Machine Learning Algorithms Logistic Regression Research Articles

Development and validation of prediction models for papillary thyroid cancer structural recurrence using machine learning approaches

BackgroundAlthough papillary thyroid cancer (PTC) patients are known to have an excellent prognosis, up to 30% of patients experience disease recurrence after initial treatment. Accurately predicting disease prognosis remains a challenge given that the predictive value of several predictors remains controversial. Thus, we investigated whether machine learning (ML) approaches based on comprehensive predictors can predict the risk of structural recurrence for PTC patients.MethodsA total of 2244 patients treated with thyroid surgery and radioiodine were included. Twenty-nine perioperative variables consisting of four dimensions (demographic characteristics and comorbidities, tumor-related variables, lymph node (LN)-related variables, and metabolic and inflammatory markers) were analyzed. We applied five ML algorithms—logistic regression (LR), support vector machine (SVM), extreme gradient boosting (XGBoost), random forest (RF), and neural network (NN)—to develop the models. The area under the receiver operating characteristic (AUC-ROC) curve, calibration curve, and variable importance were used to evaluate the models’ performance.ResultsDuring a median follow-up of 45.5 months, 179 patients (8.0%) experienced structural recurrence. The non-stimulated thyroglobulin, LN dissection, number of LNs dissected, lymph node metastasis ratio, N stage, comorbidity of hypertension, comorbidity of diabetes, body mass index, and low-density lipoprotein were used to develop the models. All models showed a greater AUC (AUC = 0.738 to 0.767) than did the ATA risk stratification (AUC = 0.620, DeLong test: P < 0.01). The SVM, XGBoost, and RF model showed greater sensitivity (0.568, 0.595, 0.676), specificity (0.903, 0.857, 0.784), accuracy (0.875, 0.835, 0.775), positive predictive value (PPV) (0.344, 0.272, 0.219), negative predictive value (NPV) (0.959, 0.959, 0.964), and F1 score (0.429, 0.373, 0.331) than did the ATA risk stratification (sensitivity = 0.432, specificity = 0.770, accuracy = 0.742, PPV = 0.144, NPV = 0.938, F1 score = 0.216). The RF model had generally consistent calibration compared with the other models. The Tg and the LNR were the top 2 important variables in all the models, the N stage was the top 5 important variables in all the models.ConclusionsThe RF model achieved the expected prediction performance with generally good discrimination, calibration and interpretability in this study. This study sheds light on the potential of ML approaches for improving the accuracy of risk stratification for PTC patients.Trial registrationRetrospectively registered at www.chictr.org.cn (trial registration number: ChiCTR2300075574, date of registration: 2023-09-08).

Revolutionizing gastric cancer diagnosis through advanced machine learning approaches

&lt;p class="Keywords"&gt;Early detection of gastric cancer through a Computer-Aided Detection (CAD) system has the potential to significantly reduce the mortality rate associated with this disease. This study aims to investigate the effects of class imbalance on the performance of machine learning classifiers in this context. Using a dataset of 145,787 screening records from NHS Liverpool Hospital, we employed stratified sampling to create balanced and unbalanced datasets and evaluated the performance of four machine learning algorithms—Logistic Regression, Support Vector Machine, Naive Bayes, and Multilayer Perceptron—under five different test conditions. The study’s novelty lies in its detailed examination of class imbalance in gastric cancer diagnosis, emphasizing the crucial role of balanced datasets in machine learning-based early detection systems. For the MLP model under 10-fold cross-validation, the Class 0 sensitivity (non-cancer cases) of the unbalanced dataset was 0.968, higher than the balanced dataset’s 0.902. However, the Class 1 sensitivity (cancer cases) and Positive Predictive Value (PPV) of the unbalanced dataset were much lower (0.383 and 0.527) than those of the balanced dataset (0.959 and 0.907), indicating a significant improvement in identifying true positive cases when using a balanced dataset. These findings highlight the negative effect of class imbalance on prediction accuracy for positive cancer cases and underscore the importance of addressing this imbalance for more reliable and accurate predictions in medical diagnosis and screening. This approach has the potential to improve patient outcomes and may contribute to strategies aimed at reducing the mortality rate associated with gastric cancer.&lt;/p&gt;

Machine learning determines stemness associated with simple and basal-like canine mammary carcinomas

Simple and complex carcinomas are the most common type of malignant Canine Mammary Tumors (CMTs), with simple carcinomas exhibiting aggressive behavior and poorer prognostic. Stemness is an ability associated with cancer initiation, malignancy, and therapeutic resistance, but is still few elucidated in canine mammary tumor subtypes. Here, we first validated, using CMT samples, a previously published canine one-class logistic regression machine learning algorithm (OCLR) to predict stemness (mRNAsi) in canine cancer cells. Then, using the canine mRNAsi, we observed that simple carcinomas exhibit higher stemness than complex carcinomas and other histological subtypes. Also, we confirmed that stemness is higher and associated with basal-like CMTs and with NMF2 metagene signature, a tumor-specific DNA-repair metagene signature. Using correlation analysis, we selected the top 50 genes correlated with higher stemness, and the top 50 genes correlated with lower stemness and further performed a gene set enrichment analysis to observe the biological processes enriched for these genes. Finally, we suggested two promise stemness-associated targets in CMTs, POLA2 and APEX1, especially in simple carcinomas. Thus, our work elucidates stemness as a potential mechanism behind the aggressiveness and development of canine mammary tumors, especially in simple carcinomas, describing evidence of a promising strategy to target this disease.

A machine learning one-class logistic regression model to predict stemness for single cell transcriptomics and spatial omics

Cell annotation is a crucial methodological component to interpreting single cell and spatial omics data. These approaches were developed for single cell analysis but are often biased, manually curated and yet unproven in spatial omics. Here we apply a stemness model for assessing oncogenic states to single cell and spatial omic cancer datasets. This one-class logistic regression machine learning algorithm is used to extract transcriptomic features from non-transformed stem cells to identify dedifferentiated cell states in tumors. We found this method identifies single cell states in metastatic tumor cell populations without the requirement of cell annotation. This machine learning model identified stem-like cell populations not identified in single cell or spatial transcriptomic analysis using existing methods. For the first time, we demonstrate the application of a ML tool across five emerging spatial transcriptomic and proteomic technologies to identify oncogenic stem-like cell types in the tumor microenvironment.

ENSEMBLE STACKING DALAM ANALISA SENTIMEN REAKSI VETERAN MILITER AS TERHADAP PENGAMBILALIHAN AFGHANISTAN OLEH TALIBAN

Abstrak— Sentiment analysis can be used to glean information about user opinions and identify social or political trends. There have been many studies on sentiment analysis using machine learning or lexicon-based methods that have been quite impressive. However, machine learning models often have difficulty generalizing to new data due to various reasons, such as overfitting and limited training data. These models are also prone to bias and variance, which negatively affect the accuracy of their predictions. This study discusses the application of the ensemble stacking method in sentiment analysis with the topic of the takeover of Afghanistan by the Taliban. By monitoring social media, the author uses a dataset in the form of comments on YouTube news channels related to the topic raised. Several studies have shown how the ensemble stacking method predicts better than the single model. The research was carried out by creating a sentiment classification model with logistic regression machine learning algorithms, SVM, KNN, and CART then the ensemble stacking classifier formed by the base learner of the four algorithms. As a result, for a single classifier, the highest average accuracy is the logistic regression algorithm of 74.6 percent. The four algorithms are compiled and predicted by logistic regression, and the stacking ensemble classifier that is applied produces better accuracy than the stand-alone classifier, which is 75.3 percent

Heart Disease Prediction Using Machine Learning

Abstract: Machine learning has been widely adopted by many other businesses as a result about its tremendous success in ecommerce, retail, &amp; marketing. Healthcare sector is one about these. Aforementioned sector is regarded for being "information rich" &amp; offers considerable potential for making wise decisions &amp; spotting hidden patterns. We can deal with aforementioned scenario using cutting-edge machine learning algorithms. Heart diseases are among topics certain are most frequently discussed in healthcare sector. A vital organ is heart. It circulates blood &amp; gives it towards all organs so they can work. A key piece about research in aforementioned area is prediction about heart disease incidence. Machine learning algorithm Logistic Regression was employed in aforementioned study. Using aforementioned technique, a prediction system is developed towards determine if a given patient is at risk for heart disease or not. primary goal is towards determine most accurate algorithm certain is suitable for this.

Performance Evaluation of Various Machine Learning Algorithms in Prediction of Chronic kidney disease (CKD)

With a high rate of morbidity and mortality as well as the ability to spread other diseases, chronic kidney disease (CKD) is a major worldwide health concern. Patients sometimes overlook the disease in the early stages of CKD since there are no evident symptoms. Early diagnosis of CKD enables patients to receive effective treatment in time to slow the disease's progression. Due to their quick and precise detection capabilities, machine learning models can help therapists accomplish this goal efficiently. In this research, we suggest a machine learning approach to CKD diagnosis. The website KAGGLE provided the CKD data set, which has a significant number of missing values.. The mean value is used to fill in the blanks; for object data types (strings), we utilized the most frequent object (string) to replace the missing values. Since patients may overlook particular measurements for a variety of reasons, missing values are typically observed in real-world medical scenarios. Four machine learning algorithms—Logistic Regression, SVM, Random Forest Classifier, and Decision Tree Classifier—were applied to create models after successfully completing the incomplete data set. Random Forest has the highest accuracy of these machine learning models.

Characterization of stemness features and construction of a stemness subtype classifier to predict survival and treatment responses in lung squamous cell carcinoma

BackgroundCancer stemness has been proven to affect tumorigenesis, metastasis, and drug resistance in various cancers, including lung squamous cell carcinoma (LUSC). We intended to develop a clinically applicable stemness subtype classifier that could assist physicians in predicting patient prognosis and treatment response.MethodsThis study collected RNA-seq data from TCGA and GEO databases to calculate transcriptional stemness indices (mRNAsi) using the one-class logistic regression machine learning algorithm. Unsupervised consensus clustering was conducted to identify a stemness-based classification. Immune infiltration analysis (ESTIMATE and ssGSEA algorithms) methods were used to investigate the immune infiltration status of different subtypes. Tumor Immune Dysfunction and Exclusion (TIDE) and Immunophenotype Score (IPS) were used to evaluate the immunotherapy response. The pRRophetic algorithm was used to estimate the efficiency of chemotherapeutic and targeted agents. Two machine learning algorithms (LASSO and RF) and multivariate logistic regression analysis were performed to construct a novel stemness-related classifier.ResultsWe observed that patients in the high-mRNAsi group had a better prognosis than those in the low-mRNAsi group. Next, we identified 190 stemness-related differentially expressed genes (DEGs) that could categorize LUSC patients into two stemness subtypes. Patients in the stemness subtype B group with higher mRNAsi scores exhibited better overall survival (OS) than those in the stemness subtype A group. Immunotherapy prediction demonstrated that stemness subtype A has a better response to immune checkpoint inhibitors (ICIs). Furthermore, the drug response prediction indicated that stemness subtype A had a better response to chemotherapy but was more resistant to epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs). Finally, we constructed a nine-gene-based classifier to predict patients’ stemness subtype and validated it in independent GEO validation sets. The expression levels of these genes were also validated in clinical tumor specimens.ConclusionThe stemness-related classifier could serve as a potential prognostic and treatment predictor and assist physicians in selecting effective treatment strategies for patients with LUSC in clinical practice.

Data-Driven Two-Stage Framework for Identification and Characterization of Different Antibiotic-Resistant Escherichia coli Isolates Based on Mass Spectrometry Data.

In clinical microbiology, matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS) is frequently employed for rapid microbial identification. However, rapid identification of antimicrobial resistance (AMR) in Escherichia coli based on a large amount of MALDI-TOF MS data has not yet been reported. This may be because building a prediction model to cover all E. coli isolates would be challenging given the high diversity of the E. coli population. This study aimed to develop a MALDI-TOF MS-based, data-driven, two-stage framework for characterizing different AMRs in E. coli. Specifically, amoxicillin (AMC), ceftazidime (CAZ), ciprofloxacin (CIP), ceftriaxone (CRO), and cefuroxime (CXM) were used. In the first stage, we split the data into two groups based on informative peaks according to the importance of the random forest. In the second stage, prediction models were constructed using four different machine learning algorithms-logistic regression, support vector machine, random forest, and extreme gradient boosting (XGBoost). The findings demonstrate that XGBoost outperformed the other four machine learning models. The values of the area under the receiver operating characteristic curve were 0.62, 0.72, 0.87, 0.72, and 0.72 for AMC, CAZ, CIP, CRO, and CXM, respectively. This implies that a data-driven, two-stage framework could improve accuracy by approximately 2.8%. As a result, we developed AMR prediction models for E. coli using a data-driven two-stage framework, which is promising for assisting physicians in making decisions. Further, the analysis of informative peaks in future studies could potentially reveal new insights. IMPORTANCE Based on a large amount of matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS) clinical data, comprising 37,918 Escherichia coli isolates, a data-driven two-stage framework was established to evaluate the antimicrobial resistance of E. coli. Five antibiotics, including amoxicillin (AMC), ceftazidime (CAZ), ciprofloxacin (CIP), ceftriaxone (CRO), and cefuroxime (CXM), were considered for the two-stage model training, and the values of the area under the receiver operating characteristic curve (AUC) were 0.62 for AMC, 0.72 for CAZ, 0.87 for CIP, 0.72 for CRO, and 0.72 for CXM. Further investigations revealed that the informative peak m/z 9714 appeared with some important peaks at m/z 6809, m/z 7650, m/z 10534, and m/z 11783 for CIP and at m/z 6809, m/z 10475, and m/z 8447 for CAZ, CRO, and CXM. This framework has the potential to improve the accuracy by approximately 2.8%, indicating a promising potential for further research.

Preoperative prediction of macrotrabecular-massive hepatocellular carcinoma through dynamic contrast-enhanced magnetic resonance imaging-based radiomics.

Macrotrabecular-massive hepatocellular carcinoma (MTM-HCC) is closely related to aggressive phenotype, gene mutation, carcinogenic pathway, and immunohistochemical markers and is a strong independent predictor of early recurrence and poor prognosis. With the development of imaging technology, successful applications of contrast-enhanced magnetic resonance imaging (MRI) have been reported in identifying the MTM-HCC subtype. Radiomics, as an objective and beneficial method for tumour evaluation, is used to convert medical images into high-throughput quantification features that greatly push the development of precision medicine. To establish and verify a nomogram for preoperatively identifying MTM-HCC by comparing different machine learning algorithms. This retrospective study enrolled 232 (training set, 162; test set, 70) hepatocellular carcinoma patients from April 2018 to September 2021. A total of 3111 radiomics features were extracted from dynamic contrast-enhanced MRI, followed by dimension reduction of these features. Logistic regression (LR), K-nearest neighbour (KNN), Bayes, Tree, and support vector machine (SVM) algorithms were used to select the best radiomics signature. We used the relative standard deviation (RSD) and bootstrap methods to quantify the stability of these five algorithms. The algorithm with the lowest RSD represented the best stability, and it was used to construct the best radiomics model. Multivariable logistic analysis was used to select the useful clinical and radiological features, and different predictive models were established. Finally, the predictive performances of the different models were assessed by evaluating the area under the curve (AUC). The RSD values based on LR, KNN, Bayes, Tree, and SVM were 3.8%, 8.6%, 4.3%, 17.7%, and 17.4%, respectively. Therefore, the LR machine learning algorithm was selected to construct the best radiomics signature, which performed well with AUCs of 0.766 and 0.739 in the training and test sets, respectively. In the multivariable analysis, age [odds ratio (OR) = 0.956, P = 0.034], alpha-fetoprotein (OR = 10.066, P < 0.001), tumour size (OR = 3.316, P = 0.002), tumour-to-liver apparent diffusion coefficient (ADC) ratio (OR = 0.156, P = 0.037), and radiomics score (OR = 2.923, P < 0.001) were independent predictors of MTM-HCC. Among the different models, the predictive performances of the clinical-radiomics model and radiological-radiomics model were significantly improved compared to those of the clinical model (AUCs: 0.888 vs 0.836, P = 0.046) and radiological model (AUCs: 0.796 vs 0.688, P = 0.012), respectively, in the training set, highlighting the improved predictive performance of radiomics. The nomogram performed best, with AUCs of 0.896 and 0.805 in the training and test sets, respectively. The nomogram containing radiomics, age, alpha-fetoprotein, tumour size, and tumour-to-liver ADC ratio revealed excellent predictive ability in preoperatively identifying the MTM-HCC subtype.

Loan Amount Prediction Based on Credit History and Other Variables

Loan approval is a very important process for banking organizations. The system approved or reject the loan applications. Recovery of loans is a major contributing parameter in the financial statements of a bank. It is very difficult to predict the possibility of payment of loan by the customer. In recent years many researchers worked on loan approvalprediction systems. Machine Learning (ML)techniques are very useful in predicting outcomes for large amount of data. In this paper three machine learning algorithms, Logistic Regression (LR), Decision Tree (DT) and Random Forest (RF) are applied to predict the loan approval of customers. The experimental results conclude that the accuracy of Logistic Regression machine learning algorithm is better as compared to Decision Tree and Random Forest machine learning approaches.

Characterization of the microenvironment of diabetic foot ulcers and potential drug identification based on scRNA-seq.

Diabetes foot ulcers (DFUs) are a type of foot infection, ulcer, and/or deep tissue destruction caused by neuropathy and vascular disease in the distal extremities of diabetic patients. Its pathogenesis and its microenvironment are not entirely understood. Initially, the GSE165816 data set from the GEO database was utilized for single cell analysis to reveal the microenvironment and functional status of DFUs. The GSE199939 RNA-seq data set was utilized for external validation. On the basis of the logistic regression machine learning algorithm (OCLR), pseudo time series analysis, dryness index analysis, and drug target gene analysis were then performed. By constructing drug-gene and gene-gene networks, we can locate the most recent DFUs treatments. Finally, immunofluorescence technology was used to detect the cell-related markers of the DFUs microenvironment, and qPCR was used to detect the expression of drug targets in DFUs. Firstly, we used the Cell Maker database to obtain information about human cells and related gene markers, and manually reviewed a total of 45 kinds of cells and maker information that may appear in the DFUs microenvironment, which were divided into 17 cell clusters after annotation. Subsequently, we counted the proportions of DM and DFUs in different types of cells, and the results showed that the proportions of macrophages, white blood cells, and monocytes were higher in patients with DFUs, while the proportions of pluripotent stem cells and stromal cells were higher in patients with DM. The Pseudo-time series analysis of cells in DFUs showed that the differentiation pathways of immune cells, mesenchymal cells and stem cells were similar in the three states, while the other cells were distributed in different stages. At the level of a single cell, the scores of both multipotential stem cells and hematopoietic stem cells were significantly lower in DFU healing and non-healing than in DM. Additionally, the highly expressed genes in DFU were chosen as drug targets. We identified seven potential target genes and discovered twenty drugs with high significance. Finally, the colocalization relationship between CD19, ITGAM, and HLA-DR expression in monocytes and macrophages of DFU skin tissue and healthy subjects was analyzed by laser confocal microscopy with the immunofluorescence triple labeling method. The results showed that the expressions of CD19, ITGAM, and HLA-DR in the skin of DFUs were significantly higher than those in the skin of healthy subjects, and the co-localization relationship was significant in DFUs. This study can serve as a resource for the treatment of DFUs.

The Comparison of Accuracy on Classification Climate Change Data with Logistic Regression

Machine learning methods can be used to generate climate change models. The goal of this study is to use logistic regression machine learning algorithms to classify data on greenhouse gas emissions. The data used is climate change data of several countries obtained from The World Bank, with total greenhouse gas emissions as the response variable and 61 other attributes as explanatory variables. This data is preprocessed using min-max normalization to handle unbalanced ranges, and then the data is split into 70% training data and 30% testing data. Based on the logistic regression modeling, it was discovered that the data from the min-max transformation resulted in better modeling than the data modeling without the transformation process. The accuracy, precision, sensitivity, and specificity of the transformation are 87.60%, 87.76%, 87.04%, and 88.14%, respectively

Predicting In-Hospital Mortality After Acute Myeloid Leukemia Therapy: Through Supervised Machine Learning Algorithms

Despite careful patient selection, induction chemotherapy for acute myeloid leukemia (AML) is associated with a considerable risk for treatment-related mortality (5%-20%). We evaluated machine learning (ML) algorithms trained using factors available at the time of admission for AML therapy to predict death during the hospitalization. We included AML discharges with age > 17 years who received inpatient chemotherapy from State Inpatient Database from Arizona, Florida, New York, Maryland, Washington, and New Jersey for years 2008-2014. The primary objective was to predict inpatient mortality in patients undergoing chemotherapy using covariates present before initiation of chemotherapy. ML algorithms logistic regression (LR), decision tree, and random forest were compared. 29,613 hospitalizations for patients with AML were included in the analysis each with 4,177 features. The median age was 58.9 (18-101) years, 13,689 (53.7%) were male, and 20,203 (69%) were White. The mean time from admission to chemotherapy was 3 days (95% CI, 2.9 to 3.1), and 2,682 (9.1%) died during the hospitalization. Both LR and random forest models achieved an area under the curve (AUC) score of 0.78, whereas decision tree achieved an AUC of 0.70. The baseline LR model with age yielded an AUC of 0.62. To clinically balance and minimize false positives, we selected a decision threshold of 0.7 and at this threshold, 51 of our test set of 5,923 could have potentially averted treatment-related mortality. Using readily accessible variables, inpatient mortality of patients on track for chemotherapy to treat AML can be predicted through ML algorithms. The model also predicted inpatient mortality when tested on different data representations and paves the way for future research.

Down-Regulation of lncRNA MBNL1-AS1 Promotes Tumor Stem Cell-like Characteristics and Prostate Cancer Progression through miR-221-3p/CDKN1B/C-myc Axis

The recurrence, progression, and drug resistance of prostate cancer (PC) is closely related to the cancer stem cells (CSCs). Therefore, it is necessary to find the key regulators of prostate cancer stem cells (PCSCs). Here, we analyzed the results of a single-class logistic regression machine learning algorithm (OCLR) to identify the PCSC-associated lncRNA MBNL1-AS1. The effects of MBNL1-AS1 on the stemness of CSCs was assessed using qPCR, western blot and sphere-forming assays. The role of MBNL1-AS1 in mediating the proliferation and invasion of the PC cell lines was examined using Transwell, wounding-healing, CCK-8, EdU and animal assays. Dual-luciferase and ChIRP assays were used to examine the molecular mechanism of MBNL1-AS1 in PCSCs. MBNL1-AS1 was shown to be negatively correlated with stemness index (mRNAsi), and even prognosis, tumor progression, recurrence, and drug resistance in PC patients. The knockdown of MBNL1-AS1 significantly affected the stemness of the PC cells, and subsequently their invasive and proliferative abilities. Molecular mechanism studies suggested that MBNL1-AS1 regulates CDKN1B through competitive binding to miR-221-3p, which led to the inhibition of the Wnt signaling pathway to affect PCSCs. In conclusion, our study identified MBNL1-AS1 as a key regulator of PCSCs and examined its mechanism of action in the malignant progression of PC.

Machine learning outperforms traditional logistic regression and offers new possibilities for cardiovascular risk prediction: A study involving 143,043 Chinese patients with hypertension.

Identifying people at risk of cardiovascular diseases (CVD) is a cornerstone of preventive cardiology. We developed machine learning (ML) algorithms and investigated their performance in predicting patients' current CVD risk (coronary heart disease and stroke in this study). We compared traditional logistic regression (LR) with five ML algorithms LR with Elastic-Net, Random Forest (RF), XGBoost (XGB), Support Vector Machine, Deep Learning, and an Ensemble model averaging predictions from RF, XGB, and Deep Learning for CVD risk prediction using pre-existing patient-level data from a multi-center, cross-sectional study (the Microalbuminuria Screening in Hypertensive Patients Project initiated by the China International Exchange and Promotive Association for Medical and Healthcare) that enrolled 143,043 patients with hypertension from 600 tertiary, secondary, or community hospitals. Each of the five ML algorithms incorporated 18 variables, such as demographics, examinations, comorbidities, and treatment regimens, and were trained and evaluated using 5-fold cross-validation. Predictive accuracy was assessed by the area under the receiver operating curve (AUROC). Patients' mean age was 62 ± 12 years and 57% were men. Advanced ML algorithms outperformed the traditional LR model. Particularly, the Ensemble model had superior discrimination with an AUROC of 0.760 than LR (AUC = 0.737) and other tested models. We establishes an Ensemble model that shows better performance in predicting patients' current CVD risk using routine information compared to the traditional LR model. ML can help physicians design follow-up plans with more accurate results, offering new possibilities for short-term risk prediction and early detection. Further, ML models can be trained with longitudinal data and used to predict long-term CVD risks, thereby informing CVD prevention.

A Rough Set Approach to Dimensionality Reduction for Performance Enhancement in Machine Learning

Machine learning uses complex mathematical algorithms to turn data set into a model for a problem domain. Analysing high dimensional data in their raw form usually causes computational overhead because the higher the size of the data, the higher the time it takes to process it. Therefore, there is a need for a more robust dimensionality reduction approach, among other existing methods, for feature projection (extraction) and selection from data set, which can be passed to a machine learning algorithm for optimal performance. This paper presents a generic mathematical approach for transforming data from a high dimensional space to low dimensional space in such a manner that the intrinsic dimension of the original data is preserved using the concept of indiscernibility, reducts, and the core of the rough set theory. The flue detection dataset available on the Kaggle website was used in this research for demonstration purposes. The original and reduced datasets were tested using a logistic regression machine learning algorithm yielding the same accuracy of 97% with a training time of 25 min and 11 min respectively.

Predicting responders to prone positioning in mechanically ventilated patients with COVID-19 using machine learning

BackgroundFor mechanically ventilated critically ill COVID-19 patients, prone positioning has quickly become an important treatment strategy, however, prone positioning is labor intensive and comes with potential adverse effects. Therefore, identifying which critically ill intubated COVID-19 patients will benefit may help allocate labor resources.MethodsFrom the multi-center Dutch Data Warehouse of COVID-19 ICU patients from 25 hospitals, we selected all 3619 episodes of prone positioning in 1142 invasively mechanically ventilated patients. We excluded episodes longer than 24 h. Berlin ARDS criteria were not formally documented. We used supervised machine learning algorithms Logistic Regression, Random Forest, Naive Bayes, K-Nearest Neighbors, Support Vector Machine and Extreme Gradient Boosting on readily available and clinically relevant features to predict success of prone positioning after 4 h (window of 1 to 7 h) based on various possible outcomes. These outcomes were defined as improvements of at least 10% in PaO2/FiO2 ratio, ventilatory ratio, respiratory system compliance, or mechanical power. Separate models were created for each of these outcomes. Re-supination within 4 h after pronation was labeled as failure. We also developed models using a 20 mmHg improvement cut-off for PaO2/FiO2 ratio and using a combined outcome parameter. For all models, we evaluated feature importance expressed as contribution to predictive performance based on their relative ranking.ResultsThe median duration of prone episodes was 17 h (11–20, median and IQR, N = 2632). Despite extensive modeling using a plethora of machine learning techniques and a large number of potentially clinically relevant features, discrimination between responders and non-responders remained poor with an area under the receiver operator characteristic curve of 0.62 for PaO2/FiO2 ratio using Logistic Regression, Random Forest and XGBoost. Feature importance was inconsistent between models for different outcomes. Notably, not even being a previous responder to prone positioning, or PEEP-levels before prone positioning, provided any meaningful contribution to predicting a successful next proning episode.ConclusionsIn mechanically ventilated COVID-19 patients, predicting the success of prone positioning using clinically relevant and readily available parameters from electronic health records is currently not feasible. Given the current evidence base, a liberal approach to proning in all patients with severe COVID-19 ARDS is therefore justified and in particular regardless of previous results of proning.

Differences in brain connectivity of meditators during assessing neurocognition via gamified experimental logic task: A machine learning approach.

Meditation practices increase attention, memory, and self-awareness. The neuroscientific study of meditation has helped gain useful insights into the functional changes in the brain. In this study, we have assessed the performance of meditators with different years of practice while performing an engaging task rather than studying the meditation practice itself. This task helps assess many neural processes simultaneously and represents task performance in presence of multiple audio-visual distractors as in a real-life scenario. The long-term practice of meditation could bring neuroplastic changes in the way cognitive processing is carried out. It could be conscious and effortful in short-term practitioners and relatively unconscious and effortless in long-term practitioners. Our goal is to understand if it is possible to differentiate between long-term and short-term meditators solely based on their cognitive processing. A group of proficient Rajayoga meditators from the Brahma Kumaris were recruited based on their meditation experience-Long-Term Practitioners (n = 12, mean 13,596h) and Short-Term Practitioners (n = 10, mean 1095h). A task-based functional Magnetic Resonance Imaging was acquired while the subjects performed the task. Functional Connectivity Analysis was performed to derive the correlation measures to be used as features for classification. Five supervised Machine Learning algorithms Logistic Regression, Support Vector Machine, Decision Tree, Random Forest, and Gradient Boosted Tree were used for classification. Among all the classifiers Gradient Boosted Tree performed the best with an accuracy of 77% when all the four Functional Connectivity Metrics were used. Connectivity in visual areas, cerebellum, left rostral prefrontal cortex, and middle frontal gyrus was found to be higher in long-term meditators. Such a classification demonstrates that long-term meditation practice brings about neuroplastic changes that influence cognitive processing.

Feasibility of Machine Learning in the Prediction of Short-Term Outcomes Following Anterior Cervical Discectomy and Fusion

Increased emphasis is being placed on efficiency and resource utilization when performing anterior cervical discectomy and fusion (ACDF), and accurate prediction of complications is increasingly important to optimize care. This study aimed to compare predictive models for postoperative complications following ACDF using machine learning (ML) models based on traditional comorbidity indices. In this retrospective case series, the American College of Surgeons National Surgical Quality Improvement Program database was queried between 2011 and 2017 for all elective, primary ACDF cases. Levels of surgery, use of interbody implants, and graft selection were calculated by procedural codes. Six ML algorithms were constructed using available preoperative and intraoperative features. The overall dataset was randomly split into training (80%) and validation (20%) subsets wherein the training subset optimized the model, and the validation subset was evaluated for accuracy. ML models were compared with models constructed from American Society of Anesthesiologists classification and frailty index alone. There were 42,194 ACDF cases eligible for inclusion. Mean age was 47.7 ± 11.6 years, body mass index was 30.4 ± 6.7, and levels of operation were 1.6 ± 0.7. ML algorithms uniformly outperformed comorbidity indices in predicting complications. Logistic regression ML algorithm was the best performing for predicting any adverse event (area under the curve [AUC] 0.73), transfusion (AUC 0.90), surgical site infection (AUC 0.63), and pneumonia (AUC 0.80). Gradient boosting trees ML algorithm was the best performing for predicting extended length of stay (AUC 0.73). ML algorithms modeled the development of postoperative adverse events with superior accuracy to that of comorbidity indices and may guide preoperative clinical decision making before ACDF.