modifications (mean F1 scores of 0·845 in multilabel and 0·852 in single-label ECGs). Our model is more accurate than physicians working in cardiology departments at distinguishing a range of distinct arrhythmias in single-label and multilabel ECGs, laying a promising foundation for computational decision-support systems in clinical applications. National Natural Science Foundation of China and Hubei Science and Technology Project.

Novel biomarkers of a peripheral blood interferon signature associated with drug-na\xefve early arthritis patients distinguish persistent from self-limiting disease course

We profiled gene expression signatures to distinguish rheumatoid arthritis (RA) from non-inflammatory arthralgia (NIA), self-limiting arthritis (SLA), and undifferentiated arthritis (UA) as compared to healthy controls as novel potential biomarkers for therapeutic responsiveness. Global gene expression profiles of PBMCs from 43 drug-naïve patients presenting with joint symptoms were evaluated and differentially expressed genes identified by comparative analysis with 24 healthy volunteers. Patients were assessed at presentation with follow up at 6 and 12 months. Gene ontology and network pathway analysis were performed using DAVID Bioinformatics Resources v6.7. Gene expression profiles were also determined after disease-modifying anti-rheumatic drug (DMARD) treatment in the inflammatory arthritis groups (i.e. RA and UA) and confirmed by qRT-PCR. Receiver operating characteristic (ROC) curves analysis and Area Under the Curve (AUC) estimation were performed to assess the diagnostic value of candidate gene expression signatures. A type I interferon (IFN) gene signature distinguished DMARD-naïve patients who will subsequently develop persistent inflammatory arthritis (i.e. RA and UA) from those with NIA. In patients with RA, the IFN signature is characterised by up-regulation of SIGLEC1 (p = 0.00597) and MS4A4A (p = 0.00000904). We also identified, EPHB2 (p = 0.000542) and PDZK1IP1 (p = 0.0206) with RA-specific gene expression profiles and elevated expression of the ST6GALNAC1 (p = 0.0023) gene in UA. ROC and AUC risk score analysis suggested that MSA4A (AUC: 0.894, 0.644, 0.720), PDZK1IP1 (AUC: 0.785, 0.806, 0.977), and EPHB2 (AUC: 0.794, 0.723, 0.620) at 0, 6, and 12 months follow-up can accurately discriminate patients with RA from healthy controls and may have practical value for RA diagnosis. In patients with early inflammatory arthritis, ST6GALNAC1 is a potential biomarker for UA as compared with healthy controls whereas EPHB2, MS4A4A, and particularly PDZK1IP1 may discriminate RA patients. SIGLEC1 may also be a useful marker of disease activity in UA.

Microaneurysms detection in color fundus images using machine learning based on directional local contrast

BackgroundAs one of the majorcomplications of diabetes, diabetic retinopathy (DR) is a leadingcause of visual impairment and blindness due to delayed diagnosisand intervention. Microaneurysms appear as the earliest symptom ofDR. Accurate and reliable detection of microaneurysms in colorfundus images has great importance for DR screening.MethodsA microaneurysms' detection methodusing machine learning based on directional local contrast (DLC) isproposed for the early diagnosis of DR. First, blood vessels wereenhanced and segmented using improved enhancement function based onanalyzing eigenvalues of Hessian matrix. Next, with blood vesselsexcluded, microaneurysm candidate regions were obtained using shapecharacteristics and connected components analysis. After imagesegmented to patches, the features of each microaneurysm candidatepatch were extracted, and each candidate patch was classified intomicroaneurysm or non-microaneurysm. The main contributions of ourstudy are (1) making use of directional local contrast inmicroaneurysms' detection for the first time, which does make sensefor better microaneurysms' classification. (2) Applying threedifferent machine learning techniques for classification andcomparing their performance for microaneurysms' detection. Theproposed algorithm was trained and tested on e-ophtha MA database,and further tested on another independent DIARETDB1 database.Results of microaneurysms' detection on the two databases wereevaluated on lesion level and compared with existing algorithms.ResultsThe proposed method has achieved better performance compared with existing algorithms on accuracy and computation time. On e-ophtha MA and DIARETDB1 databases, the area under curve (AUC) of receiver operating characteristic (ROC) curve was 0.87 and 0.86, respectively. The free-response ROC (FROC) score on the two databases was 0.374 and 0.210, respectively. The computation time per image with resolution of 2544×1969, 1400×960 and 1500×1152 is 29 s, 3 s and 2.6 s, respectively.ConclusionsThe proposed methodusing machine learning based on directional local contrast of imagepatches can effectively detect microaneurysms in color fundus imagesand provide an effective scientific basis for early clinical DRdiagnosis.

Developing novel in silico prediction models for assessing chemical reproductive toxicity using the naïve Bayes classifier method.

Assessment of reproductive toxicity is one of the important safety considerations in drug development. Thus, in the present research, the naïve Bayes (NB)-classifier method was applied to develop binary classification models. Six important molecular descriptors for reproductive toxicity were selected by the genetic algorithm. Then, 110 classification models were developed using six molecular descriptors and10 types of fingerprints with 11 different maximum diameters. Among these established models, the model based on six molecular descriptors and the SciTegic extended-connectivity fingerprints with 20 maximum diameters (LCFC_20) displayed the best prediction performance for reproductive toxicity (NB-1), which gave a 0.884 receiver operating characteristic (ROC) score and 91.8% overall prediction accuracy for the Training Set, and produced a 0.888 ROC score and 83.0% overall accuracy for the external Test Set I. In addition, for the external rat multi-generation reproductive toxicity dataset (Test Set II), the NB-1 model generated a 0.806 ROC score and 85.1% concordance. The generated prediction results indicated that the NB-1 model could give robust and reliable predictions for a reproductive toxicity potential of chemicals. Thus, the established model could be applied to filter early-stage molecules for potential reproductive adverse effects. In addition, six important molecular descriptors and new structural alerts for reproductive toxicity were identified, which could help medicinal chemists rationally guide the optimization of lead compounds and select chemicals with the best prospects of being safe and effective.

Text-based emotion prediction system using machine learning approach

Text-based input becomes a common channel for humans in sharing their opinions/emotions to the product or service through online social media, shopping platform etc. Humans are easy to make errors in interpreting emotions, especially the emotion that derived from text based. The main aim of this study is to develop text-based emotion recognition and prediction system. Several market challenges facing in the advancement of emotion analysis with accuracy being the main issue. Therefore, four supervised machine learning classification algorithms such as Multinomial Naïve Bayes, Support Vector Machine, Decision Trees, and k-Nearest Neighbors were investigated. The model was developed based on Ekman’s six basic emotions which are anger, fear, disgust, joy, guilt and sadness. Data pre-processing techniques such as stemming, stop-words, digits and punctuation marks removal, spelling correction, and tokenization were implemented. A benchmark of ISEAR (International Survey on Emotion Antecedents and Reactions) dataset was used to test all models. Multinomial Naïve Bayes classifier resulted the best performance with an average accuracy of 64.08%. Finally, the best model was integrated to graphical user interface using Python Tkinter library to complete the whole system development. Besides, the detailed performance of the best model such as tf-idf and count vectorizer, confusion matrix, precision-recall rate, as well as ROC (Receiver Operating Characteristic) score were also discussed. Text-based emotion prediction system to interpret and understand human emotions was successfully developed.

Real-World Rates and Management of Thrombocytopenia Due to Cancer Treatment in Advanced/Metastatic Female Genitourinary Malignancies

OBJECTIVES Thrombocytopenia is a common, potentially treatment-limiting hematologic side effect in cancer treatment (Kuter DJ, 2018). This is particularly true for female genitourinary (fGU) cancers, such as ovarian and endometrial, where treatment involves agents such as platinum-based agents, taxanes, PARP inhibitors, and gemcitabine (Lord R et al., 2018; Berek JS et al., 2018; Cassidy CA et al., 2001; Ten Berg MJ et al., 2001; Mahner S., 2015). Evidence has shown that dose reduction can be an effective way to address thrombocytopenia and avoid discontinuation of treatment (Lord R et al., 2018; Berek JS et al., 2018). In this study, we explore the use of RWD to understand real-world rates of thrombocytopenia in fGU cancer, including rates of occurrence, management and impacts on chemotherapy treatment, dose reduction, and discontinuation. METHODS Analyses were conducted on nationally representative insurance claims database and Electronic Medical Records (EMR). Inclusion criteria of adult fGU cancer patients with prior exposures to gemcitabine, platinum-based agents, taxanes, or PARP inhibitor with 1-month continuous enrollment (or activity, for EMR) were applied. Thrombocytopenia post-treatment exposure was identified using ICD codes. Baseline descriptive characteristics and subsequent dosing changes were assessed. Results were also stratified for the top 2 fGU malignancies present in the data, ovarian and endometrial cancer (approximately two-fifths and one-third respectively). Factors associated with thrombocytopenia were assessed using logistic regression modeling and ensemble predictive model approaches (e.g., XG Boost). Covariates for regression modeling included patient demographics, pre-existing medical conditions, drug exposures and lab test procedures. Model performance was assessed using Receiver-Operating Characteristic (ROC) scores). RESULTS Thrombocytopenia incidence ranged from 10-15%, with approximately 6% of cases diagnosed within the first 60 days of therapy. Early onset of thrombocytopenia (within 30 days of initiating therapy) occurred in less than 5% of patients and increased to 13% after more than 90 days. A significant proportion of patients with thrombocytopenia showed evidence of treatment discontinuation, with relatively few showing evidence of dose reduction in chemotherapy. Of patients with TCP, up to 15% of patients were prescribed corticosteroids and less than 10% of patients required platelet transfusion. ROC scores for ensemble model approaches were on the order of 85-90%. Factors associated with development of TCP included age, history of arterial disease, and concomitant GI malignancy. CONCLUSION Thrombocytopenia is a common condition in ovarian and endometrial cancer, leading to treatment discontinuation in a high proportion of patients. Given relatively low rates of chemotherapy dose titration in TCP patients, there is an opportunity to improve management of this hematologic complication and enable patients to remain on chemotherapy. Predictive model findings, while preliminary, also suggest opportunity to proactively identify patients with high risk of developing TCP, thereby facilitating earlier identification and pre-emptive treatment modification. Disclosures Ransom: Medidata Solutions: Employment. Galaznik:Medidata Solutions: Employment. Rusli:Medidata Solutions: Employment. Lempernesse:Medidata Solutions: Employment. Berger:Medidata Solutions: Consultancy.

Sequential data assimilation for 1D self-exciting processes with application to urban crime data

A number of models – such as the Hawkes process and log Gaussian Cox process – have been used to understand how crime rates evolve in time and/or space. Within the context of these models and actual crime data, parameters are often estimated using maximum likelihood estimation (MLE) on batch data, but this approach has several limitations such as limited tracking in real-time and uncertainty quantification. For practical purposes, it would be desirable to move beyond batch data estimation to sequential data assimilation. A novel and general Bayesian sequential data assimilation algorithm is developed for joint state-parameter estimation for an inhomogeneous Poisson process by deriving an approximating Poisson-Gamma ‘Kalman’ filter that allows for uncertainty quantification. The ensemble-based implementation of the filter is developed in a similar approach to the ensemble Kalman filter, making the filter applicable to large-scale real world applications unlike nonlinear filters such as the particle filter. The filter has the advantage that it is independent of the underlying model for the process intensity, and can therefore be used for many different crime models, as well as other application domains. The performance of the filter is demonstrated on synthetic data and real Los Angeles gang crime data and compared against a very large sample-size particle filter, showing its effectiveness in practice. In addition the forecast skill of the Hawkes model is investigated for a forecast system using the Receiver Operating Characteristic (ROC) to provide a useful indicator for when predictive policing software for a crime type is likely to be useful. The ROC and Brier scores are used to compare and analyze the forecast skill of sequential data assimilation and MLE. It is found that sequential data assimilation produces improved probabilistic forecasts over the MLE.

Ligand and structure-based approaches for the identification of SIRT1 activators

SIRT1 is a NAD+-dependent deacetylase that involved in various important metabolic pathways. Combined ligand and structure-based approach was utilized for identification of SIRT1 activators. Pharmacophore models were developed using DISCOtech and refined with GASP module of Sybyl X software. Pharmacophore models were composed of two hydrogen bond acceptor (HBA) atoms, two hydrogen bond donor (HBD) sites and one hydrophobic (HY) feature. The pharmacophore models were validated through receiver operating characteristic (ROC) and Güner–Henry (GH) scoring methods. Model-2 was selected as best model among the model 1–3, based on ROC and GH score value, and found reliable in identification of SIRT1 activators. Model-2 (3D search query) was searched against Zinc database. Several compounds with different chemical scaffold were retrieved as hits. Currently, there is no experimental SIRT1 3D structure available, therefore, we modeled SIRT1 protein structure using homology modeling. Compounds with Qfit value of more than 86 were selected for docking study into the SIRT1 homology model to explore the binding mode of retrieved hits in the active allosteric site. Finally, in silico ADMET prediction study was performed with two best docked compounds. Combination of ligand and structure-based modeling methods identified active hits, which may be good lead compounds to develop novel SIRT1 activators.

Histogram analysis of gadoxetic acid-enhanced MRI for quantitative hepatic fibrosis measurement.

PurposeThe diagnosis and monitoring of liver fibrosis is an important clinical issue; however, this is usually achieved by invasive methods such as biopsy. We aimed to determine whether histogram analysis of hepatobiliary phase images of gadoxetic acid-enhanced magnetic resonance imaging (MRI) can provide non-invasive quantitative measurement of liver fibrosis.MethodsThis retrospective study was approved by the institutional ethics committee, and a waiver of informed consent was obtained. Hepatobiliary phase images of preoperative gadoxetic acid-enhanced MRI studies of 105 patients (69 males, 36 females; age 56.1±12.2) with pathologically documented liver fibrosis grades were analyzed. Fibrosis staging was F0/F1/F2/F3/F4 (METAVIR system) for 11/20/13/15/46 patients, respectively. Four regions-of-interest (ROI, each about 2 cm2) were placed on predetermined locations of representative images. The measured signal intensity of pixels in each ROI was used to calculate corrected coefficient of variation (cCV), skewness, and kurtosis. An average value of each parameter was calculated for comparison. Statistical analysis was performed by ANOVA, receiver operating characteristic (ROC) curve analysis, and linear regression.ResultsThe cCV showed statistically significant differences among pathological fibrosis grades (P<0.001) whereas skewness and kurtosis did not. Univariable linear regression analysis suggested cCV to be a meaningful parameter in predicting the fibrosis grade (P<0.001, β = 0.40 and standard error = 0.06). For discriminating F0-3 from F4, the area under ROC score was 0.857, standard deviation 0.036, 95% confidence interval 0.785–0.928.ConclusionHistogram analysis of hepatobiliary phase images of gadoxetic acid-enhanced MRI can provide non-invasive quantitative measurements of hepatic fibrosis.

Copula 함수 기반의 이변량 결합가뭄지수 산정 및 평가

본 연구에서는 가뭄해석을 위한 이변량 결합가뭄지수를 산정하고 국내 활용성을 평가하였다. 이변량 결합가뭄지수 산정을 위해 결합분포함수는 Clayton copula, 매개변수 추정은 교정방법을 이용하였으며, 입력변수로는 1977~2012년 동안의 강수량 및 토양수분량 자료를 선정하였다. 우리나라에 대한 이변량 결합가뭄지수를 산정한 후, 기존 가뭄지수인 SPI 및 SSI와 같이 시계열 분석, 지역별 분석 및 ROC 분석을 수행하여 활용성을 평가하였다. 분석결과 SPI는 가뭄의 시작, SSI는 가뭄의 해갈을 적절히 고려하였으며, 이변량 결합가뭄지수는 SPI 및 SSI의 거동 특성을 적절히 반영하는 것으로 나타났다. 또한 이변량 결합가뭄지수가 SPI 및 SSI에 비해 ROC score가 높게 산정되었으며, 지역별 분석에서도 지역의 가뭄 상황을 적절히 재현하여 활용성이 우수하게 나타났다. 이변량 결합가뭄지수는 기존 가뭄지수의 해석적 한계를 보완하였다는 측면에서 국내 가뭄해석의 활용성이 높다고 판단된다. The objective of this study is to evaluate the utilization of bivariate joint drought index in South Korea. In order to develop the bivariate joint drought index, in this study, Clayton copula was used to estimate the joint distribution function and the calibration method was employed for parameter estimation. Precipitation and soil moisture data were selected as input data of bivariate joint drought index for period of 1977~2012. The time series analysis, ROC (Receiver Operating Characteristic) analysis, spatial analysis were used to evaluate the bivariate joint drought index with SPI (Standardized Precipitation Index) and SSI (Standardized Soil moisture Index). As a result, SPI performed better for drought onset and SSI for drought demise. On the other hand the bivariate joint drought index captured both drought onset and demise very well. The ROC score of bivariate joint drought index was higher than that of SPI and SSI, and it also reflected the local drought situations. The bivariate joint drought index overcomes the limitations of existing drought indices and is useful for drought analysis.

Grey-scale inversion improves detection of lung nodules

Objective The current study aims to establish whether detection of solitary pulmonary nodules can be improved by inverting the grey scale of posteroanterior (PA) chests. Methods 30 PA chest images were presented on 2 occasions to 16 senior radiologists on either primary or secondary class displays in the standard or inverted mode. 15 images within each group contained a single nodule positioned in a range of anatomical sites. A receiver operating characteristic (ROC) methodology was used to explore differences between the presentation modes. Results Improved ROC scores were evident with inverted (Az 0.77) compared with standard (Az 0.73) (p=0.02) images; however, this difference was seen only with the primary displays. The benefits seen are most likely owing to increased nodule luminance with the inverted images, particularly when using primary displays. Conclusion This study demonstrates that the inverted image can offer advantages in lung nodule detection over the standard presentation mode when images are viewed on high-specification viewing systems. The study has demonstrated that there is an improvement in the detectability of lung nodules on an inverted image with a primary display monitor that is not evident with secondary displays. This is likely to be the result of increased nodule luminance on primary displays when images are presented in the inverted mode.

Grey-scale inversion improves detection of lung nodules

The current study aims to establish whether detection of solitary pulmonary nodules can be improved by inverting the grey scale of posteroanterior (PA) chests. 30 PA chest images were presented on 2 occasions to 16 senior radiologists on either primary or secondary class displays in the standard or inverted mode. 15 images within each group contained a single nodule positioned in a range of anatomical sites. A receiver operating characteristic (ROC) methodology was used to explore differences between the presentation modes. Improved ROC scores were evident with inverted (Az 0.77) compared with standard (Az 0.73) (p=0.02) images; however, this difference was seen only with the primary displays. The benefits seen are most likely owing to increased nodule luminance with the inverted images, particularly when using primary displays. This study demonstrates that the inverted image can offer advantages in lung nodule detection over the standard presentation mode when images are viewed on high-specification viewing systems. The study has demonstrated that there is an improvement in the detectability of lung nodules on an inverted image with a primary display monitor that is not evident with secondary displays. This is likely to be the result of increased nodule luminance on primary displays when images are presented in the inverted mode.

Clinical Experience Using the Mann Assessment of Swallowing Ability for Identification of Patients at Risk for Aspiration in a Mixed-Disease Population

To determine the clinical performance characteristics of the Mann Assessment of Swallowing Ability (MASA) for predicting aspiration (determined by videofluoroscopic swallowing study [VFSS]) in a mixed population. We selected 133 cases clinically evaluated using MASA and VFSS from January through June 2007. Ordinal risk rating (ORR) and total numeric score (TNS) were evaluated as predictors of aspiration on VFSS. To account for missing items, the maximum possible score was determined and a weighted percentage score calculated for each patient. We used receiver operating characteristic (ROC) analysis to compare the sensitivity and specificity of ORR and TNS for predicting aspiration. VFSS identified 51 (38.4%) aspirators, while ORR identified 54 (40.6%) as probable or definite aspiration and TNS 19 (14.3%) as moderate to severe aspiration risk. ROC analysis demonstrated an area under the curve of 0.74, 95% CI [0.66, 0.82], for ORR and 0.51, 95% CI [0.41, 0.61], for TNS. These ROC scores suggest that the MASA ORR is better at predicting aspiration on VFSS than the numeric score. In this sample, the subjective ORR had good predictive ability, while the percentage TNS failed to predict aspiration on VFSS. The MASA ORR assessment was a better predictor for a patient's aspiration risk in this population.