Machine learning predicts cancer-associated deep vein thrombosis using clinically available variables.

PURPOSE: To develop and validate machine learning (ML) models for cancer-associated deep vein thrombosis (DVT) and to compare the performance of these models with the Khorana score (KS). METHODS: We randomly extracted data of 2100 patients with cancer between Jan. 1, 2017, and Oct. 31, 2019, and 1035 patients who underwent Doppler ultrasonography were enrolled. Univariate analysis and Lasso regression were applied to select important predictors. Model training and hyperparameter tuning were implemented on 70% of the data using a ten-fold cross-validation method. The remaining 30% of the data were used to compare the performance with seven indicators (area under the receiver operating characteristic curve [AUC], sensitivity, specificity, accuracy, balanced accuracy, Brier score, and calibration curve), among all five ML models (linear discriminant analysis [LDA], logistic regression [LR], classification tree [CT], random forest [RF], and support vector machine [SVM]), and the KS. RESULTS: The incidence of cancer-associated DVT was 22.3%. The top five predictors were D-dimer level, age, Charlson Comorbidity Index (CCI), length of stay (LOS), and previous VTE (venous thromboembolism) history according to RF. Only LDA (AUC = 0.773) and LR (AUC = 0.772) outperformed KS (AUC = 0.642), and combination with D-dimer showed improved performance in all models. A nomogram and web calculator https://webcalculatorofcancerassociateddvt.shinyapps.io/dynnomapp/ were used to visualize the best recommended LR model. CONCLUSION: This study developed and validated cancer-associated DVT predictive models using five ML algorithms and visualized the best recommended model using a nomogram and web calculator. The nomogram and web calculator developed in this study may assist doctors and nurses in evaluating individualized cancer-associated DVT risk and making decisions. However, other prospective cohort studies should be conducted to externally validate the recommended model.

Predicting disease risks by matching quantiles estimation for censored data.

In time to event data analysis, it is often of interest to predict quantities such as t-year survival rate or the survival function over a continuum of time. A commonly used approach is to relate the survival time to the covariates by a semiparametric regression model and then use the fitted model for prediction, which usually results in direct estimation of the conditional hazard function or the conditional estimating equation. Its prediction accuracy, however, relies on the correct specification of the covariate-survival association which is often difficult in practice, especially when patient populations are heterogeneous or the underlying model is complex. In this paper, from a prediction perspective, we propose a disease-risk prediction approach by matching an optimal combination of covariates with the survival time in terms of distribution quantiles. The proposed method is easy to implement and works flexibly without assuming a priori model. The redistribution-of-mass technique is adopted to accommodate censoring. We establish theoretical properties of the proposed method. Simulation studies and a real data example are also provided to further illustrate its practical utilities.

[Characteristics of mass distributions of aerosol particle and its inorganic water-soluble ions in summer over a suburb farmland in Beijing].

Agricultural activity is one of the important sources of aerosol particle. To understand the mass distribution and sources of aerosol particle and its inorganic water-soluble ions in the suburb farmland of Beijing, particle samples were collected with a MOUDI cascade impactor in the summer of 2004 in a suburb vegetable field. The mass distributions of the particle and its inorganic water-soluble ions in the diameter range of 0.18 to approximately 18 microm were measured. The dominant ions in the fine particle were SO4(2-), NOS3(-) and NH4+. The association of day to day variation of the concentration of these ions with temperature, humidity and solar radiation suggests that they are formed by the reaction of NH3 released from the vegetable field with the acid species produced from photochemical reactions. K+ in the fine particle is likely from the vegetation emission and biomass burning. Ca2+, Mg2+, NO3(-) and SO4(2-) in the coarse particle are suggested to come from the mechanical process by which the soil particle entered the atmosphere, and from the reactions of the acid species at the surface of the soil particle. The results show that fertilizer and soil are possibly important factors determining the aerosol particle over the agricultural fields, and the vegetable fields in suburb Beijing could contribute significantly to the aerosol particle.