[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82382-en":3,"doc-seo-82382-105":29,"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":13,"seo_description":14,"update_tm":27,"read_time":28},82382,1099514068365,"Aurelia","https://ap-avatar.wpscdn.com/avatar/10000253d8d9f28188e?_k=1776742907772140068",8,"Research & Report","SAGEAgent: A Self-Evolving Agent for Cost-Aware Modality Acquisition in Multimodal Survival Prediction","Multimodal clinical oncology questions whether every cancer patient needs a complete diagnostic workup for accurate survival prediction. Existing methods either assume all modalities are available or treat missing data passively, without reasoning about whether the next test in an ordered workflow is justified for a specific patient. SAGEAgent (Sequential Acquisition Guided by Experience) is a training-free, self-evolving LLM-based agent that sequentially decides acquire-versus-stop to balance predictive accuracy with clinical invasiveness. Across glioma cohorts using TCGA-LGG, TCGA-GBM, and BraTS, it preserves competitive accuracy while cutting average acquisition burden by 55%.","arXiv :2607 .0952 1v 1 [ cs .AI] 10 Jul 2026  \nSAGEAgent: A Self-Evolving Agent for Cost-Aware Modality Acquisition in Multimodal Survival Prediction  \nChongyu Qu 1 , Can Cui 1 , Zhengyi Lu 1 , Junchao Zhu 1 , Tianyuan Yao 1 , Junlin Guo 1 , Juming Xiong 1 , Yanfan Zhu 1 , Yuechen Yang 1 , Bennett A. Landman 1 ,2 ,  \nand Yuankai Huo 1 B  \n1 Vanderbilt University, Nashville TN 37235, USA  \n2 Vanderbilt University Medical Center, Nashville TN, 37232 USA  \n[yuankai.huo@vanderbilt.edu](yuankai.huo@vanderbilt.edu)  \nAbstract. Does every cancer patient truly need a complete diagnostic workup for accurate survival prediction? In multimodal clinical oncology, diagnostic modalities follow a clinically mandated order of escalating burden—from demographics collected at intake to genomic profiling requiring specialized tissue analysis. Current multimodal survival methods either assume all modalities are available or passively handle missing data, but none actively reason about whether acquiring the next modality is justified for a given patient along this ordered workflow. We formulate this as a sequential decision problem and propose SAGEAgent (Sequential Acquisition Guided by Experience), a selfevolving LLM-based clinical agent that decides which diagnostic modalities to acquire for each patient, balancing predictive accuracy against clinical invasiveness. SAGEAgent reasons about each patient’s evolving diagnostic state through clinical tools that translate numerical predictions into text, an episodic memory that retrieves similar past cases, anda semantic memory that accumulates reusable decision patterns from experience. Experiments on a glioma cohort combining TCGA-LGG, TCGA-GBM, and BraTS with four diagnostic modalities demonstrate that SAGEAgent achieves competitive survival prediction accuracy while reducing average acquisition burden by 55% . Code is publicly available at: [https://github.com/Chongyu1117/SAGEAgent](https://github.com/Chongyu1117/SAGEAgent)  \nKeywords: Modality Acquisition · LLM Agent · Survival Prediction.  \n1 Introduction  \nAccurate survival prediction in oncology increasingly relies on integrating multiple diagnostic modalities [3, 15], ranging from routine demographics and radiology to invasive biopsy and molecular profiling [16] . These modalities follow a strict clinical order: imaging precedes biopsy, which in turn provides the specimen for genomic analysis. Previous multimodal fusion methods [5,4 ,8] combine all available modalities through a fixed strategy or handle missing data at test  \n2 C. Qu et al.  \nFig. 1. Overview of modality acquisition strategies. (a) Existing approaches: static fusion applies all modalities uniformly, RL learns adaptive policies without explanation, and LLM-based methods reason over independent tests without clinical ordering. (b) SAGEAgent follows a clinically acquisition order, deciding at each stage whether to acquire or stop based on clinical tools, learned rules, and similar past cases. Patient A stops early (83% burden saved); Patient B proceeds through all modalities.  \ntime, but treat prediction as a static problem—the model outputs a risk score from whatever is observed, without reasoning about whether the next diagnostic step is worth it for this specific patient.  \nSequential test acquisition has been studied from two directions. Reinforcement learning (RL) methods [2,20] train cost-aware acquisition policies but produce black-box decisions that limit clinical trust. Large language model (LLM) based approaches [13, 18 , 1] offer transparent reasoning but operate on independent diagnostic tests without accounting for the sequential dependencies inherent in clinical modality acquisition. Neither direction models the clinically mandated ordering (see Fig. 1(a) for a comparison) .  \nWe propose SAGEAgent (Sequential Acquisition Guided by Experience), a training-free, self-evolving LLM-based framework for cost-aware sequential modality acquisition in multimodal survival pr","cbCaiowi8SHzCoZZ","https://ap.wps.com/l/cbCaiowi8SHzCoZZ","pdf",2741493,1,10,"English","en",105,"# 1 Introduction\n## Clinical motivation and limitations of existing approaches\n## Proposed approach and sequential acquisition order\n# 2 Method Overview\n## SAGEAgent framework at each clinical stage\n## Memory components and self-evolution\n# 3 Architecture and Reasoning\n## Agent flow and signal sources","[{\"question\":\"What problem does SAGEAgent address in multimodal survival prediction?\",\"answer\":\"SAGEAgent targets whether acquiring the next diagnostic modality is necessary for each patient, instead of always using all modalities or handling missing data passively.\"},{\"question\":\"How does SAGEAgent decide whether to acquire or stop at each stage?\",\"answer\":\"At each clinical order stage (demographics → radiology → pathology → genomics), it uses clinical tools to express predictions in text, an episodic memory to retrieve similar past cases, and a semantic memory of reusable decision patterns to choose stop versus acquire.\"},{\"question\":\"What are the reported results of SAGEAgent on glioma datasets?\",\"answer\":\"On a glioma cohort combining TCGA-LGG, TCGA-GBM, and BraTS with nested 5×5 cross-validation, SAGEAgent maintains competitive survival prediction accuracy while reducing average acquisition burden by 55% compared with full workup.\"}]",1784180046,25,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"sageagent-a-self-evolving-agent-for-cost-aware-modality-acquisition-in-multimodal-survival-prediction","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/sageagent-a-self-evolving-agent-for-cost-aware-modality-acquisition-in-multimodal-survival-prediction/82382/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",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},"What problem does SAGEAgent address in multimodal survival prediction?","Question",{"text":75,"@type":76},"SAGEAgent targets whether acquiring the next diagnostic modality is necessary for each patient, instead of always using all modalities or handling missing data passively.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does SAGEAgent decide whether to acquire or stop at each stage?",{"text":80,"@type":76},"At each clinical order stage (demographics → radiology → pathology → genomics), it uses clinical tools to express predictions in text, an episodic memory to retrieve similar past cases, and a semantic memory of reusable decision patterns to choose stop versus acquire.",{"name":82,"@type":73,"acceptedAnswer":83},"What are the reported results of SAGEAgent on glioma datasets?",{"text":84,"@type":76},"On a glioma cohort combining TCGA-LGG, TCGA-GBM, and BraTS with nested 5×5 cross-validation, SAGEAgent maintains competitive survival prediction accuracy while reducing average acquisition burden by 55% compared with full workup.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":21,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":21,"slug":133},"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]