{"id":"https://openalex.org/W7127319731","doi":"https://doi.org/10.1109/tmi.2026.3661001","title":"Med-R1: Reinforcement Learning for Generalizable Medical Reasoning in Vision-Language Models","display_name":"Med-R1: Reinforcement Learning for Generalizable Medical Reasoning in Vision-Language Models","publication_year":2026,"publication_date":"2026-02-03","ids":{"openalex":"https://openalex.org/W7127319731","doi":"https://doi.org/10.1109/tmi.2026.3661001","pmid":"https://pubmed.ncbi.nlm.nih.gov/41632678"},"language":"en","primary_location":{"id":"doi:10.1109/tmi.2026.3661001","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tmi.2026.3661001","pdf_url":null,"source":{"id":"https://openalex.org/S58069681","display_name":"IEEE Transactions on Medical Imaging","issn_l":"0278-0062","issn":["0278-0062","1558-254X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Medical Imaging","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","pubmed"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5111304741","display_name":"Yuxiang Lai","orcid":null},"institutions":[{"id":"https://openalex.org/I150468666","display_name":"Emory University","ror":"https://ror.org/03czfpz43","country_code":"US","type":"education","lineage":["https://openalex.org/I150468666"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yuxiang Lai","raw_affiliation_strings":["Department of Computer Science and Informatics, Emory University, Atlanta, GA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science and Informatics, Emory University, Atlanta, GA, USA","institution_ids":["https://openalex.org/I150468666"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088198600","display_name":"Jike Zhong","orcid":null},"institutions":[{"id":"https://openalex.org/I1174212","display_name":"University of Southern California","ror":"https://ror.org/03taz7m60","country_code":"US","type":"education","lineage":["https://openalex.org/I1174212"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jike Zhong","raw_affiliation_strings":["Department of Computer Science, University of Southern California, Los Angeles, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science, University of Southern California, Los Angeles, CA, USA","institution_ids":["https://openalex.org/I1174212"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124897205","display_name":"Ming Li","orcid":null},"institutions":[{"id":"https://openalex.org/I74801974","display_name":"The University of Tokyo","ror":"https://ror.org/057zh3y96","country_code":"JP","type":"education","lineage":["https://openalex.org/I74801974"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Ming Li","raw_affiliation_strings":["Department of Computer Science, The University of Tokyo, Tokyo, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science, The University of Tokyo, Tokyo, Japan","institution_ids":["https://openalex.org/I74801974"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102591920","display_name":"Shitian Zhao","orcid":null},"institutions":[{"id":"https://openalex.org/I145311948","display_name":"Johns Hopkins University","ror":"https://ror.org/00za53h95","country_code":"US","type":"education","lineage":["https://openalex.org/I145311948"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shitian Zhao","raw_affiliation_strings":["Department of Computer Science, Johns Hopkins University, Baltimore, MD, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Johns Hopkins University, Baltimore, MD, USA","institution_ids":["https://openalex.org/I145311948"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5121352455","display_name":"Yuheng Li","orcid":null},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yuheng Li","raw_affiliation_strings":["Department of Biomedical Engineering, Emory and Georgia Institute of Technology, Atlanta, GA, USA"],"raw_orcid":"https://orcid.org/0009-0001-0249-8952","affiliations":[{"raw_affiliation_string":"Department of Biomedical Engineering, Emory and Georgia Institute of Technology, Atlanta, GA, USA","institution_ids":["https://openalex.org/I130701444"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5042745248","display_name":"Konstantinos Psounis","orcid":"https://orcid.org/0000-0002-6342-1078"},"institutions":[{"id":"https://openalex.org/I1174212","display_name":"University of Southern California","ror":"https://ror.org/03taz7m60","country_code":"US","type":"education","lineage":["https://openalex.org/I1174212"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Konstantinos Psounis","raw_affiliation_strings":["Department of Computer Science, University of Southern California, Los Angeles, CA, USA"],"raw_orcid":"https://orcid.org/0000-0002-6342-1078","affiliations":[{"raw_affiliation_string":"Department of Computer Science, University of Southern California, Los Angeles, CA, USA","institution_ids":["https://openalex.org/I1174212"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5104330156","display_name":"Xiaofeng Yang","orcid":null},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xiaofeng Yang","raw_affiliation_strings":["Department of Biomedical Engineering, Emory and Georgia Institute of Technology, Atlanta, GA, USA"],"raw_orcid":"https://orcid.org/0000-0001-9023-5855","affiliations":[{"raw_affiliation_string":"Department of Biomedical Engineering, Emory and Georgia Institute of Technology, Atlanta, GA, USA","institution_ids":["https://openalex.org/I130701444"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":7,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":162.5775,"has_fulltext":false,"cited_by_count":10,"citation_normalized_percentile":{"value":0.99965176,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":99,"max":100},"biblio":{"volume":"45","issue":"6","first_page":"2727","last_page":"2737"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.8683000206947327,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.8683000206947327,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.034699998795986176,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.01769999973475933,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.7328000068664551},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.6722999811172485},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.5447999835014343},{"id":"https://openalex.org/keywords/reliability","display_name":"Reliability (semiconductor)","score":0.5163000226020813},{"id":"https://openalex.org/keywords/medical-imaging","display_name":"Medical imaging","score":0.4754999876022339},{"id":"https://openalex.org/keywords/limit","display_name":"Limit (mathematics)","score":0.4528000056743622},{"id":"https://openalex.org/keywords/component","display_name":"Component (thermodynamics)","score":0.40619999170303345},{"id":"https://openalex.org/keywords/quality","display_name":"Quality (philosophy)","score":0.3971000015735626}],"concepts":[{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.7328000068664551},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7263000011444092},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.708899974822998},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6722999811172485},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.6313999891281128},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5447999835014343},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.5163000226020813},{"id":"https://openalex.org/C31601959","wikidata":"https://www.wikidata.org/wiki/Q931309","display_name":"Medical imaging","level":2,"score":0.4754999876022339},{"id":"https://openalex.org/C151201525","wikidata":"https://www.wikidata.org/wiki/Q177239","display_name":"Limit (mathematics)","level":2,"score":0.4528000056743622},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.40619999170303345},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.3971000015735626},{"id":"https://openalex.org/C534262118","wikidata":"https://www.wikidata.org/wiki/Q177719","display_name":"Medical diagnosis","level":2,"score":0.3808000087738037},{"id":"https://openalex.org/C109747225","wikidata":"https://www.wikidata.org/wiki/Q815758","display_name":"Scarcity","level":2,"score":0.36329999566078186},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.31290000677108765},{"id":"https://openalex.org/C105002631","wikidata":"https://www.wikidata.org/wiki/Q4833645","display_name":"Subject-matter expert","level":3,"score":0.29829999804496765},{"id":"https://openalex.org/C2781011336","wikidata":"https://www.wikidata.org/wiki/Q1465133","display_name":"Time limit","level":2,"score":0.28630000352859497},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2786000072956085},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.2630000114440918},{"id":"https://openalex.org/C58328972","wikidata":"https://www.wikidata.org/wiki/Q184609","display_name":"Expert system","level":2,"score":0.2533999979496002}],"mesh":[{"descriptor_ui":"D000098408","descriptor_name":"Reinforcement Machine Learning","qualifier_ui":null,"qualifier_name":null,"is_major_topic":true},{"descriptor_ui":"D000098408","descriptor_name":"Reinforcement Machine Learning","qualifier_ui":null,"qualifier_name":null,"is_major_topic":true},{"descriptor_ui":"D000465","descriptor_name":"Algorithms","qualifier_ui":null,"qualifier_name":null,"is_major_topic":false},{"descriptor_ui":"D000465","descriptor_name":"Algorithms","qualifier_ui":null,"qualifier_name":null,"is_major_topic":false},{"descriptor_ui":"D003952","descriptor_name":"Diagnostic Imaging","qualifier_ui":"Q000379","qualifier_name":"methods","is_major_topic":true},{"descriptor_ui":"D003952","descriptor_name":"Diagnostic Imaging","qualifier_ui":"Q000379","qualifier_name":"methods","is_major_topic":true},{"descriptor_ui":"D006801","descriptor_name":"Humans","qualifier_ui":null,"qualifier_name":null,"is_major_topic":false},{"descriptor_ui":"D006801","descriptor_name":"Humans","qualifier_ui":null,"qualifier_name":null,"is_major_topic":false},{"descriptor_ui":"D009323","descriptor_name":"Natural Language Processing","qualifier_ui":null,"qualifier_name":null,"is_major_topic":true},{"descriptor_ui":"D009323","descriptor_name":"Natural Language Processing","qualifier_ui":null,"qualifier_name":null,"is_major_topic":true}],"locations_count":2,"locations":[{"id":"doi:10.1109/tmi.2026.3661001","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tmi.2026.3661001","pdf_url":null,"source":{"id":"https://openalex.org/S58069681","display_name":"IEEE Transactions on Medical Imaging","issn_l":"0278-0062","issn":["0278-0062","1558-254X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Medical Imaging","raw_type":"journal-article"},{"id":"pmid:41632678","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/41632678","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE transactions on medical imaging","raw_type":null}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G1873159316","display_name":null,"funder_award_id":"R01EB032680","funder_id":"https://openalex.org/F4320332161","funder_display_name":"National Institutes of Health"},{"id":"https://openalex.org/G1956664977","display_name":null,"funder_award_id":"R01DE033512","funder_id":"https://openalex.org/F4320332161","funder_display_name":"National Institutes of Health"},{"id":"https://openalex.org/G3329856668","display_name":null,"funder_award_id":"R01CA272991","funder_id":"https://openalex.org/F4320332161","funder_display_name":"National Institutes of Health"},{"id":"https://openalex.org/G3850380546","display_name":null,"funder_award_id":"U54CA274513","funder_id":"https://openalex.org/F4320332161","funder_display_name":"National Institutes of Health"}],"funders":[{"id":"https://openalex.org/F4320332161","display_name":"National Institutes of Health","ror":"https://ror.org/01cwqze88"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Vision-language":[0],"models":[1],"(VLMs)":[2],"have":[3],"achieved":[4],"impressive":[5],"progress":[6],"in":[7,14,64,85,117,152,164,218],"natural":[8],"image":[9],"reasoning,":[10],"yet":[11],"their":[12],"potential":[13],"medical":[15,38,86,109,219],"imaging":[16,110],"remains":[17],"underexplored.":[18],"Medical":[19],"vision-language":[20],"tasks":[21],"demand":[22],"precise":[23],"understanding":[24],"and":[25,40,57,83,125,241],"clinically":[26],"coherent":[27],"answers,":[28],"which":[29],"are":[30],"difficult":[31],"to":[32,35,80,95],"achieve":[33],"due":[34],"complexity":[36],"of":[37,43,52,169,224,238],"data":[39],"the":[41,50,161,191,204,221,236,239,242,245],"scarcity":[42],"high-quality":[44],"expert":[45],"annotations.":[46,101],"These":[47,214],"challenges":[48,190],"limit":[49],"effectiveness":[51],"conventional":[53],"supervised":[54],"fine-tuning":[55],"(SFT)":[56],"Chain-of-Thought":[58],"(CoT)":[59],"strategies":[60],"that":[61,174,194,203],"work":[62],"well":[63],"general":[65],"domains.":[66],"To":[67,135],"address":[68],"these":[69],"challenges,":[70],"we":[71,139,200],"propose":[72],"Med-R1,":[73,165],"a":[74,114,166],"reinforcement":[75],"learning":[76,98],"(RL)-enhanced":[77],"VLM":[78],"designed":[79],"improve":[81],"generalization":[82,184],"reliability":[84],"reasoning.":[87],"Med-R1":[88,105,112,142,147,206],"adopts":[89],"Group":[90],"Relative":[91],"Policy":[92],"Optimization":[93],"(GRPO)":[94],"encourage":[96],"reward-guided":[97],"beyond":[99],"static":[100],"We":[102,158],"comprehensively":[103],"evaluate":[104,141],"across":[106],"eight":[107],"distinct":[108],"modalities.":[111],"achieves":[113],"29.94%":[115],"improvement":[116],"average":[118],"accuracy":[119],"over":[120],"its":[121],"base":[122],"model":[123,129],"Qwen2-VL-2B,":[124],"even":[126],"outperforms":[127,148],"Qwen2-VL-72B-a":[128],"with":[130,185],"$36\\times":[131],"$":[132],"more":[133,195],"parameters.":[134],"assess":[136],"cross-task":[137],"generalization,":[138,154],"further":[140,159,208],"on":[143,235],"five":[144],"question":[145],"types.":[146],"Qwen2-VL-2B":[149],"by":[150],"32.06%":[151],"question-type":[153],"also":[155,189,201],"surpassing":[156],"Qwen2-VL-72B.":[157],"explore":[160],"thinking":[162],"process":[163],"crucial":[167],"component":[168],"Deepseek-R1.":[170],"Our":[171],"results":[172],"show":[173],"omitting":[175],"intermediate":[176],"rationales":[177],"(No-Thinking":[178],"Med-R1)":[179],"not":[180,228],"only":[181],"improves":[182,209],"cross-domain":[183],"less":[186],"training,":[187],"but":[188],"common":[192],"assumption":[193],"reasoning":[196,226,240,246],"always":[197],"helps.":[198],"Nevertheless,":[199],"find":[202],"Think-After":[205],"variant":[207],"performance":[210,233],"while":[211],"maintaining":[212],"interpretability.":[213],"findings":[215],"suggest":[216],"that,":[217],"VQA,":[220],"mere":[222],"presence":[223],"explicit":[225],"does":[227],"guarantee":[229],"better":[230],"performance.":[231],"Instead,":[232],"depends":[234],"quality":[237],"position":[243],"where":[244],"is":[247],"generated.":[248]},"counts_by_year":[{"year":2026,"cited_by_count":10}],"updated_date":"2026-06-23T13:55:30.953635","created_date":"2026-02-04T00:00:00"}