Measuring Classifier Performance: On the Incoherence of the Area under the ROC Curve and What to Do about It

2011 ◽  
pp. 119-128
Keyword(s):  
Roc Curve ◽  
Classifier Performance

Head Position Identification

1967 ◽  
Vol 10 (3) ◽  
pp. 438-448
Author(s):  
H. N. Wright
Keyword(s):  
Roc Curve ◽  
Head Position ◽  
Hearing Impaired ◽  
Type Ii ◽  
Experimental Conditions ◽  
Partial Explanation ◽  
Complex Sounds ◽  
Binaural Cues ◽  
Criterion Behavior ◽  
Stimulus Degradation

A binaural recording of traffic sounds that reached an artificial head oriented in five different positions was presented to five subjects, each of whom responded under four different criteria. The results showed that it is possible to examine the ability of listeners to localize sound while listening through earphones and that the criterion adopted by an individual listener is independent of his performance. For the experimental conditions used, the Type II ROC curve generated by manipulating criterion behavior was linear and consistent with a guessing model. Further experiments involving different degrees of stimulus degradation suggested a partial explanation for this finding and illustrated the various types of monaural and binaural cues used by normal and hearing-impaired listeners to localize complex sounds.

Prediction of Post spinal anesthesia hypotension in patients undergoing cesarean section using Perfusion index

2020 ◽  
Vol 5 (3) ◽  
pp. 1241-1245
Author(s):  
Kumud Pyakurel ◽  
Lalit Kumar Rajbanshi ◽  
Ramesh Bhattarai ◽  
Sonia Dahal
Keyword(s):  
Spinal Anesthesia ◽  
Cesarean Delivery ◽  
Supine Position ◽  
Roc Curve ◽  
Anesthetic Management ◽  
College Teaching ◽  
Perfusion Index ◽  
Sensory Level ◽  
Roc Curve Analysis ◽  
Group I

Introduction: Spinal anesthesia induced hypotension frequently complicates Cesarean delivery. This is usually due to sudden sympatholysis causing decreased venous return which can be aggravated by physiological changes of pregnancy leading to change in baseline peripheral vascular tone. Strategies to prevent hypotensive episodes should be the primary aim of anesthetic management. A simple noninvasive measurement of perfusion index derived from pulse oximeter predicting hypotension during the routine intraoperative course could provide a new dynamism to the management and improving the safe execution of anesthesia. Objectives: The primary objective of this study was to compare incidence of hypotension following SAB for LSCS in patients with baseline PI ≤ 3.5 to those with PI > 3.5. The secondary objectives were to compare PI, HR, SBP, MAP at various time intervals and also to study the side effects between the two groups. Methodology: This prospective observational study was conducted at Nobel Medical College Teaching Hospital from to July 2019 to October 2019. 73 Term parturients presenting for elective cesarean delivery were included for the study. Upon arrival in the operation room, standard monitors were attached and baseline HR, SBP, DBP, MAP, PI and SPO2 were recorded in supine position. The patients with baseline PI ≤ 3.5 were enrolled into Group I and those with a PI > 3.5 were enrolled into Group II. Spinal Anesthesia with 10mg of 0.5% heavy Bupivacaine and 20mcg Fentanyl ( total 2.4ml) was given at L3-L4 interspace in sitting position using midline approach. Patient was then returned to supine position with left lateral tilt of 15 degrees to facilitate left uterine displacement. Upper sensory level was checked at 5 minutes using alcohol swab. Once T-6 level was reached, surgery was started. Maternal SBP, DBP, MAP, HR and PI were recorded at 1 minute intervals between spinal injection and delivery and then 3 minutes until end of surgery. Clinically relevant hypotension was defined as the decrease in MAP by 20% or more from baseline value. Results: The incidence of hypotension in Group I was 18.8% (6/30) compared to 81.3% (26/38). This was clinically and statistically highly significant (P = 0.000, odds ratio 0.11). On Spearman’s rank correlation we found highly significant correlation between baseline PI >3.5 and number of episodes of hypotension (rs 0.482, P = 0.000). The sensitivity and specificity of baseline PI with cut-off 3.5 for predicting hypotension were 81.3% and 66.7% respectively. The ROC curve analysis showed 3.53 as appropriate cut‑off for our findings. The area under the ROC curve (AUC) was 0.734 [Figure 6](Lower bound 0.608 and upper bound 0.861, P=0.001).  Conclusion : This study demonstrates that baseline PI of > 3.5 correlates with incidence of hypotension after spinal anesthesia for cesarean delivery in healthy parturients compared to a baseline PI of < 3.5.

VALUE OF COMBINATION OF HAP SCORE AND BISAP SCORE IN PREDICTING SEVERITY IN PATIENTS OF ACUTE PANCREATITIS

2019 ◽  
pp. 96-100
Author(s):  
Thi Ngoc Suong Le ◽  
Pham Chi Tran ◽  
Van Huy Tran
Keyword(s):  
Acute Pancreatitis ◽  
Positive Predictive Value ◽  
Roc Curve ◽  
Predictive Value ◽  
Prognostic Value ◽  
Acute Inflammation ◽  
Clinical Symptoms ◽  
Mortality Rates ◽  
Multiple Organ ◽  
Central Hospital

Acute pancreatitis (AP) is an acute inflammation of the pancreas, usually occurs suddenly with a variety of clinical symptoms, complications of multiple organ failure and high mortality rates. Objectives: To determine the value of combination of HAP score and BISAP score in predicting the severity of acute pancreatitis of the Atlanta 2012 Classification. Patients and Methods: 75 patients of acute pancreatitis hospitalized at Hue Central Hospital between March 2017 and July 2018; HAP and BISHAP score is calculated within the first 24 hours. The severity of AP was classified by the revised Atlanta criteria 2012. Results: When combining the HAP and BISAP scores in predicting the severity of acute pancreatitis, the area under the ROC curve was 0,923 with sensitivity value was 66.7%, specificity value was 97.1%; positive predictive value was 66.7%, negative predictive value was 97.1%. Conclusion: The combination of HAP and BISAP scores increased the sensitivity, predictive value, and prognostic value in predicting the severity of acute pancreatitis of the revised Atlanta 2012 classification in compare to each single scores. Key words: HAPscore, BiSAP score, acute pancreatitis, predicting severity

Binary Spectrum Feature for Improved Classifier Performance

2020 ◽  
Author(s):  
Nalika Ulapane ◽  
Karthick Thiyagarajan ◽  
sarath kodagoda
Keyword(s):  
Machine Learning ◽  
Classification Performance ◽  
Feature Reduction ◽  
Sensor Data ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
Svm Classifier ◽  
Monitoring Task ◽  
Classifier Performance ◽  
Spectrum Feature

<div>Classification has become a vital task in modern machine learning and Artificial Intelligence applications, including smart sensing. Numerous machine learning techniques are available to perform classification. Similarly, numerous practices, such as feature selection (i.e., selection of a subset of descriptor variables that optimally describe the output), are available to improve classifier performance. In this paper, we consider the case of a given supervised learning classification task that has to be performed making use of continuous-valued features. It is assumed that an optimal subset of features has already been selected. Therefore, no further feature reduction, or feature addition, is to be carried out. Then, we attempt to improve the classification performance by passing the given feature set through a transformation that produces a new feature set which we have named the “Binary Spectrum”. Via a case study example done on some Pulsed Eddy Current sensor data captured from an infrastructure monitoring task, we demonstrate how the classification accuracy of a Support Vector Machine (SVM) classifier increases through the use of this Binary Spectrum feature, indicating the feature transformation’s potential for broader usage.</div><div><br></div>

Expression Levels of MicroRNA-125b in Serum Exosomes of Patients with Asthma of Different Severity and its Diagnostic Significance

2019 ◽  
Vol 20 (10) ◽  
pp. 781-784 ◽  
Author(s):  
Meizhen Zhao ◽  
Li Juanjuan ◽  
Fan Weijia ◽  
Xie Jing ◽  
Huang Qiuhua ◽  
...  
Keyword(s):  
Correlation Analysis ◽  
Roc Curve ◽  
Quantitative Polymerase Chain Reaction ◽  
Persistent Asthma ◽  
Asthma Severity ◽  
Control Group ◽  
Spearman Correlation ◽  
Diagnostic Efficacy ◽  
Expression Levels ◽  
Serum Exosomes

Background: This study aimed to investigate the expression levels of microRNA (miRNA)-125b in serum exosomes and its diagnostic efficacy for asthma severity. Methods: The study included 80 patients with untreated asthma and 80 healthy volunteers. The patients were divided into 4 groups according to disease severity: 20 with the intermittent state, 20 with the mildly persistent state, 20 with the moderately persistent state, and 20 with the severely persistent state. The expression levels of miRNA-125b in serum exosomes of each group were detected using a quantitative polymerase chain reaction and compared. The Spearman correlation analysis was used to study the correlation between the expression levels of miRNA-125b in serum exosomes and asthma severity. The diagnostic efficacy of the expression levels of miRNA-125b in exosomes for asthma severity was evaluated using the Receiver Operating Characteristic (ROC) curve. Results: The expression levels of miRNA-125b in serum exosomes of patients with intermittent, mildly persistent, moderately persistent, and severely persistent asthma were all higher than those in the healthy control group, with statistically significant differences. The expression levels of miRNA-125b were also statistically significantly different among patients in each group. The Spearman correlation analysis showed a positive correlation of the relative expression of miRNA-125b in serum exosomes with asthma severity. The area under the ROC curve of the diagnostic efficacy of miRNA-125b in serum exosomes for patients with intermittent, mildly, moderately, and severely persistent asthma was 0.7770, 0.8573, 0.9111, and 0.9995, respectively. Conclusion: The expression levels of miRNA-125b in serum exosomes had a high diagnostic efficacy and might serve as a noninvasive diagnostic marker for asthma severity.

New Hybrid Method for Left Ventricular Ejection Fraction Assessment from Radionuclide Ventriculography Images

2020 ◽  
Vol 16 ◽  
Author(s):  
Halima Dziri ◽  
Mohamed Ali Cherni ◽  
Dorra Ben Sellem
Keyword(s):  
Left Ventricular Ejection Fraction ◽  
Left Ventricle ◽  
Ejection Fraction ◽  
Positive Predictive Value ◽  
Roc Curve ◽  
Radionuclide Ventriculography ◽  
Left Ventricular ◽  
The Other ◽  
Left Ventricular Ejection ◽  
Ventricular Ejection Fraction

Background: In this paper, we propose a new efficient method of radionuclide ventriculography image segmentation to estimate the left ventricular ejection fraction. This parameter is an important prognostic factor for diagnosing abnormal cardiac function. Methods: The proposed method combines the Chan-Vese and the mathematical morphology algorithms. It was applied to diastolic and systolic images obtained from the Nuclear Medicine Department of Salah AZAIEZ Institute.In order to validate our proposed method, we compare the obtained results to those of two methods of the literature. The first one is based on mathematical morphology, while the second one uses the basic Chan-Vese algorithm. To evaluate the quality of segmentation, we compute accuracy, positive predictive value and area under the ROC curve. We also compare the left ventricle ejection fraction estimated by our method to that of the reference given by the software of the gamma-camera and validated by the expert, using Pearson’s correlation coefficient, ANOVA test and linear regression. Results and conclusion: Static results show that the proposed method is very efficient in the detection of the left ventricle. The accuracy was 98.60%, higher than that of the other two methods (95.52% and 98.50%). Likewise, the positive predictive value was the highest (86.40% vs. 83.63% 71.82%). The area under the ROC curve was also the most important (0.998% vs. 0.926% 0.919%). On the other hand, Pearson's correlation coefficient was the highest (99% vs. 98% 37%). The correlation was significantly positive (p<0.001).

Sign in / Sign up

Export Citation Format

Share Document