Stratified 10-fold Cross-validation Research Articles

A Machine Learning Model for the Prognosis of Pulseless Electrical Activity during Out-of-Hospital Cardiac Arrest.

Pulseless electrical activity (PEA) is characterized by the disassociation of the mechanical and electrical activity of the heart and appears as the initial rhythm in 20–30% of out-of-hospital cardiac arrest (OHCA) cases. Predicting whether a patient in PEA will convert to return of spontaneous circulation (ROSC) is important because different therapeutic strategies are needed depending on the type of PEA. The aim of this study was to develop a machine learning model to differentiate PEA with unfavorable (unPEA) and favorable (faPEA) evolution to ROSC. An OHCA dataset of 1921 PEA signal segments from defibrillator files was used, 703 faPEA segments from 107 patients with ROSC and 1218 unPEA segments from 153 patients with no ROSC. The solution consisted of a signal-processing stage of the ECG and the thoracic impedance (TI) and the extraction of the TI circulation component (ICC), which is associated with ventricular wall movement. Then, a set of 17 features was obtained from the ECG and ICC signals, and a random forest classifier was used to differentiate faPEA from unPEA. All models were trained and tested using patientwise and stratified 10-fold cross-validation partitions. The best model showed a median (interquartile range) area under the curve (AUC) of and a balance accuracy of , improving the previously available solutions at more than four points in the AUC and three points in balanced accuracy. It was demonstrated that the evolution of PEA can be predicted using the ECG and TI signals, opening the possibility of targeted PEA treatment in OHCA.

A novel computer-aided diagnostic system for accurate detection and grading of liver tumors

Liver cancer is a major cause of morbidity and mortality in the world. The primary goals of this manuscript are the identification of novel imaging markers (morphological, functional, and anatomical/textural), and development of a computer-aided diagnostic (CAD) system to accurately detect and grade liver tumors non-invasively. A total of 95 patients with liver tumors (M = 65, F = 30, age range = 34–82 years) were enrolled in the study after consents were obtained. 38 patients had benign tumors (LR1 = 19 and LR2 = 19), 19 patients had intermediate tumors (LR3), and 38 patients had hepatocellular carcinoma (HCC) malignant tumors (LR4 = 19 and LR5 = 19). A multi-phase contrast-enhanced magnetic resonance imaging (CE-MRI) was collected to extract the imaging markers. A comprehensive CAD system was developed, which includes the following main steps: i) estimation of morphological markers using a new parametric spherical harmonic model, ii) estimation of textural markers using a novel rotation invariant gray-level co-occurrence matrix (GLCM) and gray-level run-length matrix (GLRLM) models, and iii) calculation of the functional markers by estimating the wash-in/wash-out slopes, which enable quantification of the enhancement characteristics across different CE-MR phases. These markers were subsequently processed using a two-stages random forest-based classifier to classify the liver tumor as benign, intermediate, or malignant and determine the corresponding grade (LR1, LR2, LR3, LR4, or LR5). The overall CAD system using all the identified imaging markers achieved a sensitivity of 91.8%±0.9%, specificity of 91.2%±1.9%, and F_{1} score of 0.91±0.01, using the leave-one-subject-out (LOSO) cross-validation approach. Importantly, the CAD system achieved overall accuracies of 88%pm 5%, 85%±2%, 78%±3%, 83%±4%, and 79%±3% in grading liver tumors into LR1, LR2, LR3, LR4, and LR5, respectively. In addition to LOSO, the developed CAD system was tested using randomly stratified 10-fold and 5-fold cross-validation approaches. Alternative classification algorithms, including support vector machine, naive Bayes classifier, k-nearest neighbors, and linear discriminant analysis all produced inferior results compared to the proposed two stage random forest classification model. These experiments demonstrate the feasibility of the proposed CAD system as a novel tool to objectively assess liver tumors based on the new comprehensive imaging markers. The identified imaging markers and CAD system can be used as a non-invasive diagnostic tool for early and accurate detection and grading of liver cancer.

A Scalable, Supervised Classification of Seabed Sediment Waves Using an Object-Based Image Analysis Approach

National mapping programs (e.g., INFOMAR and MAREANO) and global efforts (Seabed 2030) acquire large volumes of multibeam echosounder data to map large areas of the seafloor. Developing an objective, automated and repeatable approach to extract meaningful information from such vast quantities of data is now essential. Many automated or semi-automated approaches have been defined to achieve this goal. However, such efforts have resulted in classification schemes that are isolated or bespoke, and therefore it is necessary to form a standardised classification method. Sediment wave fields are the ideal platform for this as they maintain consistent morphologies across various spatial scales and influence the distribution of biological assemblages. Here, we apply an object-based image analysis (OBIA) workflow to multibeam bathymetry to compare the accuracy of four classifiers (two multilayer perceptrons, support vector machine, and voting ensemble) in identifying seabed sediment waves across three separate study sites. The classifiers are trained on high-spatial-resolution (0.5 m) multibeam bathymetric data from Cork Harbour, Ireland and are then applied to lower-spatial-resolution EMODnet data (25 m) from the Hemptons Turbot Bank SAC and offshore of County Wexford, Ireland. A stratified 10-fold cross-validation was enacted to assess overfitting to the sample data. Samples were taken from the lower-resolution sites and examined separately to determine the efficacy of classification. Results showed that the voting ensemble classifier achieved the most consistent accuracy scores across the high-resolution and low-resolution sites. This is the first object-based image analysis classification of bathymetric data able to cope with significant disparity in spatial resolution. Applications for this approach include benthic current speed assessments, a geomorphological classification framework for benthic biota, and a baseline for monitoring of marine protected areas.

A Combined microRNA and Chemokine Profile in Urine to Identify Rejection After Kidney Transplantation.

First, microRNA (miR) molecules were screened in transplant biopsies and urine sediments of patients with acute rejection and patients without rejection and stable graft function. Second, the expression of 15 selected miRs was quantified in an independent set of 115 urine sediments of patients with rejection and 55 urine sediments of patients without histological signs of rejection on protocol biopsy. Additionally, CXCL-9 and CXCL-10 protein levels were quantified in the urine supernatant. Levels of miR-155-5p (5.7-fold), miR-126-3p (4.2-fold), miR-21-5p (3.7-fold), miR-25-3p (2.5-fold), and miR-615-3p (0.4-fold) were significantly different between rejection and no-rejection urine sediments. CXCL-9 and CXCL-10 levels were significantly elevated in urine from recipients with rejection. In a multivariable model (sensitivity: 89.1%, specificity: 75.6%, area under the curve: 0.94, P < 0.001), miR-155-5p, miR-615-3p, and CXCL-9 levels were independent predictors of rejection. Stratified 10-fold cross validation of the model resulted in an area under the curve of 0.92. A combined urinary microRNA and chemokine profile discriminates kidney transplant rejection from stable graft conditions.

PdCSM-GPCR: predicting potent GPCR ligands with graph-based signatures.

MotivationG protein-coupled receptors (GPCRs) can selectively bind to many types of ligands, ranging from light-sensitive compounds, ions, hormones, pheromones and neurotransmitters, modulating cell physiology. Considering their role in many essential cellular processes, they are one of the most targeted protein families, with over a third of all approved drugs modulating GPCR signalling. Despite this, the large diversity of receptors and their multipass transmembrane architectures make the identification and development of novel specific, and safe GPCR ligands a challenge. While computational approaches have the potential to assist GPCR drug development, they have presented limited performance and generalization capabilities. Here, we explored the use of graph-based signatures to develop pdCSM-GPCR, a method capable of rapidly and accurately screening potential GPCR ligands.ResultsBioactivity data (IC50, EC50, Ki and Kd) for individual GPCRs were curated. After curation, we used the data for developing predictive models for 36 major GPCR targets, across 4 classes (A, B, C and F). Our models compose the most comprehensive computational resource for GPCR bioactivity prediction to date. Across stratified 10-fold cross-validation and blind tests, our approach achieved Pearson’s correlations of up to 0.89, significantly outperforming previous methods. Interpreting our results, we identified common important features of potent GPCRs ligands, which tend to have bicyclic rings, leading to higher levels of aromaticity. We believe pdCSM-GPCR will be an invaluable tool to assist screening efforts, enriching compound libraries and ranking candidates for further experimental validation.Availability and implementationpdCSM-GPCR predictive models and datasets used have been made available via a freely accessible and easy-to-use web server at http://biosig.unimelb.edu.au/pdcsm_gpcr/.Supplementary information Supplementary data are available at Bioinformatics Advances online.

Using Machine Learning to Predict Five-Year Reintervention Risk in Type B Aortic Dissection Patients After Thoracic Endovascular Aortic Repair

Purpose: Type B aortic dissection (TBAD) is a high-risk disease, commonly treated with thoracic endovascular aortic repair (TEVAR). However, for the long-term follow-up, it is associated with a high 5-year reintervention rate for patients after TEVAR. There is no accurate definition of prognostic risk factors for TBAD in medical guidelines, and there is no scientific judgment standard for patients’ quality of life or survival outcome in the next five years in clinical practice. A large amount of medical data features makes prognostic analysis difficult. However, machine learning (ML) permits lots of objective data features to be considered for clinical risk stratification and patient management. We aimed to predict the 5-year prognosis in TBAD after TEVAR by Ml, based on baseline, stent characteristics and computed tomography angiography (CTA) imaging data, and provided a certain degree of scientific basis for prognostic risk score and stratification in medical guidelines. Materials and Methods: Dataset we recorded was obtained from 172 TBAD patients undergoing TEVAR. Totally 40 features were recorded, including 14 baseline, 5 stent characteristics and 21 CTA imaging data. Information gain (IG) was used to select features highly associated with adverse outcome. Then, the Gradient Boost classifier was trained using grid search and stratified 5-fold cross-validation, and Its predictive performance was evaluated by the area under the curve (AUC) in the receiver operating characteristic (ROC). Results: Totally 60 patients underwent reintervention during follow-up. Combing 24 features selected by IG, ML model predicted prognosis well in TBAD after TEVAR, with an AUC of 0.816 and a 95% confidence interval of 0.797 to 0.837. Reintervention rate of prediction was slightly higher than the actual (48.2% vs. 34.8%). Conclusion: Machine learning, which combined with baseline, stent characteristics and CTA imaging data for personalized risk computations, effectively predicted reintervention risk in TBAD patients after TEVAR in 5-year follow-up. The model could be used to efficiently assist the clinical management of TBAD patients and prompt high-risk factors.

A hybrid multi-stage learning technique based on brain storming optimization algorithm for breast cancer recurrence prediction

Breast cancer disease is considered to be the second leading reason for death among women. Unfortunately, even if the treatment of cancer started soon after diagnosis, the cancer cells may remain in the body, and cancer may recur. Various Machine Learning (ML) methods to predict breast cancer recurrence were applied recently, and the ML methods’ performance needs to be examined to determine the proper method for prediction. Usually, the datasets contain many features which may sometimes mislead the prediction process; as some features may lead to confusion or inaccurate prediction. Thus, in this study, two breast cancer recurrence datasets were statistically analyzed and further refined by Brain Storming Optimization algorithm (BSO). The proposed multi-stages technique consists of three main stages; first, the statistical feature selection methods (SFM) which statistically select the discriminative features based on importance ranking and features correlations and statistical hypothesis testing, to be passed to the second stage. The features are ranked based on their correlation with the class variable. The second stage namely the multi classifier (MC), which evaluates each method based on three classifiers and produces a combination of features performed by two SFM and three classifiers. In the third stage, the best combination of the selected features was recognized by the BSO algorithm to search for an optimal solution that produces the highest accuracy. In addition, the BSO algorithm has been modified to deal with the feature selection problem. The performance of the proposed technique was evaluated by stratified 10-fold cross-validation. As a result, the multi-stage learning technique showed to be effective in ranking the features and improved the classification accuracy for breast cancer recurrence.

Using Fitbit data to predict clinical deterioration and unplanned treatment events in cancer outpatients.

e13560 Background: A common cause of preventable harm is the failure to detect and appropriately respond to clinical deterioration. Timely intervention is needed, particularly in cancer patients, to mitigate the effects of adverse events, disease progression, and medical error. This problem requires effective clinical surveillance, early recognition, timely notification of the appropriate clinician, and effective intervention. Methods: Applying a user-centered systems engineering design approach, we designed and implemented a surveillance-and-response system to improve the detection and response to clinical deterioration in cancer outpatients. The surveillance system predicts 7-day risk of UTEs, defined as clinically meaningful changes in the patient’s treatment course or cancer care pathway (e.g., any unplanned/unexpected: clinic or ER visit, hospital admission, or major treatment change and/or delays, and/or death). Data inputs consist of: 1) patient activity and health data collected by a Fitbit monitor; 2) geolocation data to measure activity outside the home (i.e., locations preselected at study onset); 3) clinical data from the hospital’s electronic health record; and 4) patient-reported outcomes measures (i.e., PROMs; the NCCN Distress Thermometer, the Comprehensive OpeN-Ended Survey or CONES, Global Health Score, items from the Consumer Assessment of Healthcare Providers and Systems (CAHPS)). Herein, we measured the effectiveness of Fitbit data alone to UTEs in a pilot sample of patients. Dimension reduction of Fitbit variables was first carried out by using Pearson correlation analysis to eliminate redundant variables. As UTEs are rare events, they were oversampled using the Synthetic Minority Oversampling Technique (SMOTE) to balance the dataset. A random forest classification model was trained to predict 7-day UTE risk. Model accuracy was determined by calculating the mean of Stratified 5-Fold Cross-Validation with 10 repeats. Results: Fitbit data was collected over a 6-8-week period from 14 head and neck cancer patients receiving surgical resection, outpatient chemotherapy, and/or radiotherapy. We identified six UTEs in 5 patients. A random forest classification model was developed from 10 variables derived from 7 Fitbit measures. The following variables were averaged or summed daily: average heart rate (HR), resting HR, below 50% or zone 1 of maximum HR, zone 2 and zone 3 HR combined (i.e., 70-100% of max HR), total daily calories, steps, and sleep in minutes. We achieved a model accuracy of 94% (ROC AUC: 0.984, Precision-Recall AUC: 0.985). Conclusions: Activity and health data collected by a commercial activity monitor demonstrated effectiveness in predicting patient UTEs when an oversampling procedure was used to adjust for class imbalance (i.e., low UTE rate). Future studies are recommended to verify and validate this result in a larger patient sample.

Temporal Feature-Based Classification Into Myocardial Infarction and Other CVDs Merging CNN and Bi-LSTM From ECG Signal

Heart attack else wise termed as myocardial infarction (MI) causes irreparable death of cardiac muscles yielding the focal reason for most casualties among all cardiovascular diseases (CVDs’). A 12-lead electrocardiogram (ECG) generally depicts cardiac abnormalities and so customary deep learning (DL) approaches use the whole signal for binary detection purposes, that is separating healthy control (HC), and MI classes. This paper proposes an alternative approach where 21 temporal features in lieu of the temporal signal are collected from the 12 lead data to reduce redundancy and class imbalance keeping the vital information intact. Then these extracted features are fed into a detection model consisting of a one dimensional (1-D) convolutional neural network (CNN) and a bidirectional long short-term memory (bi-LSTM) layer which classifies into three classes, namely: HC, MI, and non-myocardial infarction (non-MI) subjects for a realistic and reliable assessment. The model’s performance is evaluated using 517 records acquired from the Physikalisch-Technische Bundesanstalt (PTB) database and a state-of-art overall accuracy of 99.246%, kappa of 0.983, and macro averaged F1 score of 98.86% were achieved using stratified 5-fold cross-validation. DL methods suffer to make unbiased decisions in the case of class imbalance due to an insufficient amount of data for a particular class and thus temporal features are employed to inherently reduce this problem. The successful performance of the extracted features depends on the precise detection of fiducial points, and so multiple novel algorithms have been introduced in this paper.

Determination of COVID-19 pneumonia based on generalized convolutional neural network model from chest X-ray images

X-ray units have become one of the most advantageous candidates for triaging the new Coronavirus disease COVID-19 infected patients thanks to its relatively low radiation dose, ease of access, practical, reduced prices, and quick imaging process. This research intended to develop a reliable convolutional-neural-network (CNN) model for the classification of COVID-19 from chest X-ray views. Moreover, it is aimed to prevent bias issues due to the database. Transfer learning-based CNN model was developed by using a sum of 1,218 chest X-ray images (CXIs) consisting of 368 COVID-19 pneumonia and 850 other pneumonia cases by pre-trained architectures, including DenseNet-201, ResNet-18, and SqueezeNet. The chest X-ray images were acquired from publicly available databases, and each individual image was carefully selected to prevent any bias problem. A stratified 5-fold cross-validation approach was utilized with a ratio of 90% for training and 10% for the testing (unseen folds), in which 20% of training data was used as a validation set to prevent overfitting problems. The binary classification performances of the proposed CNN models were evaluated by the testing data. The activation mapping approach was implemented to improve the causality and visuality of the radiograph. The outcomes demonstrated that the proposed CNN model built on DenseNet-201 architecture outperformed amongst the others with the highest accuracy, precision, recall, and F1-scores of 94.96%, 89.74%, 94.59%, and 92.11%, respectively. The results indicated that the reliable diagnosis of COVID-19 pneumonia from CXIs based on the CNN model opens the door to accelerate triage, save critical time, and prioritize resources besides assisting the radiologists.

A NOVEL HYBRID APPROACH OF DEEP LEARNING NETWORK ALONG WITH TRANSFER LEARNING FOR BRAIN TUMOR CLASSIFICATION

Brain tumors occur mainly due to the abnormal and uncontrollable partitioning that happened in the cell. In recent times, the DL[deep learning] method has helped the medical field by diagnosing the medical image process of several disease types. Among which brain tumor-based medical image classification is the most common one. It is done through Deep NN[neural Networks], where the TL[transfer Learning] on the MR image for brain tumor image classification helps in improving the metrics of classification by developing image-level stratified and patient-level models. In this research, almost 3640 T1-based MR[Magnetic Resonance] images from 233 patients with different types of tumors such as the [glioma, meningioma, pituitary] have been collected for research purpose. The image views are based on three views known as the sagittal, coronal, and axial views, respectively. The average brain image in all three views was found to be 14. Image classification is done with the help of cross-trained data with the use of the already trained Inception V3-model. In the MR imagimg dataset, the 3-image-level and 1-patient-level have been a model to proceed with the process. The evaluation process during image classification is based on loss, accuracy, recall, kappa, precision, and AUC. The validation process is done through 4 methods: 10-fold cross-validation, holdout validation, group 10-fold cross-validation, stratified 10fold cross-validation. With the use of the cropped and uncropped dataset, the capacity of the generialization and its improvement process is detected. A siginificant result was gained with the utilization of 10-fold cross-validation, where it acquired an accuracy of 99.82%. The brain tumor classification from the MR image can be done with the help of DNN along with the transfer learning process. Our patient-level organization model noticed the best outcomes in arrangement to improve exactness.

Constructing an Emotion Estimation Model Based on EEG/HRV Indexes Using Feature Extraction and Feature Selection Algorithms

In human emotion estimation using an electroencephalogram (EEG) and heart rate variability (HRV), there are two main issues as far as we know. The first is that measurement devices for physiological signals are expensive and not easy to wear. The second is that unnecessary physiological indexes have not been removed, which is likely to decrease the accuracy of machine learning models. In this study, we used single-channel EEG sensor and photoplethysmography (PPG) sensor, which are inexpensive and easy to wear. We collected data from 25 participants (18 males and 7 females) and used a deep learning algorithm to construct an emotion classification model based on Arousal–Valence space using several feature combinations obtained from physiological indexes selected based on our criteria including our proposed feature selection methods. We then performed accuracy verification, applying a stratified 10-fold cross-validation method to the constructed models. The results showed that model accuracies are as high as 90% to 99% by applying the features selection methods we proposed, which suggests that a small number of physiological indexes, even from inexpensive sensors, can be used to construct an accurate emotion classification model if an appropriate feature selection method is applied. Our research results contribute to the improvement of an emotion classification model with a higher accuracy, less cost, and that is less time consuming, which has the potential to be further applied to various areas of applications.

MRI index lesion radiomics and machine learning for detection of extraprostatic extension of disease: a multicenter study

ObjectivesTo build a machine learning (ML) model to detect extraprostatic extension (EPE) of prostate cancer (PCa), based on radiomics features extracted from prostate MRI index lesions.MethodsConsecutive MRI exams of patients undergoing radical prostatectomy for PCa were retrospectively collected from three institutions. Axial T2-weighted and apparent diffusion coefficient map images were annotated to obtain index lesion volumes of interest for radiomics feature extraction. Data from one institution was used for training, feature selection (using reproducibility, variance and pairwise correlation analyses, and a correlation-based subset evaluator), and tuning of a support vector machine (SVM) algorithm, with stratified 10-fold cross-validation. The model was tested on the two remaining institutions’ data and compared with a baseline reference and expert radiologist assessment of EPE.ResultsIn total, 193 patients were included. From an initial dataset of 2436 features, 2287 were excluded due to either poor stability, low variance, or high collinearity. Among the remaining, 14 features were used to train the ML model, which reached an overall accuracy of 83% in the training set. In the two external test sets, the SVM achieved an accuracy of 79% and 74% respectively, not statistically different from that of the radiologist (81–83%, p = 0.39–1) and outperforming the baseline reference (p = 0.001–0.02).ConclusionsA ML model solely based on radiomics features demonstrated high accuracy for EPE detection and good generalizability in a multicenter setting. Paired to qualitative EPE assessment, this approach could aid radiologists in this challenging task.Key Points• Predicting the presence of EPE in prostate cancer patients is a challenging task for radiologists.• A support vector machine algorithm achieved high diagnostic accuracy for EPE detection, with good generalizability when tested on multiple external datasets.• The performance of the algorithm was not significantly different from that of an experienced radiologist.

Automated detection of COVID-19 from CT scan using convolutional neural network

Under the prevailing circumstances of the global pandemic of COVID-19, early diagnosis and accurate detection of COVID-19 through tests/screening and, subsequently, isolation of the infected people would be a proactive measure. Artificial intelligence (AI) based solutions, using Convolutional Neural Network (CNN) and exploiting the Deep Learning model’s diagnostic capabilities, have been studied in this paper. Transfer Learning approach, based on VGG16 and ResNet50 architectures, has been used to develop an algorithm to detect COVID-19 from CT scan images consisting of Healthy (Normal), COVID-19, and Pneumonia categories. This paper adopts data augmentation and fine-tuning techniques to improve and optimize the VGG16 and ResNet50 model. Further, stratified 5-fold cross-validation has been conducted to test the robustness and effectiveness of the model. The proposed model performs exceptionally well in case of binary classification (COVID-19 vs. Normal) with an average classification accuracy of more than 99% in both VGG16 and ResNet50 based models. In multiclass classification (COVID-19 vs. Normal vs. Pneumonia), the proposed model achieves an average classification accuracy of 86.74% and 88.52% using VGG16 and ResNet50 architectures as baseline, respectively. Experimental results show that the proposed model achieves superior performance and can be used for automated detection of COVID-19 from CT scans.

A comparative study of machine learning classifiers for injury severity prediction of crashes involving three-wheeled motorized rickshaw

Motorcycles and motorcyclists have a variety of attributes that have been found to be a potential contributor to the high liability of vulnerable road users (VRUs). Vulnerable Road Users (VRUs) that include pedestrians, bicyclists, cycle-rickshaw occupants, and motorcyclists constitute by far the highest share of road traffic accidents in developing countries. Motorized three-wheeled Rickshaws (3W-MR) is a popular public transport mode in almost all Pakistani cities and is used primarily for short trips to carry passengers and small-scale goods movement. Despite being an important mode of public transport in the developing world, little work has been done to understand the factors affecting the injury severity of three-wheeled motorized vehicles. Crash injury severity prediction is a promising research target in traffic safety. Traditional statistical models have underlying assumptions and predefined associations, which can yield misleading results if flouted. Machine learning(ML) is an emerging non-parametric method that can effectively capture the non-linear effects of both continuous and discrete variables without prior assumptions and achieve better prediction accuracy. This research analyzed injury severity of three-wheeled motorized rickshaws (3W-MR) using various machine learning-based identification algorithms, i.e., Decision jungle (DJ), Random Forest (RF), and Decision Tree (DT). Three years of crash data (from 2017 to 2019) was collected from Provincial Emergency Response Service RESCUE 1122 for Rawalpindi city, Pakistan. A total of 2,743 3W-MR crashes were reported during the study period that resulted in 258 fatalities. The predictive performance of proposed ML models was assessed using several evaluation metrics such as overall accuracy, macro-average precision, macro-average recall, and geometric means of individual class accuracies. Results revealed that DJ with an overall accuracy of 83.7 % outperformed the DT and RF-based on a stratified 10-fold cross-validation approach. Finally, Spearman correlation analysis showed that factors such as the lighting condition, crashes involving young drivers (aged 20–30 years), facilities with high-speed limits (over 60 mph), weekday, off-peak, and shiny weather conditions were more likely to worsen injury severity of 3W-MR crashes. The outcomes of this study could provide necessary and essential guidance to road safety agencies, particularly in the study area, for proactive implementation of appropriate countermeasures to curb road safety issues pertaining to three-wheeled motorized vehicles.

Development of a multivariable prediction model for early revision of total knee arthroplasty – The effect of including patient-reported outcome measures

BackgroundRevision TKA is a serious adverse event with substantial consequences for the patient. As revision is becoming increasingly common in patients under 65 years, the need for improved preoperative patient selection is imminently needed. Therefore, this study aimed to identify the most important factors of early revision and to develop a prediction model of early revision including assessment of the effect of incorporating data on patient-reported outcome measures (PROMs). Material and methodsA cohort of 538 patients undergoing primary TKA was included. Multiple logistic regression using forward selection of variables was applied to identify the best predictors of early revision and to develop a prediction model. The model was internally validated with stratified 5-fold cross-validation. This procedure was repeated without including data on PROMs to develop a model for comparison. The models were evaluated on their discriminative capacity using area under the receiver operating characteristic curve (AUC). ResultsThe most important factors of early revision were age (OR 0.63 [0.42, 0.95]; P = 0.03), preoperative EQ-5D (OR 0.07 [0.01, 0.51]; P = 0.01), and number of comorbidities (OR 1.01 [0.97, 1.25]; P = 0.15). The AUCs of the models with and without PROMs were 0.65 and 0.61, respectively. The difference between the AUCs was not statistically significant (P = 0.32). ConclusionsAlthough more work is needed in order to reach a clinically meaningful quality of the predictions, our results show that the inclusion of PROMs seems to improve the quality of the prediction model.

Applying machine learning to balance performance and stability of high energy density materials

SummaryThe long-standing performance-stability contradiction issue of high energy density materials (HEDMs) is of extremely complex and multi-parameter nature. Herein, machine learning was employed to handle 28 feature descriptors and 5 properties of detonation and stability of 153 HEDMs, wherein all 21,648 data used were obtained through high-throughput crystal-level quantum mechanics calculations on supercomputers. Among five models, namely, extreme gradient boosting regression tree (XGBoost), adaptive boosting, random forest, multi-layer perceptron, and kernel ridge regression, were respectively trained and evaluated by stratified sampling and 5-fold cross-validation method. Among them, XGBoost model produced the best scoring metrics in predicting the detonation velocity, detonation pressure, heat of explosion, decomposition temperature, and lattice energy of HEDMs, and XGBoost predictions agreed best with the 1,383 experimental data collected from massive literatures. Feature importance analysis was conducted to obtain data-driven insight into the causality of the performance-stability contradiction and delivered the optimal range of key features for more efficient rational design of advanced HEDMs.

Discrimination of malignant from benign thyroid lesions through neural networks using FTIR signals obtained from tissues

The current gold standard in cancer diagnosis-the microscopic examination of hematoxylin and eosin (H&E)-stained biopsies-is prone to bias since it greatly relies on visual examination. Hence, there is a need to develop a more sensitive and specific method for diagnosing cancer. Here, Fourier transform infrared (FTIR) spectroscopy of thyroid tumors (n =164; 76 malignant, 88 benign) was performed and five (5) neural network (NN) models were designed to discriminate the obtained spectral data. PCA-LDA was used as classical benchmark for comparison. Each NN model was evaluated using a stratified 10-fold cross-validation method to avoid overfitting, and the performance metrics-accuracy, area under the curve (AUC), positive predictive value (PPV), negative predictive value (NPV), specificity rate (SR), and recall rate (RR)-were averaged for comparison. All NN models were able to perform excellently as classifiers, and all were able to surpass the LDA model in terms of accuracy. Among the NN models, the RNN model performed best, having an AUC of 95.29% ± 6.08%, an accuracy of 98.06% ± 2.87%, a PPV of 98.57% ± 4.52%, a NPV of 93.18% ± 7.93%, a SR value of 98.89% ± 3.51%, and a RR value of 91.25% ± 10.29%. The RNN model outperformed the LDA model for all metrics except for the AUC, NPV, and RR. In conclusion, NN-based tools were able to predict thyroid cancer based on infrared spectroscopy of tissues with a high level of diagnostic performance in comparison to the gold standard.

SNF-NN: computational method to predict drug-disease interactions using similarity network fusion and neural networks

BackgroundDrug repositioning is an emerging approach in pharmaceutical research for identifying novel therapeutic potentials for approved drugs and discover therapies for untreated diseases. Due to its time and cost efficiency, drug repositioning plays an instrumental role in optimizing the drug development process compared to the traditional de novo drug discovery process. Advances in the genomics, together with the enormous growth of large-scale publicly available data and the availability of high-performance computing capabilities, have further motivated the development of computational drug repositioning approaches. More recently, the rise of machine learning techniques, together with the availability of powerful computers, has made the area of computational drug repositioning an area of intense activities.ResultsIn this study, a novel framework SNF-NN based on deep learning is presented, where novel drug-disease interactions are predicted using drug-related similarity information, disease-related similarity information, and known drug-disease interactions. Heterogeneous similarity information related to drugs and disease is fed to the proposed framework in order to predict novel drug-disease interactions. SNF-NN uses similarity selection, similarity network fusion, and a highly tuned novel neural network model to predict new drug-disease interactions. The robustness of SNF-NN is evaluated by comparing its performance with nine baseline machine learning methods. The proposed framework outperforms all baseline methods (AUC-ROC = 0.867, and AUC-PR=0.876) using stratified 10-fold cross-validation. To further demonstrate the reliability and robustness of SNF-NN, two datasets are used to fairly validate the proposed framework’s performance against seven recent state-of-the-art methods for drug-disease interaction prediction. SNF-NN achieves remarkable performance in stratified 10-fold cross-validation with AUC-ROC ranging from 0.879 to 0.931 and AUC-PR from 0.856 to 0.903. Moreover, the efficiency of SNF-NN is verified by validating predicted unknown drug-disease interactions against clinical trials and published studies.ConclusionIn conclusion, computational drug repositioning research can significantly benefit from integrating similarity measures in heterogeneous networks and deep learning models for predicting novel drug-disease interactions. The data and implementation of SNF-NN are available at http://pages.cpsc.ucalgary.ca/ tnjarada/snf-nn.php.

Identification of North American softwoods via machine-learning

This manuscript reports the feasibility of a sequential convolutional neural network (CNN) machine-learning model that correctly identifies 11 North American softwood species from 14× magnified macroscopic end-grain images. The convolutional network contained a large kernel size, max pooling layers, and leaky rectified linear units to accelerate training. To reduce overfitting of training data, we employed L2 regularization, custom initialization, and stratified 5-fold cross-validation techniques. The database consisted of 1789 wood end-grain images. The training data set consisted of 1431 images, whereas the validation set had approximately 358 images. In both sets, the input image size was 227 pixels × 227 pixels. Data augmentation was performed on-the-fly by flipping, rotating, and zooming the images. We tested the performance of the CNN against precision, sensitivity, specificity, F1 score, and adjusted accuracy. The adjusted accuracy for the entire model was 94.0%. Confusion matrices indicated the lowest performance was in correctly classifying ponderosa pine (Pinus ponderosa Douglas ex P. Lawson &amp; C. Lawson) and eastern spruce (Picea spp. A. Dietr.) group with an average sensitivity of 89.0% for each. Even though high validation accuracy (&gt;94.0%) was achieved, we concluded that a much larger data set is needed for wood identification to obtain industrially accurate identification of softwoods, mainly due to their visual and macroscopic similarities.