2026-10-12T07:49:04Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33865700 2026-09-16T17:31:11Z category_146 category_12 category_734 category_931 category_64 portal_5 item_type_3 month_year_09_2026

<p>Examples of best rules for predicting long liver (top) and prostate (bottom) MRI exams, with individual features highlighted. These rules were discovered by RL-XAI algorithms, which identified the optimal choices of variables and thresholds. Impact was defined as rule frequency multiplied by its average difference in duration – the impact on shortening the entire workflow. {} with arrows indicate specific acquisition subsequences contributing to longer scan times. Numbers in parentheses are 95% confidence intervals computed via 10,000 bootstrap samples. “3T Vida” refers to the specific scanner model, (shims, coils) – to scanner hardware settings, and (tse, haste, fl3d_vibe, ep2d_diff) and similar – to various scanning sequences performed by MRI scanner.</p> Sean Hartmann (19548544) Andrew Sharp (342500) Heather Johnston (11514937) Michael S. Gee (24998606) Susie Y. Huang (2699911) Mukesh G. Harisinghani (24998609) Jens Gühring (24998612) Wei-Ching Lo (832669) Vibhas Deshpande (554047) Oleg S. Pianykh (8963060) Space Science Cell Biology Biological Sciences not elsewhere classified Information Systems not elsewhere classified Cancer xlink "> n three outpatient facilities rich device logs operationally meaningful rules new scanning sequence median scan time five 3t scanners average scan duration 20 top rules 1000 top rules imaging exams longer selected boolean rules oriented optimization metrics top rule resulted rule &# 8217 optimal boolean rule learning ai models long liver exams learning explainable ai learning ai exams performed selected using explainable rule requiring optimization process optimization word environment transform feature sustainable improvements supporting process support measurable statistical significance smaller set similar optimizations prostate protocols prostate cases practical efficiency permutation test model rule main goal january 2024 january 2019 f1 score domain expertise discover short 25 minutes 001 ). <p>Examples of best rules for predicting long liver (top) and prostate (bottom) MRI exams, with individual features highlighted. These rules were discovered by RL-XAI algorithms, which identified the optimal choices of variables and thresholds. Impact was defined as rule frequency multiplied by its average difference in duration – the impact on shortening the entire workflow. {} with arrows indicate specific acquisition subsequences contributing to longer scan times. Numbers in parentheses are 95% confidence intervals computed via 10,000 bootstrap samples. “3T Vida” refers to the specific scanner model, (shims, coils) – to scanner hardware settings, and (tse, haste, fl3d_vibe, ep2d_diff) and similar – to various scanning sequences performed by MRI scanner.</p> 2026-09-16T17:31:02Z Dataset Dataset 10.1371/journal.pone.0358070.t002 https://figshare.com/articles/dataset/_p_Examples_of_best_rules_for_predicting_long_liver_top_and_prostate_bottom_MRI_exams_with_individual_features_highlighted_These_rules_were_discovered_by_RL-XAI_algorithms_which_identified_the_optimal_choices_of_variables_and_thresholds_Imp/33865700 CC BY 4.0