Tree-based Machine Learning Models Research Articles

Computing the Hazard Ratios Associated With Explanatory Variables Using Machine Learning Models of Survival Data.

The application of Cox proportional hazards (CoxPH) models to survival data and the derivation of hazard ratio (HR) are well established. Although nonlinear, tree-based machine learning (ML) models have been developed and applied to the survival analysis, no methodology exists for computing HRs associated with explanatory variables from such models. We describe a novel way to compute HRs from tree-based ML models using the SHapley Additive exPlanation values, which is a locally accurate and consistent methodology to quantify explanatory variables' contribution to predictions. We used three sets of publicly available survival data consisting of patients with colon, breast, or pan cancer and compared the performance of CoxPH with the state-of-the-art ML model, XGBoost. To compute the HR for explanatory variables from the XGBoost model, the SHapley Additive exPlanation values were exponentiated and the ratio of the means over the two subgroups was calculated. The CI was computed via bootstrapping the training data and generating the ML model 1,000 times. Across the three data sets, we systematically compared HRs for all explanatory variables. Open-source libraries in Python and R were used in the analyses. For the colon and breast cancer data sets, the performance of CoxPH and XGBoost was comparable, and we showed good consistency in the computed HRs. In the pan-cancer data set, we showed agreement in most variables but also an opposite finding in two of the explanatory variables between the CoxPH and XGBoost result. Subsequent Kaplan-Meier plots supported the finding of the XGBoost model. Enabling the derivation of HR from ML models can help to improve the identification of risk factors from complex survival data sets and to enhance the prediction of clinical trial outcomes.

A tree-based machine learning model to approach morphologic assessment of malignant salivary gland tumors

Malignant salivary gland tumors represent a challenge for pathologists due to their low frequency and morphologic overlap. In recent years machine learning techniques have been applied to the field of pathology to improve diagnostic performance. In the present work, we fitted a machine learning algorithm to approach the diagnosis of malignant salivary gland tumors. Twelve morphologic variables were scored across 115 samples representing the most commonly encountered malignant salivary gland tumors. The sample was randomly split into a discovery and validation set. A recursive partitioning algorithm was used to systematically screen and organize candidate variables into a classification tree using the discovery set. A cross-validation strategy was used to tune the algorithm hyperparameters. Inter-observer concordance was calculated by independent evaluation of 26 randomly selected cases. The five-tiered tree built, required the evaluation of 6 morphological variables. Basaloid appearance, presence of mucous cells, necrosis, cribriform pattern, clear cells and keratinization were selected by the algorithm to build the tree. This diagnostic tool correctly classified 89.9% and 84.6% of the samples in the discovery and validation sets respectively. Misclassification pattern was consistent between both sets. Misclassified tumors belonged to one of three histologic types: epithelial-myoepithelial, polymorphous and mucoepidermoid carcinomas. Other histotypes demonstrated perfect recall in both the discovery and validation sets. Overall inter-observer concordance was good, with median kappa scores between the expert evaluator and training pathologists being 0.81. Overall, our classification tool developed using a recursive partitioning algorithm can effectively guide the morphological approach to malignant salivary gland tumors.

Predicting Flow Rate Escalation for Pediatric Patients on High Flow Nasal Cannula Using Machine Learning.

Background: High flow nasal cannula (HFNC) is commonly used as non-invasive respiratory support in critically ill children. There are limited data to inform consensus on optimal device parameters, determinants of successful patient response, and indications for escalation of support. Clinical scores, such as the respiratory rate-oxygenation (ROX) index, have been described as a means to predict HFNC non-response, but are limited to evaluating for escalations to invasive mechanical ventilation (MV). In the presence of apparent HFNC non-response, a clinician may choose to increase the HFNC flow rate to hypothetically prevent further respiratory deterioration, transition to an alternative non-invasive interface, or intubation for MV. To date, no models have been assessed to predict subsequent escalations of HFNC flow rates after HFNC initiation.Objective: To evaluate the abilities of tree-based machine learning algorithms to predict HFNC flow rate escalations.Methods: We performed a retrospective, cohort study assessing children admitted for acute respiratory failure under 24 months of age placed on HFNC in the Johns Hopkins Children's Center pediatric intensive care unit from January 2019 through January 2020. We excluded encounters with gaps in recorded clinical data, encounters in which MV treatment occurred prior to HFNC, and cases electively intubated in the operating room. The primary study outcome was discriminatory capacity of generated machine learning algorithms to predict HFNC flow rate escalations as compared to each other and ROX indices using area under the receiver operating characteristic (AUROC) analyses. In an exploratory fashion, model feature importance rankings were assessed by comparing Shapley values.Results: Our gradient boosting model with a time window of 8 h and lead time of 1 h before HFNC flow rate escalation achieved an AUROC with a 95% confidence interval of 0.810 ± 0.003. In comparison, the ROX index achieved an AUROC of 0.525 ± 0.000.Conclusion: In this single-center, retrospective cohort study assessing children under 24 months of age receiving HFNC for acute respiratory failure, tree-based machine learning models outperformed the ROX index in predicting subsequent flow rate escalations. Further validation studies are needed to ensure generalizability for bedside application.

Machine Learning-Assisted Development of Sensitive Electrode Materials for Mixed Potential-Type NO2 Gas Sensors.

Yttrium-stabilized zirconia (YSZ)-based mixed potential-type NOx sensors have broad application prospects in automotive exhaust gas detection. Great efforts continue to be made in developing high-performance sensitive electrode materials for mixed potential-type NO2 gas sensors. However, only five kinds of new sensing electrode materials have been developed for this type of gas sensor in the last 3 years. In this work, four different tree-based machine learning models were trained to find potentially sensitive electrode materials for NO2 detection. More than 400 materials were selected from 8000 materials by the above machine learning models. To further verify the reliability of the model, 13 of these materials containing unexploited elements were selected as sensitive electrode materials for making sensors and testing their gas-sensing performances. The experimental results showed that all 13 materials exhibited good gas-sensing performance for NO2. More interestingly, an electrode material BPO4, which does not contain any metal elements, was also screened out and showed good sensing properties to NO2. In a short period of time, 13 new sensitive electrode materials for NO2 detection were targeted and screened, which was difficult to achieve by a trial-and-error procedure.

Temporal shift and accuracy of machine learning in heart transplant outcomes

Abstract Background Accurate prediction of outcomes following a heart transplant is critical to explaining risks and benefits to patients and decision-making when considering potential organ offers. Given the large number of potential variables to be considered, this task may be most efficiently performed using machine learning (ML). Purpose We trained and tested different ML algorithms to accurately predict outcomes following a cardiac transplant using the United Network of Organ Sharing (UNOS) database. Methods We included 67 939 adult and pediatric patients enrolled in the UNOS database between January 1994 and December 2016 who underwent cardiac transplantation (median age 53 [IQR 38 – 60], 72.7% males). In our models, as an input, we included 114 features that have been collected from recipients and donors prior to transplant. The primary outcome was all-cause mortality at one-year post-transplant. We evaluated three different ML methods: XGBoost, Random Forest (RF) and L2 regularized logistic regression. Algorithms were trained and tested using shuffled 10-fold cross-validation (CV) as well as rolling CV. In the rolling CV, to mimic prospective procedure, ML models were trained by incrementally adding patients according to transplant year and testing models on the data in the following year. The hyperparameters, controlling the learning process, were tuned using Bayesian optimization. Prognostic accuracy for one-year all-cause mortality was characterized using the area under the receiver-operating characteristic curve (AUC). Results In total, 8,394 patients died within 1 year of transplant. We observed a substantial difference in prognostic accuracy between the shuffled 10-fold CV and the rolling CV. In the 10-fold CV, XGBoost and RF achieved high predictive performance with AUC of 0.848 (95% CI: 0.842–0.854) and 0.891 (95% CI: 0.886–0.896), respectively. In the rolling CV, which is a more realistic setting, AUC dropped to 0.673 (95% CI: 0.661–0.684) for XGBoost and 0.670 (0.657–0.683) for RF. Predictive performance of L2 regularized logistic regression remained stable across the two CV procedures, achieving AUC 0.669 (95% CI: 0.662–0.676) in the 10-fold shuffled CV and 0.665 (95% CI: 0.649–0.680) in the rolling CV procedure (Figure). Conclusions Our study suggests that ML models could be used to predict mortality in the first year post-transplant. We also show that the choice of CV procedure is crucial for evaluating ML models, particularly in data collected over a long period of time. The difference between the shuffled and rolling CV in the predictive performance of the tree-based ML models might indicate temporal dataset shift. In the rolling CV, all three methods achieved similar predictive performance. Funding Acknowledgement Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Research Foundation Flanders (FWO)

Comparison between Deep Learning and Tree-Based Machine Learning Approaches for Landslide Susceptibility Mapping

The efficiency of deep learning and tree-based machine learning approaches has gained immense popularity in various fields. One deep learning model viz. convolution neural network (CNN), artificial neural network (ANN) and four tree-based machine learning models, namely, alternative decision tree (ADTree), classification and regression tree (CART), functional tree and logistic model tree (LMT), were used for landslide susceptibility mapping in the East Sikkim Himalaya region of India, and the results were compared. Landslide areas were delimited and mapped as landslide inventory (LIM) after gathering information from historical records and periodic field investigations. In LIM, 91 landslides were plotted and classified into training (64 landslides) and testing (27 landslides) subsets randomly to train and validate the models. A total of 21 landslides conditioning factors (LCFs) were considered as model inputs, and the results of each model were categorised under five susceptibility classes. The receiver operating characteristics curve and 21 statistical measures were used to evaluate and prioritise the models. The CNN deep learning model achieved the priority rank 1 with area under the curve of 0.918 and 0.933 by using the training and testing data, quantifying 23.02% and 14.40% area as very high and highly susceptible followed by ANN, ADtree, CART, FTree and LMT models. This research might be useful in landslide studies, especially in locations with comparable geophysical and climatological characteristics, to aid in decision making for land use planning.

Theft Prediction Model Based on Spatial Clustering to Reflect Spatial Characteristics of Adjacent Lands

Previous studies have shown that when a crime occurs, the risk of crime in adjacent areas increases. To reflect this, previous grid-based crime prediction studies combined all the cells surrounding the event location to be predicted for use in model training. However, the actual land is continuous rather than a set of independent cells as in a geographic information system. Because the patterns that occur according to the detailed method of crime vary, it is necessary to reflect the spatial characteristics of the adjacent land in crime prediction. In this study, cells with similar spatial characteristics were classified using the Max-p region model (a spatial clustering technique), and the performance was compared to the existing method using random forest (a tree-based machine learning model). According to the results, the F1 score of the model using spatial clustering increased by approximately 2%. Accordingly, there are differences in the physical environmental factors influenced by the detailed method of crime. The findings reveal that crime involving the same offender is likely to occur around the area of the original crime, indicating that a repeated crime is likely in areas with similar spatial features to the area where the crime occurred.

Molecular insights on ABL kinase activation using tree-based machine learning models and molecular docking.

Abelson kinase (c-Abl) is a non-receptor tyrosine kinase involved in several biological processes essential for cell differentiation, migration, proliferation, and survival. This enzyme's activation might be an alternative strategy for treating diseases such as neutropenia induced by chemotherapy, prostate, and breast cancer. Recently, a series of compounds that promote the activation of c-Abl has been identified, opening a promising ground for c-Abl drug development. Structure-based drug design (SBDD) and ligand-based drug design (LBDD) methodologies have significantly impacted recent drug development initiatives. Here, we combined SBDD and LBDD approaches to characterize critical chemical properties and interactions of identified c-Abl's activators. We used molecular docking simulations combined with tree-based machine learning models—decision tree, AdaBoost, and random forest to understand the c-Abl activators' structural features required for binding to myristoyl pocket, and consequently, to promote enzyme and cellular activation. We obtained predictive and robust models with Matthews correlation coefficient values higher than 0.4 for all endpoints and identified characteristics that led to constructing a structure–activity relationship model (SAR).Graphic abstract Supplementary InformationThe online version contains supplementary material available at 10.1007/s11030-021-10261-z.

Can Clinical Symptoms and Laboratory Results Predict CT Abnormality? Initial Findings Using Novel Machine Learning Techniques in Children With COVID-19 Infections.

The rapid spread of coronavirus 2019 disease (COVID-19) has manifested a global public health crisis, and chest CT has been proven to be a powerful tool for screening, triage, evaluation and prognosis in COVID-19 patients. However, CT is not only costly but also associated with an increased incidence of cancer, in particular for children. This study will question whether clinical symptoms and laboratory results can predict the CT outcomes for the pediatric patients with positive RT-PCR testing results in order to determine the necessity of CT for such a vulnerable group. Clinical data were collected from 244 consecutive pediatric patients (16 years of age and under) treated at Wuhan Children's Hospital with positive RT-PCR testing, and the chest CT were performed within 3 days of clinical data collection, from January 21 to March 8, 2020. This study was approved by the local ethics committee of Wuhan Children's Hospital. Advanced decision tree based machine learning models were developed for the prediction of CT outcomes. Results have shown that age, lymphocyte, neutrophils, ferritin and C-reactive protein are the most related clinical indicators for predicting CT outcomes for pediatric patients with positive RT-PCR testing. Our decision support system has managed to achieve an AUC of 0.84 with 0.82 accuracy and 0.84 sensitivity for predicting CT outcomes. Our model can effectively predict CT outcomes, and our findings have indicated that the use of CT should be reconsidered for pediatric patients, as it may not be indispensable.

Himawari-8-derived diurnal variations in ground-level PM<sub>2.5</sub> pollution across China using the fast space-time Light Gradient Boosting Machine (LightGBM)

Abstract. Fine particulate matter with a diameter of less than 2.5 µm (PM2.5) has been used as an important atmospheric environmental parameter mainly because of its impact on human health. PM2.5 is affected by both natural and anthropogenic factors that usually have strong diurnal variations. Such information helps toward understanding the causes of air pollution, as well as our adaptation to it. Most existing PM2.5 products have been derived from polar-orbiting satellites. This study exploits the use of the next-generation geostationary meteorological satellite Himawari-8/AHI (Advanced Himawari Imager) to document the diurnal variation in PM2.5. Given the huge volume of satellite data, based on the idea of gradient boosting, a highly efficient tree-based Light Gradient Boosting Machine (LightGBM) method by involving the spatiotemporal characteristics of air pollution, namely the space-time LightGBM (STLG) model, is developed. An hourly PM2.5 dataset for China (i.e., ChinaHighPM2.5) at a 5 km spatial resolution is derived based on Himawari-8/AHI aerosol products with additional environmental variables. Hourly PM2.5 estimates (number of data samples = 1 415 188) are well correlated with ground measurements in China (cross-validation coefficient of determination, CV-R2 = 0.85), with a root-mean-square error (RMSE) and mean absolute error (MAE) of 13.62 and 8.49 µg m−3, respectively. Our model captures well the PM2.5 diurnal variations showing that pollution increases gradually in the morning, reaching a peak at about 10:00 LT (GMT+8), then decreases steadily until sunset. The proposed approach outperforms most traditional statistical regression and tree-based machine-learning models with a much lower computational burden in terms of speed and memory, making it most suitable for routine pollution monitoring.

Early COVID-19 quarantine: A machine learning approach to model what differentiated the top 25% well-being scorers

This study focused on the interaction of demographics and well-being. Diener's subjective well-being (SWB) was successfully validated with Exploratory Graph Analysis and Confirmatory Factor Analysis to track well-being differences of the COVID-19 quarantined individuals. Six tree-based Machine Learning models were trained to classify top 25% SWB scorers during COVID-19 quarantine, after data-splitting (train 70%, test 30%). The model input variables were demographics, to avoid overlapping of inputs-outputs. A 10-fold cross-validation method (70%–30%) was then implemented in the training session to select the optimal Machine Learning model among the six tested. A CART classification was the optimal algorithm (Train-Accuracy = 0.77, Test-Accuracy = 0.75). A clean, three-node tree suggested that if someone spends time on perceived creative activities during the COVID-19 quarantine, under clearly described conditions, he/she had high probabilities to be a top subjective well-being scorer. The key importance of creative activities was subsequently cross-validated with three different model configurations: (1) a different tree-based model (Test-Accuracy =0.75); (2) a different operationalization of subjective well-being (Test-Accuracy =0.75) and (3) a different construct (depression; Test-Accuracy =0.73). This is an integrative approach to study individual differences in subjective well-being, bridging Exploratory Graph Analysis and Machine Learning in a single research cycle with multiples cross-validations.

Failure mode identification of column base plate connection using data-driven machine learning techniques

Column base plate (CBP) connection is one of the most important structural elements of steel structures since the failure of these base plate connections can result in the collapse of the entire structure. The prediction of failure mode of CBP connection plays a significant role in ductile behavior of the structure which is critical for damage assessment or retrofitting strategies after any natural hazard. This study introduces a rapid failure mode identification technique for CBP connections by exploring the recent advances in machine learning (ML) techniques. A comprehensive database is assembled with 189 available experimental results for CBP connections including various parameters affecting the CBP behavior. To establish the best classification model, a total of nine different ML algorithms such as Support vector machine, Naïve bayes, K-nearest neighbors, Decision tree, Random forest, Adaboost, XGboost, LightGBM, and Catboost are considered in this study. Comparing the developed ML models, the Decision tree based ML model is suggested in this study which has an overall accuracy of 91% for identifying the failure mode of CBP connections. It is also found that base plate thickness, embedment length, and anchor rod diameter are the influential parameters that govern the failure mode of CBP connections. Furthermore, an open-source classification model is provided to rapidly identify the failure mode of CBP connection by allowing modifications for future developments.

Predictive modelling as a tool for effective municipal waste management policy at different territorial levels

Nowadays, the European municipal waste management policy based on the circular economy paradigm demands the closing of material and financial loops at all territorial levels of public administration. The effective planning of treatment capacities (especially sorting plants, recycling, and energy recovery facilities) and municipal waste management policy requires an accurate prognosis of municipal waste generation, and therefore, the knowledge of behavioral, socio-economic, and demographic factors influencing the waste management (and recycling) behavior of households, and other municipal waste producers. To enable public bodies at different territorial levels to undertake an effective action resulting in circular economy we evaluated various factors influencing the generation of municipal waste fractions at regional, micro-regional and municipal level in the Czech Republic. Principal components were used as input for traditional models (multivariable linear regression, generalized linear model) as well as tree-based machine learning models (regression trees, random forest, gradient boosted regression trees). Study results suggest that the linear regression model usually offers a good trade-off between model accuracy and interpretability. When the most important goal of the prediction is supposed to be accuracy, the random forest is generally the best choice. The quality of developed models depends mostly on the chosen territorial level and municipal waste fraction. The performance of these models deteriorates significantly for lower territorial levels because of worse data quality and bigger variability. Only the age structure seems to be important across territorial levels and municipal waste fractions. Nevertheless, also other factors are of high significance in explaining the generation of municipal waste fractions at different territorial levels (e.g. number of economic subjects, expenditures, population density and the level of education). Therefore, there is not one single effective public policy dealing with circular economy strategy that fits all territorial levels. Public representatives should focus on policies effective at specific territorial level. However, performance of the models is poor for lower territorial levels (municipality and micro-regions). Thus, results for municipalities and micro-regions are weak and should be treated as such.

Using environmental variables and Fourier Transform Infrared Spectroscopy to predict soil organic carbon

Soil Organic Carbon (SOC) content is a key element for soil fertility and productivity, nutrient availability and potentially represents a measurement of the sink for greenhouse gas abatement. Improving our knowledge on the spatial distribution of SOC is hence essential for sustainable nutrient management and carbon storage capacity. The objective of this study was to evaluate the performance of six tree-based machine-learning models when using environmental variables (i.e., remote sensing and terrain attributes - scenario 1), Fourier Transform Infrared Spectroscopy (FTIR) data (scenario 2) and combination of environmental variables and FTIR data (scenario 3) as predictors in prediction of SOC content. The models included Random Forest, Cubist, Conditional Inference Forest, Conditional Inference Trees, Extreme Gradient Boosting and Classification, Regression Trees. Furthermore, we explored if the Bat optimization algorithm can improve the prediction accuracy of the models. The study was conducted across a 7000 ha field in the Miandoab County, Northern Iran, with a total of 80 soil samples collected systematically in a regular grid (700 × 1000 m). According to Leave-One-Out Cross-Validation, the best prediction performance was achieved by the Cubist+Bat model when environmental variables and FTIR spectra (scenario 3) were used (Coefficient of determination = 0.73, Concordance Correlation Coefficient = 0.77, Root Mean Square Error = 0.36, Mean Absolute Error = 0.31, Median Absolute Error = 0.28). FTIR data had the highest influence on the prediction accuracy of SOC. Therefore, it can be concluded that the combination of environmental variables and FTIR data with Cubist+Bat model as a precise approach to monitor SOC in semi-arid soils of Iran. The final Digital Soil Map (DSM) of SOC revealed that improvements in prediction might be possible with the collection of more soil samples in areas where the land use and topography changed over short spatial scales.

Meteorology-driven variability of air pollution (PM<sub>1</sub>) revealed with explainable machine learning

Abstract. Air pollution, in particular high concentrations of particulate matter smaller than 1 µm in diameter (PM1), continues to be a major health problem, and meteorology is known to substantially influence atmospheric PM concentrations. However, the scientific understanding of the ways in which complex interactions of meteorological factors lead to high-pollution episodes is inconclusive. In this study, a novel, data-driven approach based on empirical relationships is used to characterize and better understand the meteorology-driven component of PM1 variability. A tree-based machine learning model is set up to reproduce concentrations of speciated PM1 at a suburban site southwest of Paris, France, using meteorological variables as input features. The model is able to capture the majority of occurring variance of mean afternoon total PM1 concentrations (coefficient of determination (R2) of 0.58), with model performance depending on the individual PM1 species predicted. Based on the models, an isolation and quantification of individual, season-specific meteorological influences for process understanding at the measurement site is achieved using SHapley Additive exPlanation (SHAP) regression values. Model results suggest that winter pollution episodes are often driven by a combination of shallow mixed layer heights (MLHs), low temperatures, low wind speeds, or inflow from northeastern wind directions. Contributions of MLHs to the winter pollution episodes are quantified to be on average ∼5 µg/m3 for MLHs below <500 m a.g.l. Temperatures below freezing initiate formation processes and increase local emissions related to residential heating, amounting to a contribution to predicted PM1 concentrations of as much as ∼9 µg/m3. Northeasterly winds are found to contribute ∼5 µg/m3 to predicted PM1 concentrations (combined effects of u- and v-wind components), by advecting particles from source regions, e.g. central Europe or the Paris region. Meteorological drivers of unusually high PM1 concentrations in summer are temperatures above ∼25 ∘C (contributions of up to ∼2.5 µg/m3), dry spells of several days (maximum contributions of ∼1.5 µg/m3), and wind speeds below ∼2 m/s (maximum contributions of ∼3 µg/m3), which cause a lack of dispersion. High-resolution case studies are conducted showing a large variability of processes that can lead to high-pollution episodes. The identification of these meteorological conditions that increase air pollution could help policy makers to adapt policy measures, issue warnings to the public, or assess the effectiveness of air pollution measures.

Event-Triggered Interval-Based Anomaly Detection and Attack Identification Methods for an In-Vehicle Network

Vehicle communication technology has been steadily progressing alongside the convergence of the in-vehicle network (IVN) and wireless communication technology. The communication with various external networks further reinforces the connectivity between the inside and outside of a vehicle. However, this bears risks of malicious packet attacks on computer-assisted mechanical mechanisms that are capable of hijacking the vehicle's functions. The present study proposes a method to detect and identify abnormalities in vehicular networks based on the periodic event-triggered interval of the controller area network (CAN) messages. To this end, we first define four attack scenarios and then extract normal and abnormal driving data corresponding to these scenarios. Next, we analyze the CAN ID's event-triggered interval and measure statistical moments depending on the defined time-window. Finally, we conduct extensive evaluations of the proposed methods' performance by considering different attack scenarios and three types of machine learning models. The results demonstrate that the proposed method can effectively detect an abnormality in the IVN, with up to 99% accuracy. Our results suggest that when tree-based machine learning models are used as the classifier, the proposed method of attack identification can achieve more than 94% accuracy.

Predictive of Osteoporosis by Tree-based Machine Learning Model in Post-menopause Woman

Predictive of Osteoporosis by Tree-based Machine Learning Model in Post-menopause Woman

Healthy Aging Within an Image: Using Muscle Radiodensitometry and Lifestyle Factors to Predict Diabetes and Hypertension.

The strong age dependency of many deleterious health outcomes likely reflects the cumulative effects from a variety of risk and protective factors that occur over one's life course. This notion has become increasingly explored in the etiology of chronic disease and associated comorbidities in aging. Our recent work has shown the robust classification of individuals at risk for cardiovascular pathophysiology using CT-based soft tissue radiodensity parameters obtained from nonlinear trimodal regression analysis (NTRA). Past and present lifestyle influences the incidence of comorbidities like hypertension (HTN), diabetes (DM) and cardiac diseases. 2,943 elderly subjects from the AGES-Reykjavik study were sorted into a three-level binary-tree structure defined by: 1) lifestyle factors (smoking and self-reported physical activity level), 2) comorbid HTN or DM, and 3) cardiac pathophysiology. NTRA parameters were extracted from mid-thigh CT cross-sections to quantify radiodensitometric changes in three tissue types: lean muscle, fat, and loose-connective tissue. Between-group differences were assessed at each binary-tree level, which were then used in tree-based machine learning (ML) models to classify subjects with DM or HTN. Classification scores for detecting HTN or DM based on lifestyle factors were excellent (AUCROC: 0.978 and 0.990, respectively). Finally, tissue importance analysis underlined the comparatively-high significance of connective tissue parameters in ML classification, while predictive models of DM onset from five-year longitudinal data gave a classification accuracy of 94.9%. Altogether, this work serves as an important milestone toward the construction of predictive tools for assessing the impact of lifestyle factors and healthy aging based on a single image.

Deciphering the Allosteric Process of the Phaeodactylum tricornutum Aureochrome 1a LOV Domain.

The conformational-driven allosteric protein diatom Phaeodactylum tricornutum aureochrome 1a (PtAu1a) differs from other light-oxygen-voltage (LOV) proteins for its uncommon structural topology. The mechanism of signaling transduction in the PtAu1a LOV domain (AuLOV) including flanking helices remains unclear because of this dissimilarity, which hinders the study of PtAu1a as an optogenetic tool. To clarify this mechanism, we employed a combination of tree-based machine learning models, Markov state models, machine-learning-based community analysis, and transition path theory to quantitatively analyze the allosteric process. Our results are in good agreement with the reported experimental findings and reveal a previously overlooked Cα helix and protein linkers as important in promoting the protein conformational changes. This integrated approach can be considered as a general workflow and applied on other allosteric proteins to provide detailed information about their allosteric mechanisms.

A methodology of steel microstructure recognition using SEM images by machine learning based on textural analysis

We attempt image-based recognition of scanning electron microscopy (SEM) of low-carbon steels using machine learning (ML) methodology utilizing feature extraction with textural analysis. Specimens are subjected to eight different heat treatments to generate microstructures of martensite, upper bainite, lower bainite, and these mixed structures are selected as candidates for the proposed recognition method. Additionally, two types of SEM sources, i.e. field emission (FE) and tungsten (W), are used, and the images individually apply to the classifications. To extract features, a textural analysis based on gray-level co-occurrence matrix (GLCM) is adopted based on the captured SEM images. For evaluating GLCM features, the original SEM images are cropped into 16 mini-images, and the following three recognition schemes are then applied: recognize mini-images individually, recognize original image categories by majority vote on predictions of mini-images, and recognize original image by using feature averages of the mini-images for training. We also adopt two decision tree-based machine learning models of random forest (RF) and gradient boosting machine (GBM). The voting schemes for both models accomplish satisfactory accuracies of about 85% for eight steel microstructures. Especially, GBM is found to exhibit more stable performance than that of RF, as differences in accuracies between types of SEM sources are lower. In addition, differences between the SEM sources are discussed by focusing on important features estimated by these ML models, and the features from FE-SEM tend to be distributed more broadly than those of W-SEM. Although a textural value called homogeneity is commonly selected as an important feature, regardless of sources and models, the other important GLCM features exhibit different trends for both SEM sources. Models trained by FE- and W-SEM tend to prefer indices of high sensitivity with regards to local brightness differences and smoothness, respectively. Due to these differences in GLCM features and their importance, models trained with images from one SEM source cannot adequately recognize the other SEM images.