Package: AutoScore 1.1.0

Feng Xie

AutoScore: An Interpretable Machine Learning-Based Automatic Clinical Score Generator

A novel interpretable machine learning-based framework to automate the development of a clinical scoring model for predefined outcomes. Our novel framework consists of six modules: variable ranking with machine learning, variable transformation, score derivation, model selection, domain knowledge-based score fine-tuning, and performance evaluation.The details are described in our research paper<doi:10.2196/21798>. Users or clinicians could seamlessly generate parsimonious sparse-score risk models (i.e., risk scores), which can be easily implemented and validated in clinical practice. We hope to see its application in various medical case studies.

Authors:Feng Xie [aut, cre], Yilin Ning [aut], Han Yuan [aut], Mingxuan Liu [aut], Siqi Li [aut], Ehsan Saffari [aut], Bibhas Chakraborty [aut], Nan Liu [aut]

AutoScore_1.1.0.tar.gz
AutoScore_1.1.0.zip(r-4.7)AutoScore_1.1.0.zip(r-4.6)AutoScore_1.1.0.zip(r-4.5)
AutoScore_1.1.0.tgz(r-4.6-any)AutoScore_1.1.0.tgz(r-4.5-any)
AutoScore_1.1.0.tar.gz(r-4.7-any)AutoScore_1.1.0.tar.gz(r-4.6-any)
AutoScore_1.1.0.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION |NEWS
card.svg |card.png
AutoScore/json (API)

# Install 'AutoScore' in R:

install.packages('AutoScore', repos = c('https://nliulab.r-universe.dev', 'https://cloud.r-project.org'))

Bug tracker:https://github.com/nliulab/autoscore/issues

Datasets:

sample_data - 20000 simulated ICU admission data, with the same distribution as the data in the MIMIC-III ICU database
sample_data_ordinal - Simulated ED data with ordinal outcome
sample_data_ordinal_small - Simulated ED data with ordinal outcome
sample_data_small - 1000 simulated ICU admission data, with the same distribution as the data in the MIMIC-III ICU database
sample_data_survival - 20000 simulated MIMIC sample data with survival outcomes
sample_data_survival_small - 1000 simulated MIMIC sample data with survival outcomes
sample_data_with_missing - 20000 simulated ICU admission data with missing values

On CRAN:

7.22 score 38 stars 36 scripts 403 downloads 2 mentions 38 exports 168 dependencies

Last updated from:eb5a2d9e82. Checks:9 OK. Indexed: yes.

Target	Result	Time
linux-devel-x86_64	OK	272
source / vignettes	OK	232
linux-release-x86_64	OK	169
macos-release-arm64	OK	163
macos-oldrel-arm64	OK	220
windows-devel	OK	159
windows-release	OK	157
windows-oldrel	OK	145
wasm-release	OK	194

Exports:AutoScore_fine_tuning AutoScore_fine_tuning_Ordinal AutoScore_fine_tuning_Survival AutoScore_parsimony AutoScore_parsimony_Ordinal AutoScore_parsimony_Survival AutoScore_rank AutoScore_rank_Ordinal AutoScore_rank_Survival AutoScore_testing AutoScore_testing_Ordinal AutoScore_testing_Survival AutoScore_weighting AutoScore_weighting_Ordinal AutoScore_weighting_Survival check_data check_data_ordinal check_data_survival compute_descriptive_table compute_multi_variable_table compute_multi_variable_table_ordinal compute_multi_variable_table_survival compute_uni_variable_table compute_uni_variable_table_ordinal compute_uni_variable_table_survival conversion_table conversion_table_ordinal conversion_table_survival plot_importance plot_predicted_risk plot_survival_km print_performance_ci_survival print_performance_ordinal print_performance_survival print_roc_performance print_scoring_table split_data transform_df_fixed

Dependencies:abind askpass backports base64enc bit bit64 boot broom bslib cachem car carData checkmate class cli clipr cluster codetools colorspace commonmark corrplot cowplot cpp11 crayon crosstalk curl data.table data.tree DBI Deriv DiagrammeR digest doBy dplyr e1071 evaluate exactRankTests farver fastmap fontawesome forcats forecast foreign Formula fracdiff fs gdata generics ggplot2 ggpubr ggrepel ggsci ggsignif ggtext glue gmodels gridExtra gridtext gtable gtools haven highr Hmisc hms htmlTable htmltools htmlwidgets httr igraph isoband jpeg jquerylib jsonlite knitr labeling labelled later lattice lazyeval lifecycle litedown lme4 lmtest magrittr markdown MASS Matrix MatrixModels maxstat memoise mgcv mime minqa mitools modelr multcomp mvtnorm nlme nloptr nnet numDeriv openssl ordinal otel pbkrtest pillar pkgconfig plotly png polspline polynom prettyunits pROC progress promises proxy purrr quantreg R6 randomForest randomForestSRC rappdirs rbibutils RColorBrewer Rcpp RcppArmadillo RcppEigen Rdpack readr reformulas rlang rmarkdown rms rpart rstatix rstudioapi S7 sandwich sass scales SparseM stringi stringr survAUC survey survival survminer sys tableone TH.data tibble tidyr tidyselect timeDate tinytex tzdb ucminf urca utf8 vctrs viridisLite visNetwork vroom withr xfun xml2 yaml zoo

AutoScore: An Interpretable Machine Learning-Based Automatic Clinical Score Generator

Rendered frombrief_intro.Rmdusingknitr::rmarkdown

Last update: 2022-10-13
Started: 2022-09-21

Help page	Topics
Internal Function: Add baselines after second-step logistic regression (part of AutoScore Module 3)	add_baseline
Internal Function: Automatically assign scores to each subjects given new data set and scoring table (Used for intermediate and final evaluation)	assign_score
AutoScore STEP(iv): Fine-tune the score by revising cut_vec with domain knowledge (AutoScore Module 5)	AutoScore_fine_tuning
AutoScore STEP(iv) for ordinal outcomes: Fine-tune the score by revising 'cut_vec' with domain knowledge (AutoScore Module 5)	AutoScore_fine_tuning_Ordinal
AutoScore STEP(iv) for survival outcomes: Fine-tune the score by revising cut_vec with domain knowledge (AutoScore Module 5)	AutoScore_fine_tuning_Survival
Internal function: impute missing values in the training and validation sets	AutoScore_impute
AutoScore STEP(ii): Select the best model with parsimony plot (AutoScore Modules 2+3+4)	AutoScore_parsimony
AutoScore STEP(ii) for ordinal outcomes: Select the best model with parsimony plot (AutoScore Modules 2+3+4)	AutoScore_parsimony_Ordinal
AutoScore STEP(ii) for survival outcomes: Select the best model with parsimony plot (AutoScore Modules 2+3+4)	AutoScore_parsimony_Survival
AutoScore STEP(i): Rank variables with machine learning (AutoScore Module 1)	AutoScore_rank
AutoScore STEP (i) for ordinal outcomes: Generate variable ranking list by machine learning (AutoScore Module 1)	AutoScore_rank_Ordinal
AutoScore STEP (1) for survival outcomes: Generate variable ranking List by machine learning (Random Survival Forest) (AutoScore Module 1)	AutoScore_rank_Survival
AutoScore STEP(v): Evaluate the final score with ROC analysis (AutoScore Module 6)	AutoScore_testing
AutoScore STEP(v) for ordinal outcomes: Evaluate the final score (AutoScore Module 6)	AutoScore_testing_Ordinal
AutoScore STEP(v) for survival outcomes: Evaluate the final score with ROC analysis (AutoScore Module 6)	AutoScore_testing_Survival
AutoScore STEP(iii): Generate the initial score with the final list of variables (Re-run AutoScore Modules 2+3)	AutoScore_weighting
AutoScore STEP(iii) for ordinal outcomes: Generate the initial score with the final list of variables (Re-run AutoScore Modules 2+3)	AutoScore_weighting_Ordinal
AutoScore STEP(iii) for survival outcomes: Generate the initial score with the final list of variables (Re-run AutoScore Modules 2+3)	AutoScore_weighting_Survival
Bias-corrected and accelerated confidence intervals	bca
Internal Function: Change Reference category after first-step logistic regression (part of AutoScore Module 3)	change_reference
AutoScore function for datasets with binary outcomes: Check whether the input dataset fulfill the requirement of the AutoScore	check_data
AutoScore function for ordinal outcomes: Check whether the input dataset fulfil the requirement of the AutoScore	check_data_ordinal
AutoScore function for survival data: Check whether the input dataset fulfill the requirement of the AutoScore	check_data_survival
Internal function: Check link function	check_link
Internal function: Check predictors	check_predictor
Internal function: Compute AUC based on validation set for plotting parsimony (AutoScore Module 4)	compute_auc_val
Internal function: Compute mean AUC for ordinal outcomes based on validation set for plotting parsimony	compute_auc_val_ord
Internal function for survival outcomes: Compute AUC based on validation set for plotting parsimony	compute_auc_val_survival
AutoScore function: Descriptive Analysis	compute_descriptive_table
Internal function: Compute risk scores for ordinal data given variables selected, cut-off values and scoring table	compute_final_score_ord
Internal function: Compute mAUC for ordinal predictions	compute_mauc_ord
AutoScore function: Multivariate Analysis	compute_multi_variable_table
AutoScore-Ordinal function: Multivariate Analysis	compute_multi_variable_table_ordinal
AutoScore function for survival outcomes: Multivariate Analysis	compute_multi_variable_table_survival
Internal function: Based on given labels and scores, compute proportion of subjects observed in each outcome category in given score intervals.	compute_prob_observed
Internal function: Based on given labels and scores, compute average predicted risks in given score intervals.	compute_prob_predicted
Internal function: Compute scoring table based on training dataset (AutoScore Module 3)	compute_score_table
Internal function: Compute scoring table for ordinal outcomes based on training dataset	compute_score_table_ord
Internal function: Compute scoring table for survival outcomes based on training dataset	compute_score_table_survival
AutoScore function: Univariable Analysis	compute_uni_variable_table
AutoScore-Ordinal function: Univariable Analysis	compute_uni_variable_table_ordinal
AutoScore function for survival outcomes: Univariate Analysis	compute_uni_variable_table_survival
AutoScore function: Print conversion table based on final performance evaluation	conversion_table
AutoScore function: Print conversion table for ordinal outcomes to map score to risk	conversion_table_ordinal
AutoScore function for survival outcomes: Print conversion table	conversion_table_survival
Internal function: generate probability matrix for ordinal outcomes given thresholds, linear predictor and link function	estimate_p_mat
Internal function survival outcome: Calculate iAUC for validation set	eva_performance_iauc
Internal function: Evaluate model performance on ordinal data	evaluate_model_ord
Extract OR, CI and p-value from a proportional odds model	extract_or_ci_ord
Internal function: Find column indices in design matrix that should be 1	find_one_inds
Internal function: Compute all scores attainable.	find_possible_scores
Internal function: Calculate cut_vec from the training set (AutoScore Module 2)	get_cut_vec
Internal function: Group scores based on given score breaks, and use friendly names for first and last intervals.	group_score
Internal function: induce informative missing to sample data in the package to demonstrate how AutoScore handles missing as a separate category	induce_informative_missing
Internal function: induce informative missing in a single variable	induce_median_missing
Internal function: Inverse cloglog link	inv_cloglog
Internal function: Inverse logit link	inv_logit
Internal function: Inverse probit link	inv_probit
Internal function: Based on 'find_one_inds', make a design matrix to compute all scores attainable.	make_design_mat
Internal function: Make parsimony plot	plot_auc
Internal Function: Print plotted variable importance	plot_importance
AutoScore function for binary and ordinal outcomes: Plot predicted risk	plot_predicted_risk
Internal Function: Plotting ROC curve	plot_roc_curve
AutoScore function for survival outcomes: Print scoring performance (KM curve)	plot_survival_km
AutoScore function for survival outcomes: Print predictive performance with confidence intervals	print_performance_ci_survival
AutoScore function for ordinal outcomes: Print predictive performance	print_performance_ordinal
AutoScore function for survival outcomes: Print predictive performance	print_performance_survival
AutoScore function: Print receiver operating characteristic (ROC) performance	print_roc_performance
AutoScore Function: Print scoring tables for visualization	print_scoring_table
20000 simulated ICU admission data, with the same distribution as the data in the MIMIC-III ICU database	sample_data
Simulated ED data with ordinal outcome	sample_data_ordinal
Simulated ED data with ordinal outcome (small sample size)	sample_data_ordinal_small
1000 simulated ICU admission data, with the same distribution as the data in the MIMIC-III ICU database	sample_data_small
20000 simulated MIMIC sample data with survival outcomes	sample_data_survival
1000 simulated MIMIC sample data with survival outcomes	sample_data_survival_small
20000 simulated ICU admission data with missing values	sample_data_with_missing
AutoScore Function: Automatically splitting dataset to train, validation and test set, possibly stratified by label	split_data
Internal function: Categorizing continuous variables based on cut_vec (AutoScore Module 2)	transform_df_fixed

Package: AutoScore 1.1.0

AutoScore: An Interpretable Machine Learning-Based Automatic Clinical Score Generator

Citation

Development and contributors

Readme and manuals

Help Manual

Usage by other packages (reverse dependencies)