scholarly journals Utility of deep neural networks in predicting gross-total resection after transsphenoidal surgery for pituitary adenoma: a pilot study

2018 ◽  
Vol 45 (5) ◽  
pp. E12 ◽  
Author(s):  
Victor E. Staartjes ◽  
Carlo Serra ◽  
Giovanni Muscas ◽  
Nicolai Maldaner ◽  
Kevin Akeret ◽  
...  
Keyword(s):  
Logistic Regression ◽  
Deep Learning ◽  
Pituitary Adenoma ◽  
Pilot Study ◽  
Transsphenoidal Surgery ◽  
Class Imbalance ◽  
Learning Model ◽  
Gross Total Resection ◽  
Total Resection ◽  
Deep Learning Model

OBJECTIVEGross-total resection (GTR) is often the primary surgical goal in transsphenoidal surgery for pituitary adenoma. Existing classifications are effective at predicting GTR but are often hampered by limited discriminatory ability in moderate cases and by poor interrater agreement. Deep learning, a subset of machine learning, has recently established itself as highly effective in forecasting medical outcomes. In this pilot study, the authors aimed to evaluate the utility of using deep learning to predict GTR after transsphenoidal surgery for pituitary adenoma.METHODSData from a prospective registry were used. The authors trained a deep neural network to predict GTR from 16 preoperatively available radiological and procedural variables. Class imbalance adjustment, cross-validation, and random dropout were applied to prevent overfitting and ensure robustness of the predictive model. The authors subsequently compared the deep learning model to a conventional logistic regression model and to the Knosp classification as a gold standard.RESULTSOverall, 140 patients who underwent endoscopic transsphenoidal surgery were included. GTR was achieved in 95 patients (68%), with a mean extent of resection of 96.8% ± 10.6%. Intraoperative high-field MRI was used in 116 (83%) procedures. The deep learning model achieved excellent area under the curve (AUC; 0.96), accuracy (91%), sensitivity (94%), and specificity (89%). This represents an improvement in comparison with the Knosp classification (AUC: 0.87, accuracy: 81%, sensitivity: 92%, specificity: 70%) and a statistically significant improvement in comparison with logistic regression (AUC: 0.86, accuracy: 82%, sensitivity: 81%, specificity: 83%) (all p < 0.001).CONCLUSIONSIn this pilot study, the authors demonstrated the utility of applying deep learning to preoperatively predict the likelihood of GTR with excellent performance. Further training and validation in a prospective multicentric cohort will enable the development of an easy-to-use interface for use in clinical practice.

scholarly journals An Efficient DenseNet-Based Deep Learning Model for Malware Detection

Entropy ◽  
2021 ◽  
Vol 23 (3) ◽  
pp. 344
Author(s):  
Jeyaprakash Hemalatha ◽  
S. Abijah Roseline ◽  
Subbiah Geetha ◽  
Seifedine Kadry ◽  
Robertas Damaševičius
Keyword(s):  
Deep Learning ◽  
Detection System ◽  
Malware Detection ◽  
Class Imbalance ◽  
Learning Model ◽  
Computational Time ◽  
Learning Approaches ◽  
Security Threat ◽  
Static And Dynamic Analysis ◽  
Deep Learning Model

Recently, there has been a huge rise in malware growth, which creates a significant security threat to organizations and individuals. Despite the incessant efforts of cybersecurity research to defend against malware threats, malware developers discover new ways to evade these defense techniques. Traditional static and dynamic analysis methods are ineffective in identifying new malware and pose high overhead in terms of memory and time. Typical machine learning approaches that train a classifier based on handcrafted features are also not sufficiently potent against these evasive techniques and require more efforts due to feature-engineering. Recent malware detectors indicate performance degradation due to class imbalance in malware datasets. To resolve these challenges, this work adopts a visualization-based method, where malware binaries are depicted as two-dimensional images and classified by a deep learning model. We propose an efficient malware detection system based on deep learning. The system uses a reweighted class-balanced loss function in the final classification layer of the DenseNet model to achieve significant performance improvements in classifying malware by handling imbalanced data issues. Comprehensive experiments performed on four benchmark malware datasets show that the proposed approach can detect new malware samples with higher accuracy (98.23% for the Malimg dataset, 98.46% for the BIG 2015 dataset, 98.21% for the MaleVis dataset, and 89.48% for the unseen Malicia dataset) and reduced false-positive rates when compared with conventional malware mitigation techniques while maintaining low computational time. The proposed malware detection solution is also reliable and effective against obfuscation attacks.

scholarly journals Can a deep learning model based on intraoperative time-series monitoring data predict post-hysterectomy quality of recovery?

2021 ◽  
Vol 10 (1) ◽  
Author(s):  
Xu Zhao ◽  
Ke Liao ◽  
Wei Wang ◽  
Junmei Xu ◽  
Lingzhong Meng
Keyword(s):  
Time Series ◽  
Logistic Regression ◽  
Deep Learning ◽  
Intraoperative Monitoring ◽  
Learning Model ◽  
Monitoring Data ◽  
Quality Of Recovery ◽  
Model Based ◽  
Deep Learning Model

Abstract Background Intraoperative physiological monitoring generates a large quantity of time-series data that might be associated with postoperative outcomes. Using a deep learning model based on intraoperative time-series monitoring data to predict postoperative quality of recovery has not been previously reported. Methods Perioperative data from female patients having laparoscopic hysterectomy were prospectively collected. Deep learning, logistic regression, support vector machine, and random forest models were trained using different datasets and evaluated by 5-fold cross-validation. The quality of recovery on postoperative day 1 was assessed using the Quality of Recovery-15 scale. The quality of recovery was dichotomized into satisfactory if the score ≥122 and unsatisfactory if <122. Models’ discrimination was estimated using the area under the receiver operating characteristics curve (AUROC). Models’ calibration was visualized using the calibration plot and appraised by the Brier score. The SHapley Additive exPlanation (SHAP) approach was used to characterize different input features’ contributions. Results Data from 699 patients were used for modeling. When using preoperative data only, all four models exhibited poor performance (AUROC ranging from 0.65 to 0.68). The inclusion of the intraoperative intervention and/or monitoring data improved the performance of the deep leaning, logistic regression, and random forest models but not the support vector machine model. The AUROC of the deep learning model based on the intraoperative monitoring data only was 0.77 (95% CI, 0.72–0.81), which was indistinct from that based on the intraoperative intervention data only (AUROC, 0.79; 95% CI, 0.75–0.82) and from that based on the preoperative, intraoperative intervention, and monitoring data combined (AUROC, 0.81; 95% CI, 0.78–0.83). In contrast, when using the intraoperative monitoring data only, the logistic regression model had an AUROC of 0.72 (95% CI, 0.68–0.77), and the random forest model had an AUROC of 0.74 (95% CI, 0.73–0.76). The Brier score of the deep learning model based on the intraoperative monitoring data was 0.177, which was lower than that of other models. Conclusions Deep learning based on intraoperative time-series monitoring data can predict post-hysterectomy quality of recovery. The use of intraoperative monitoring data for outcome prediction warrants further investigation. Trial registration This trial (Identifier: NCT03641625) was registered at ClinicalTrials.gov by the principal investigator, Lingzhong Meng, on August 22, 2018.

scholarly journals Deep Learning Generates Custom-made Logistic Regression Models for Explaining how Breast Cancer Subtypes are Classified

2021 ◽  
Author(s):  
Takuma Shibahara ◽  
Chisa Wada ◽  
Yasuho Yamashita ◽  
Kazuhiro Fujita ◽  
Masamichi Sato ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Logistic Regression ◽  
Deep Learning ◽  
Linear Model ◽  
Regression Models ◽  
Learning Model ◽  
Logistic Regression Models ◽  
Custom Made ◽  
Meta Learning ◽  
Deep Learning Model

Abstract Breast cancer is the most frequently found cancer in women and the one most often subjected to genetic analysis. Nonetheless, it has been causing the largest number of women's cancer-related deaths. PAM50, the intrinsic subtype assay for breast cancer, is beneficial for diagnosis but does not explain each subtype’s mechanism. Deep learning can predict the subtypes from genetic information more accurately than conventional statistical methods. However, the previous studies did not directly use deep learning to examine which genes associate with the subtypes. To reveal the mechanisms embedded in the PAM50 subtypes, we developed an explainable deep learning model called a point-wise linear model, which uses meta-learning to generate a custom-made logistic regression for each sample. We developed an explainable deep learning model called a point-wise linear model, which uses meta-learning to generate a custom-made logistic regression for each sample. Logistic regression is familiar to physicians, and we can use it to analyze which genes are important for prediction. The custom-made logistic regression models generated by the point-wise linear model used the specific genes selected in other subtypes compared to the conventional logistic regression model: the overlap ratio is less than twenty percent. Analyzing the point-wise linear model’s inner state, we found that the point-wise linear model used genes relevant to the cell cycle-related pathways.

P5014The NPAC score for predicting survival after incident acute myocardial infarction

2019 ◽  
Vol 40 (Supplement_1) ◽  
Author(s):  
K H Li ◽  
J Ho ◽  
Z Xu ◽  
I Lakhani ◽  
G Bazoukis ◽  
...  
Keyword(s):  
Myocardial Infarction ◽  
Acute Myocardial Infarction ◽  
Logistic Regression ◽  
Deep Learning ◽  
Learning Model ◽  
Peripheral Vascular ◽  
Short Term ◽  
C Statistic ◽  
Short Term Mortality ◽  
Deep Learning Model

Abstract Background Risk stratification in acute myocardial infarction (AMI) is important for guiding clinical management. Current risk scores are mostly derived from clinical trials with stringent patient selection. We aimed to establish and evaluate a composite scoring system to predict short-term mortality after index episodes of AMI, independent of electrocardiography (ECG) pattern, in a large real-world cohort. Methods Using electronic health records, patients admitted to our regional teaching hospital (derivation cohort, n=2127) and an independent tertiary care center (validation cohort, n=1276) with index acute myocardial infarction between January 2013 and December 2017 as confirmed by principal diagnosis and laboratory findings, were identified retrospectively. Results Univariate logistic regression was used as the primary model to identify potential contributors to mortality. Stepwise forward likelihood ratio logistic regression revealed that neutrophil-to-lymphocyte ratio, peripheral vascular disease, age, and serum creatinine (NPAC) were significant predictors for 90-day mortality (Hosmer-Lemeshow test, P=0.21). Each component of the NPAC score was weighted by beta-coefficients in multivariate analysis. The C-statistic of the NPAC score was 0.75, which was higher than the conventional Charlson's score (C-statistic=0.63). Application of a deep learning model to our dataset improved the accuracy of classification with a C-statistic of 0.81. Multivariate binary logistic regression Variable β Adjusted Odds ratio (95% CI) P-value Points Age ≥65 years 1.304 3.68 (2.63–5.17) <0.001 2 Peripheral vascular disease 1.109 3.03 (1.52–6.04) 0.002 2 NLRt ≥9.51 1.100 2.73 (2.12–3.51) <0.001 1 Creatinine≥109 μmol/L 1.003 3.00 (2.35–3.85) <0.001 2 NPAC deep learning model Conclusions The NPAC score comprised of four items from routine laboratory parameters and basic clinical information and can facilitate early identification of cases at risk of short-term mortality following index myocardial infarction. Deep learning model can serve as a gate-keeper to provide more accurate prediction to facilitate clinical decision making.

scholarly journals A deep-learning model to continuously predict severe acute kidney injury based on urine output changes in critically ill patients

2021 ◽  
Author(s):  
Francesca Alfieri ◽  
Andrea Ancona ◽  
Giovanni Tripepi ◽  
Dario Crosetto ◽  
Vincenzo Randazzo ◽  
...  
Keyword(s):  
At Risk ◽  
Acute Kidney Injury ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Deep Learning ◽  
Logistic Regression Analysis ◽  
Urine Output ◽  
Kidney Injury ◽  
Learning Model ◽  
Deep Learning Model

Abstract Background Acute Kidney Injury (AKI), a frequent complication of pateints in the Intensive Care Unit (ICU), is associated with a high mortality rate. Early prediction of AKI is essential in order to trigger the use of preventive care actions. Methods The aim of this study was to ascertain the accuracy of two mathematical analysis models in obtaining a predictive score for AKI development. A deep learning model based on a urine output trends was compared with a logistic regression analysis for AKI prediction in stages 2 and 3 (defined as the simultaneous increase of serum creatinine and decrease of urine output, according to  the Acute Kidney Injury Network (AKIN) guidelines). Two retrospective datasets including 35,573 ICU patients were analyzed. Urine output data were used to train and test the logistic regression and the deep learning model. Results The deep learning model defined an area under the curve (AUC) of 0.89 (± 0.01), sensitivity = 0.8 and specificity = 0.84, which was higher than the logistic regression analysis. The deep learning model was able to predict 88% of AKI cases more than 12 h before their onset: for every 6 patients identified as being at risk of AKI by the deep learning model, 5 experienced the event. On the contrary, for every 12 patients not considered to be at risk by the model, 2 developed AKI. Conclusion In conclusion, by using urine output trends, deep learning analysis was able to predict AKI episodes more than 12 h in advance, and with a higher accuracy than the classical urine output thresholds. We suggest that this algorithm could be integrated in the ICU setting to better manage, and potentially prevent, AKI episodes. Graphic abstract

scholarly journals Hip Fracture Detection Using Artificial Intelligence: A Pilot Study

2021 ◽  
Vol 55 (3) ◽  
Author(s):  
Samuel Arsenio M. Grozman ◽  
Patricio E. Dumalo III ◽  
Lauro T. Gonzales
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Deep Learning ◽  
Hip Fracture ◽  
Pilot Study ◽  
Hip Fractures ◽  
Proximal Femur ◽  
Learning Model ◽  
Fracture Detection ◽  
Deep Learning Model

Background. Hip fractures are commonly missed on the first radiograph in up to 30% of patients. The delay in diagnosis leads to significant gaps in management and consequent morbidities. Thus, a computer-aided hip fracture recognition through the Artificial Neural Network deep learning model, which allows the program to learn and gain experience with more images processed, has been created. The study aimed to determine the accuracy and sensitivity of the artificial neural network model in detecting fractures of the hip and explored the feasibility of its use as a diagnostic screening tool. Materials and Methods. A sample size of 45 participants/samples per treatment group was computed using a confidence level of 90%, and prevalence of 0.05 for a pilot study. The program was tested by processing digital pictures of radiographs of patients with known hip fractures that included femoral neck, intertrochanteric, subtrochanteric, and proximal femur fractures taken from the database of adult patients, who have undergone surgery for a hip fracture at the Philippine General Hospital from 2016-17. The 90 (45 fractured, 45 normal) manually selected proximal femur images were run on 10 models. The models were based on AlexNet and VGG-16, which are the representative convoluted neural networks designed for image analysis. Results and Conclusion. The program had an accuracy of 70%, specificity of 42.2% and sensitivity of 97.8%. This study is proof of concept that a deep learning model fracture detection software shows potential in hip fracture detection. Further training is necessary to make this promising innovation clinically useful.

scholarly journals Automatic Identification of Peanut-Leaf Diseases Based on Stack Ensemble

2021 ◽  
Vol 11 (4) ◽  
pp. 1950
Author(s):  
Haixia Qi ◽  
Yu Liang ◽  
Quanchen Ding ◽  
Jun Zou
Keyword(s):  
Machine Learning ◽  
Logistic Regression ◽  
Deep Learning ◽  
Data Augmentation ◽  
Learning Model ◽  
Leaf Spot Disease ◽  
Automatic Identification ◽  
Rust Disease ◽  
Single Leaf ◽  
Deep Learning Model

Peanut is an important food crop, and diseases of its leaves can directly reduce its yield and quality. In order to solve the problem of automatic identification of peanut-leaf diseases, this paper uses a traditional machine-learning method to ensemble the output of a deep learning model to identify diseases of peanut leaves. The identification of peanut-leaf diseases included healthy leaves, rust disease on a single leaf, leaf-spot disease on a single leaf, scorch disease on a single leaf, and both rust disease and scorch disease on a single leaf. Three types of data-augmentation methods were used: image flipping, rotation, and scaling. In this experiment, the deep-learning model had a higher accuracy than the traditional machine-learning methods. Moreover, the deep-learning model achieved better performance when using data augmentation and a stacking ensemble. After ensemble by logistic regression, the accuracy of residual network with 50 layers (ResNet50) was as high as 97.59%, and the F1 score of dense convolutional network with 121 layers (DenseNet121) was as high as 90.50. The deep-learning model used in this experiment had the greatest improvement in F1 score after the logistic regression ensemble. Deep-learning networks with deeper network layers like ResNet50 and DenseNet121 performed better in this experiment. This study can provide a reference for the identification of peanut-leaf diseases.

Improving Sentiment Analysis using Hybrid Deep Learning Model

2020 ◽  
Vol 13 (4) ◽  
pp. 627-640 ◽  
Author(s):  
Avinash Chandra Pandey ◽  
Dharmveer Singh Rajpoot
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Sentiment Analysis ◽  
Classification Accuracy ◽  
Short Term Memory ◽  
Computational Cost ◽  
Extraction Process ◽  
Learning Model ◽  
Sentiment Classification ◽  
Deep Learning Model

Background: Sentiment analysis is a contextual mining of text which determines viewpoint of users with respect to some sentimental topics commonly present at social networking websites. Twitter is one of the social sites where people express their opinion about any topic in the form of tweets. These tweets can be examined using various sentiment classification methods to find the opinion of users. Traditional sentiment analysis methods use manually extracted features for opinion classification. The manual feature extraction process is a complicated task since it requires predefined sentiment lexicons. On the other hand, deep learning methods automatically extract relevant features from data hence; they provide better performance and richer representation competency than the traditional methods. Objective: The main aim of this paper is to enhance the sentiment classification accuracy and to reduce the computational cost. Method: To achieve the objective, a hybrid deep learning model, based on convolution neural network and bi-directional long-short term memory neural network has been introduced. Results: The proposed sentiment classification method achieves the highest accuracy for the most of the datasets. Further, from the statistical analysis efficacy of the proposed method has been validated. Conclusion: Sentiment classification accuracy can be improved by creating veracious hybrid models. Moreover, performance can also be enhanced by tuning the hyper parameters of deep leaning models.

Sign in / Sign up

Export Citation Format

Share Document