[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-125263-en":3,"doc-seo-125263-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},125263,687197207057,"Sage","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","Leveraging interpretable machine learning in intensive care - Interpretable vs black-box evaluation","Healthcare, particularly in intensive care units (ICUs), depends on medical professionals making informed decisions under complex and high-dimensional data conditions. This work assesses whether interpretable machine learning can deliver reliable predictions without losing transparency. Using the Medical Information Mart for Intensive Care dataset, the study compares interpretable generalized additive models (GAMs) against black-box boosted decision trees and random forests for mortality and length-of-stay prediction. Evaluation covers predictive performance, effects of compact feature sets, and interpretability aligned with medical knowledge, showing only a small performance decrease while preserving full interpretability.","Annals of Operations Research (2025) 347:1093–1132  \n[https://doi.org/10.1007/s10479-024-06226-8](https://doi.org/10.1007/s10479-024-06226-8)  \nORIGINAL RESEARCH  \nLeveraging interpretable machine learning in intensive care  \nLasse Bohlen1 · Julian Rosenberger2 · Patrick Zschech3 · Mathias Kraus2  \nReceived: 31 March 2023 / Accepted: 15 August 2024 / Published online: 19 September 2024 © The Author(s) 2024  \nAbstract  \nIn healthcare, especially within intensive care units (ICU), informed decision-making by medical professionals is crucial due to the complexity of medical data. Healthcare analytics seeks to support these decisions by generating accurate predictions through advanced machine learning (ML) models, such as boosted decision trees and random forests. While these models frequently exhibit accurate predictions across various medical tasks, they often lack interpretability. To address this challenge, researchers have developed interpretable ML models that balance accuracy and interpretability. In this study, we evaluate the performance gap between interpretable and black-box models in two healthcare prediction tasks, mortality and length-of-stay prediction in ICU settings. We focus specifically on the family of generalized additive models (GAMs) as powerful interpretable ML models. Our assessment uses the publicly available Medical Information Mart for Intensive Care dataset, and we analyze the models based on (i) predictive performance,(ii) the influence of compact feature sets (i.e., only few features) on predictive performance, and (iii) interpretability and consistency with medical knowledge. Our results show that interpretable models achieve competitive performance, with a minor decrease of 0.2–0.9 percentage points in area under the receiver operating characteristic relative to state-of-the-art black-box models, while preserving complete interpretability. This remains true even for parsimonious models that use only 2.2% of patient features. Our study highlights the potential of interpretable models to improve decision-making in ICUs by providing medical professionals with easily understandable and verifiable predictions.  \nKeywords Healthcare analytics · Interpretable machine learning · Generalized additive models · Length-of-stay prediction · Mortality prediction  \nLasse Bohlen and Julian Rosenberger have contributed equally to this work.  \nB  \n1  \n2  \n3  \nLasse Bohlen  \n[lasse.bohlen@fau.de](lasse.bohlen@fau.de)  \nJulian Rosenberger  \n[julian.rosenberger@ur.de](julian.rosenberger@ur.de)  \nPatrick Zschech  \n[patrick.zschech@uni-leipzig.de](patrick.zschech@uni-leipzig.de)  \nMathias Kraus  \n[mathias.kraus@ur.de](mathias.kraus@ur.de)  \nFriedrich-Alexander-Universität Erlangen-Nürnberg, Lange Gasse 20, 90403 Nürnberg, Germany Universität Regensburg, Bajuwarenstraße 4, 93053 Regensburg, Germany  \nUniversität Leipzig, Grimmaische Str. 12, 04109 Leipzig, Germany  \n1 Introduction  \nHealthcare systems across the globe face a multitude of complex challenges, such as care quality disparities, demographic shifts, and administrative obstacles including resource shortages, rising costs, and insufficient infrastructure (Roncarolo et al., 2017) . These issues are intensified by increasing demand for healthcare services, sophisticated medical technology complicating physicians’workflows, and heightened expectations for patient-centered care. In the context of intensive care units (ICUs), these challenges are further heightened due to patient acuity, the need for specialized staff, and pressure on resource allocation. ICUs account for approximately 14% of total hospital expenses, making them one of the most critical and costly components of healthcare systems (Halpern & Pastores, 2010) . Consequently, effective ICU management is essential for optimizing patient outcomes and ensuring healthcare systems’ sustainability (Bertsimas et al., 2021) . However, optimizing resource utilization in ICUs remains a daunting task due to the urgent and","cbCaieaNTcneVbM9","https://ap.wps.com/l/cbCaieaNTcneVbM9","pdf",1565280,1,40,"English","en",105,"# Introduction\n## Challenges in ICU decision-making\n## Machine learning for ICU predictions\n## Model parsimony and interpretability goals\n# Study approach and evaluation (abstract-based)","[{\"question\":\"Why is interpretability important for machine learning in intensive care units (ICUs)?\",\"answer\":\"ICU decisions rely on complex medical data, and many high-performing ML models are black boxes whose decision logic is hard for humans to understand. Interpretability helps medical professionals trust and verify predictions.\"},{\"question\":\"Which interpretable model family does the study focus on?\",\"answer\":\"The study focuses on generalized additive models (GAMs) as interpretable machine learning models that aim to balance accuracy and transparency.\"},{\"question\":\"How does the study evaluate interpretable vs black-box models?\",\"answer\":\"Models are compared on predictive performance, how performance changes with compact feature sets, and interpretability and consistency with medical knowledge, using mortality and length-of-stay tasks in ICU settings.\"}]","Leveraging interpretable machine learning in intensive care - Interpretable vs black-box evaluation | PDF",1785897787,101,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"leveraging-interpretable-machine-learning-in-intensive-care-interpretable-vs-black-box-evaluation","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/leveraging-interpretable-machine-learning-in-intensive-care-interpretable-vs-black-box-evaluation/125263/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why is interpretability important for machine learning in intensive care units (ICUs)?","Question",{"text":75,"@type":76},"ICU decisions rely on complex medical data, and many high-performing ML models are black boxes whose decision logic is hard for humans to understand. Interpretability helps medical professionals trust and verify predictions.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Which interpretable model family does the study focus on?",{"text":80,"@type":76},"The study focuses on generalized additive models (GAMs) as interpretable machine learning models that aim to balance accuracy and transparency.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the study evaluate interpretable vs black-box models?",{"text":84,"@type":76},"Models are compared on predictive performance, how performance changes with compact feature sets, and interpretability and consistency with medical knowledge, using mortality and length-of-stay tasks in ICU settings.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,119,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":21,"slug":118},7,"Healthcare","healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]