scholarly journals Support vector machine and deep-learning object detection for localisation of hard exudates

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Veronika Kurilová ◽  
Jozef Goga ◽  
Miloš Oravec ◽  
Jarmila Pavlovičová ◽  
Slavomír Kajan
Keyword(s):  
Support Vector Machine ◽  
False Positive ◽  
High Reliability ◽  
False Positive Rate ◽  
False Negative ◽  
Support Vector ◽  
Retinal Images ◽  
Svm Classifier ◽  
Hard Exudates ◽  
Positive Rate

AbstractHard exudates are one of the main clinical findings in the retinal images of patients with diabetic retinopathy. Detecting them early significantly impacts the treatment of underlying diseases; therefore, there is a need for automated systems with high reliability. We propose a novel method for identifying and localising hard exudates in retinal images. To achieve fast image pre-scanning, a support vector machine (SVM) classifier was combined with a faster region-based convolutional neural network (faster R-CNN) object detector for the localisation of exudates. Rapid pre-scanning filtered out exudate-free samples using a feature vector extracted from the pre-trained ResNet-50 network. Subsequently, the remaining samples were processed using a faster R-CNN detector for detailed analysis. When evaluating all the exudates as individual objects, the SVM classifier reduced the false positive rate by 29.7% and marginally increased the false negative rate by 16.2%. When evaluating all the images, we recorded a 50% reduction in the false positive rate, without any decrease in the number of false negatives. The interim results suggested that pre-scanning the samples using the SVM prior to implementing the deep-network object detector could simultaneously improve and speed up the current hard exudates detection method, especially when there is paucity of training data.

Deep-learning and radiomics ensemble classifier for false positive reduction in brain metastases segmentation

2021 ◽  
Author(s):  
Zi Yang ◽  
Mingli Chen ◽  
Mahdieh Kazemimoghadam ◽  
Lin Ma ◽  
Strahinja Stojadinovic ◽  
...  
Keyword(s):  
Deep Learning ◽  
Brain Metastases ◽  
False Positive ◽  
False Positive Rate ◽  
Ensemble Classifier ◽  
Support Vector ◽  
Svm Classifier ◽  
Siamese Network ◽  
Proposed Model ◽  
Positive Rate

Abstract Stereotactic radiosurgery (SRS) is now the standard of care for brain metastases (BMs) patients. The SRS treatment planning process requires precise target delineation, which in clinical workflow for patients with multiple (>4) BMs (mBMs) could become a pronounced time bottleneck. Our group has developed an automated BMs segmentation platform to assist in this process. The accuracy of the auto-segmentation, however, is influenced by the presence of false-positive segmentations, mainly caused by the injected contrast during MRI acquisition. To address this problem and further improve the segmentation performance, a deep-learning and radiomics ensemble classifier was developed to reduce the false-positive rate in segmentations. The proposed model consists of a Siamese network and a radiomic-based support vector machine (SVM) classifier. The 2D-based Siamese network contains a pair of parallel feature extractors with shared weights followed by a single classifier. This architecture is designed to identify the inter-class difference. On the other hand, the SVM model takes the radiomic features extracted from 3D segmentation volumes as the input for twofold classification, either a false-positive segmentation or a true BM. Lastly, the outputs from both models create an ensemble to generate the final label. The performance of the proposed model in the segmented mBMs testing dataset reached the accuracy (ACC), sensitivity (SEN), specificity (SPE) and area under the curve (AUC) of 0.91, 0.96, 0.90 and 0.93, respectively. After integrating the proposed model into the original segmentation platform, the average segmentation false negative rate (FNR) and the false positive over the union (FPoU) were 0.13 and 0.09, respectively, which preserved the initial FNR (0.07) and significantly improved the FPoU (0.55). The proposed method effectively reduced the false-positive rate in the BMs raw segmentations indicating that the integration of the proposed ensemble classifier into the BMs segmentation platform provides a beneficial tool for mBMs SRS management.

scholarly journals CarSite-II: an integrated classification algorithm for identifying carbonylated sites based on K-means similarity-based undersampling and synthetic minority oversampling techniques

2021 ◽  
Vol 22 (1) ◽  
Author(s):  
Yun Zuo ◽  
Jianyuan Lin ◽  
Xiangxiang Zeng ◽  
Quan Zou ◽  
Xiangrong Liu
Keyword(s):  
Reactive Oxygen Species ◽  
False Positive ◽  
False Positive Rate ◽  
False Negative ◽  
False Negative Rate ◽  
Reactive Oxygen ◽  
Support Vector ◽  
Oxygen Species ◽  
Negative Rate ◽  
Positive Rate

Abstract Background Carbonylation is a non-enzymatic irreversible protein post-translational modification, and refers to the side chain of amino acid residues being attacked by reactive oxygen species and finally converted into carbonyl products. Studies have shown that protein carbonylation caused by reactive oxygen species is involved in the etiology and pathophysiological processes of aging, neurodegenerative diseases, inflammation, diabetes, amyotrophic lateral sclerosis, Huntington’s disease, and tumor. Current experimental approaches used to predict carbonylation sites are expensive, time-consuming, and limited in protein processing abilities. Computational prediction of the carbonylation residue location in protein post-translational modifications enhances the functional characterization of proteins. Results In this study, an integrated classifier algorithm, CarSite-II, was developed to identify K, P, R, and T carbonylated sites. The resampling method K-means similarity-based undersampling and the synthetic minority oversampling technique (SMOTE-KSU) were incorporated to balance the proportions of K, P, R, and T carbonylated training samples. Next, the integrated classifier system Rotation Forest uses “support vector machine” subclassifications to divide three types of feature spaces into several subsets. CarSite-II gained Matthew’s correlation coefficient (MCC) values of 0.2287/0.3125/0.2787/0.2814, False Positive rate values of 0.2628/0.1084/0.1383/0.1313, False Negative rate values of 0.2252/0.0205/0.0976/0.0608 for K/P/R/T carbonylation sites by tenfold cross-validation, respectively. On our independent test dataset, CarSite-II yield MCC values of 0.6358/0.2910/0.4629/0.3685, False Positive rate values of 0.0165/0.0203/0.0188/0.0094, False Negative rate values of 0.1026/0.1875/0.2037/0.3333 for K/P/R/T carbonylation sites. The results show that CarSite-II achieves remarkably better performance than all currently available prediction tools. Conclusion The related results revealed that CarSite-II achieved better performance than the currently available five programs, and revealed the usefulness of the SMOTE-KSU resampling approach and integration algorithm. For the convenience of experimental scientists, the web tool of CarSite-II is available in http://47.100.136.41:8081/

scholarly journals Deteksi Pemain Basket Terklasifikasi Berbasis Histogram of Oriented Gradients

Jurnal INFORM ◽  
2018 ◽  
Vol 3 (1) ◽  
pp. 6-11
Author(s):  
Nisa ul Hafidhoh ◽  
Septian Enggar Sukmana
Keyword(s):  
Support Vector Machine ◽  
False Positive ◽  
False Positive Rate ◽  
True Positive Rate ◽  
Support Vector ◽  
Histogram Of Oriented Gradients ◽  
True Positive ◽  
Positive Rate

Pada olahraga basket jaman modern ini, kebutuhan analisis pergerakan pemain pada calon tim lawan olahraga basket perlu didukung oleh teknologi informasi yang mampu mengupayakan sistem yang otomatis. Analisis pergerakan pemain yang otomatis perlu didukung oleh sistem deteksi pemain yang handal dan akurat sehingga pemetaan pergerakan dapat dilakukan secara optimal. Tujuan dari penelitian ini adalah untuk mengembangkan metode Histogram of Oriented Gradients (HOG) menjadi sebuah metode deteksi yang handal untuk kasus deteksi pemain basket pada media. Tantangan pada penelitian ini adalah deteksi pemain tidak hanya pada saat berjalan dan berlari namun juga pada saat melompat. Untuk memperkuat fokus dan konsistensi terhadap objek yang terdeteksi, pemanfaatan metode klasifikasi Support Vector Machine (SVM) digunakan melalui kolaborasi terhadap HOG descriptor serta warna kostum pemain sehingga pembeda tim dari masing-masing pemain juga dapat dikenali. Tingkat akurasi dari evaluasi yang dihasilkan adalah 92% untuk true positive rate dan 40% untuk false positive rate.

A hybrid firefly and support vector machine classifier for phishing email detection

Kybernetes ◽  
2016 ◽  
Vol 45 (6) ◽  
pp. 977-994 ◽  
Author(s):  
Oluyinka Aderemi Adewumi ◽  
Ayobami Andronicus Akinyelu
Keyword(s):  
Support Vector Machine ◽  
False Positive Rate ◽  
False Negative ◽  
False Negative Rate ◽  
Support Vector ◽  
Data Set ◽  
Content Type ◽  
Phishing Attacks ◽  
Positive Rate ◽  
E Mail

Purpose – Phishing is one of the major challenges faced by the world of e-commerce today. Thanks to phishing attacks, billions of dollars has been lost by many companies and individuals. The global impact of phishing attacks will continue to be on the increase and thus a more efficient phishing detection technique is required. The purpose of this paper is to investigate and report the use of a nature inspired based-machine learning (ML) approach in classification of phishing e-mails. Design/methodology/approach – ML-based techniques have been shown to be efficient in detecting phishing attacks. In this paper, firefly algorithm (FFA) was integrated with support vector machine (SVM) with the primary aim of developing an improved phishing e-mail classifier (known as FFA_SVM), capable of accurately detecting new phishing patterns as they occur. From a data set consisting of 4,000 phishing and ham e-mails, a set of features, suitable for phishing e-mail detection, was extracted and used to construct the hybrid classifier. Findings – The FFA_SVM was applied to a data set consisting of up to 4,000 phishing and ham e-mails. Simulation experiments were performed to evaluate and compared the performance of the classifier. The tests yielded a classification accuracy of 99.94 percent, false positive rate of 0.06 percent and false negative rate of 0.04 percent. Originality/value – The hybrid algorithm has not been earlier apply, as in this work, to the classification and detection of phishing e-mail, to the best of the authors’ knowledge.

scholarly journals Evaluation of optimal urine screening and confirmation cut-off values for opiates, at a national reference laboratory

2021 ◽  
Vol 0 (0) ◽  
Author(s):  
Çiğdem Karakükcü ◽  
Mehmet Zahid Çıracı ◽  
Derya Kocer ◽  
Mine Yüce Faydalı ◽  
Muhittin Abdulkadir Serdar
Keyword(s):  
False Positive ◽  
False Positive Rate ◽  
False Negative ◽  
Urine Samples ◽  
Reference Laboratory ◽  
National Reference Laboratory ◽  
Urine Screening ◽  
National Reference ◽  
Positive Rate ◽  
Sensitivity Specificity

Abstract Objectives To obtain optimal immunoassay screening and LC-MS/MS confirmation cut-offs for opiate group tests to reduce false positive (FP) and false negative (FN) rates. Methods A total of 126 urine samples, −50 opiate screening negative, 76 positive according to the threshold of 300 ng/mL by CEDIA method – were confirmed by a full-validated in-house LC-MS/MS method. Sensitivity, specificity, FP, and FN rates were determined at cut-off concentrations of both 300 and 2,000 ng/mL for morphine and codeine, and 10 ng/mL for heroin metabolite 6-mono-acetyl-morphine (6-MAM). Results All CEDIA opiate negative urine samples were negative for morphine, codeine and 6-MAM. Although sensitivity was 100% for each cut-off; specificity was 54.9% at CEDIA cut-off 300 ng/mL vs. LC-MS/MS cut-off 300 ng/mL and, 75% at CEDIA cut-off 2,000 ng/mL vs. LC-MS/MS cut-off 2,000 ng/mL. False positive rate was highest (45.1%) at CEDIA cut-off 300 ng/mL. At CEDIA cut-off 2,000 ng/mL vs. LC-MS/MS cut-off 300 ng/mL, specificity increased to 82.4% and FP rate decreased to 17.6%. All 6-MAM positive samples had CEDIA concentration ≥2,000 ng/mL. Conclusions 2,000 ng/mL for screening and 300 ng/mL for confirmation cut-offs are the most efficient thresholds for the lowest rate of FP opiate results.

scholarly journals An Approach Based on the Improved SVM Algorithm for Identifying Malware in Network Traffic

2021 ◽  
Vol 2021 ◽  
pp. 1-14
Author(s):  
Bo Liu ◽  
Jinfu Chen ◽  
Songling Qin ◽  
Zufa Zhang ◽  
Yisong Liu ◽  
...  
Keyword(s):  
Network Traffic ◽  
Cyber Security ◽  
False Positive ◽  
False Positive Rate ◽  
Support Vector ◽  
Traffic Classification ◽  
Svm Algorithm ◽  
Network Traffic Classification ◽  
Positive Rate ◽  
Function Selection

Due to the growth and popularity of the internet, cyber security remains, and will continue, to be an important issue. There are many network traffic classification methods or malware identification approaches that have been proposed to solve this problem. However, the existing methods are not well suited to help security experts effectively solve this challenge due to their low accuracy and high false positive rate. To this end, we employ a machine learning-based classification approach to identify malware. The approach extracts features from network traffic and reduces the dimensionality of the features, which can effectively improve the accuracy of identification. Furthermore, we propose an improved SVM algorithm for classifying the network traffic dubbed Optimized Facile Support Vector Machine (OFSVM). The OFSVM algorithm solves the problem that the original SVM algorithm is not satisfactory for classification from two aspects, i.e., parameter optimization and kernel function selection. Therefore, in this paper, we present an approach for identifying malware in network traffic, called Network Traffic Malware Identification (NTMI). To evaluate the effectiveness of the NTMI approach proposed in this paper, we collect four real network traffic datasets and use a publicly available dataset CAIDA for our experiments. Evaluation results suggest that the NTMI approach can lead to higher accuracy while achieving a lower false positive rate compared with other identification methods. On average, the NTMI approach achieves an accuracy of 92.5% and a false positive rate of 5.527%.

Computed Tomography in Partial Carpal Arthrodesis

1997 ◽  
Vol 22 (5) ◽  
pp. 653-655
Author(s):  
J. M. SOLER-MINOVES ◽  
J. GONZALEZ-USTES ◽  
R. PÉREZ ◽  
M. GIFREU ◽  
A. M. GALLART
Keyword(s):  
Computed Tomography ◽  
False Positive ◽  
False Positive Rate ◽  
False Negative ◽  
Plain Radiography ◽  
X Rays ◽  
Positive Rate

We carried out X-rays and computed tomography in 59 wrists in patients who had previous surgical intercarpal fusions. 1.2 mm thick axial images were obtained perpendicular to the axis of the joint. CT showed whether or not the carpal fusions were united. Compared with CT, plain radiography yielded a 25% false negative and 6% false positive rate. We conclude that CT is more useful than plain X-rays for evaluating partial carpal arthrodesis.

Pathologic Assessment of Computerized Tomography Accuracy for the Evaluation of the Laryngeal Cartilaginous Frame Work in Laryngopharyngeal Carcinomas

1989 ◽  
Vol 75 (2) ◽  
pp. 156-162 ◽  
Author(s):  
Sandro Sulfaro ◽  
Francesco Querin ◽  
Luigi Barzan ◽  
Mario Lutman ◽  
Roberto Comoretto ◽  
...  
Keyword(s):  
Computerized Tomography ◽  
False Positive ◽  
False Positive Rate ◽  
False Negative ◽  
False Negative Rate ◽  
Cartilage Destruction ◽  
High False Negative Rate ◽  
Positive Rate ◽  
Pathologic Assessment ◽  
Low Sensitivity

Sixty-six whole-organ sectioned laryngopharyngectomy specimens removed for cancer during a seven-year period were uniformly examined to determine the accuracy of preoperative high resolution computerized tomography (CT) for detection of cartilaginous involvement. Our results indicate that CT has a high overall specificity (88.2%) but a low sensitivity (47.1 %); we observed a high false-negative rate (26.5%) and a fairly low false-positive rate (5.9%). Massive cartilage destruction was easily assessed by CT, whereas both small macroscopic and microscopic neoplastic foci of cartilaginous invasion were missed on CT scans. Moreover, false-positive cases were mainly due to proximity of the tumor to the cartilage. Clinical implications of these results are discussed.

scholarly journals Flow microfluorometric system for screening gynecologic cytology specimens using propidium iodide-fluorescein isothiocyanate.

1976 ◽  
Vol 24 (1) ◽  
pp. 322-331 ◽  
Author(s):  
B J Fowlkes ◽  
C J Herman ◽  
M Cassidy
Keyword(s):  
Data Analysis ◽  
Propidium Iodide ◽  
False Positive ◽  
False Positive Rate ◽  
False Negative ◽  
False Negative Rate ◽  
Fluorescein Isothiocyanate ◽  
Clinical Specimens ◽  
Positive Rate ◽  
Abnormal Cells

Seventy cervical cytology specimens have been screened by a xero resolution flow analyzer-sorter using propidium iodide and fluorescein isothiocyanate as fluorochromes for nucleus and cytoplasm, respectively. This system shows a 1% sensitivity for detection of abnormal cells using only crude visual data analysis. Screening of clinical specimens was performed on the instrument with a 5.8% false negative rate and a 11.8% false positive rate by comparison with routine visual cytologic evaluation of the same samples.

Sign in / Sign up

Export Citation Format

Share Document