{"id":"https://openalex.org/W7165175477","doi":"https://doi.org/10.48550/arxiv.2606.18797","title":"Beyond Scalar Scores: Exploring LLM-based Metrics for Clinical Significance Evaluation in Radiology Reports","display_name":"Beyond Scalar Scores: Exploring LLM-based Metrics for Clinical Significance Evaluation in Radiology Reports","publication_year":2026,"publication_date":"2026-06-17","ids":{"openalex":"https://openalex.org/W7165175477","doi":"https://doi.org/10.48550/arxiv.2606.18797"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.18797","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.18797","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.18797","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5053159289","display_name":"Qingyu Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Qingyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052597908","display_name":"R K Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Ruochen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138959458","display_name":"Liang Ding","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ding, Liang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5053890256","display_name":"Yufei Xia","orcid":"https://orcid.org/0000-0001-7805-8091"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xia, Yufei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021436608","display_name":"Y C Zhu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Youxiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5138875895","display_name":"Dacheng Tao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tao, Dacheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.49390000104904175,"subfield":{"id":"https://openalex.org/subfields/2718","display_name":"Health Informatics"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}},"topics":[{"id":"https://openalex.org/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.49390000104904175,"subfield":{"id":"https://openalex.org/subfields/2718","display_name":"Health Informatics"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}},{"id":"https://openalex.org/T11894","display_name":"Radiology practices and education","score":0.28450000286102295,"subfield":{"id":"https://openalex.org/subfields/2741","display_name":"Radiology, Nuclear Medicine and Imaging"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}},{"id":"https://openalex.org/T11775","display_name":"COVID-19 diagnosis using AI","score":0.03779999911785126,"subfield":{"id":"https://openalex.org/subfields/2741","display_name":"Radiology, Nuclear Medicine and Imaging"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.6577000021934509},{"id":"https://openalex.org/keywords/clinical-significance","display_name":"Clinical significance","score":0.6025000214576721},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5266000032424927},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.48399999737739563},{"id":"https://openalex.org/keywords/testbed","display_name":"Testbed","score":0.44020000100135803},{"id":"https://openalex.org/keywords/boundary","display_name":"Boundary (topology)","score":0.3885999917984009},{"id":"https://openalex.org/keywords/quality","display_name":"Quality (philosophy)","score":0.37959998846054077},{"id":"https://openalex.org/keywords/performance-metric","display_name":"Performance metric","score":0.3772999942302704}],"concepts":[{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.6577000021934509},{"id":"https://openalex.org/C63363279","wikidata":"https://www.wikidata.org/wiki/Q5133848","display_name":"Clinical significance","level":2,"score":0.6025000214576721},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5552999973297119},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5266000032424927},{"id":"https://openalex.org/C19527891","wikidata":"https://www.wikidata.org/wiki/Q1120908","display_name":"Medical physics","level":1,"score":0.5048999786376953},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.48399999737739563},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4422000050544739},{"id":"https://openalex.org/C31395832","wikidata":"https://www.wikidata.org/wiki/Q1318674","display_name":"Testbed","level":2,"score":0.44020000100135803},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4115000069141388},{"id":"https://openalex.org/C62354387","wikidata":"https://www.wikidata.org/wiki/Q875399","display_name":"Boundary (topology)","level":2,"score":0.3885999917984009},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.37959998846054077},{"id":"https://openalex.org/C2780898871","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Performance metric","level":2,"score":0.3772999942302704},{"id":"https://openalex.org/C31601959","wikidata":"https://www.wikidata.org/wiki/Q931309","display_name":"Medical imaging","level":2,"score":0.37720000743865967},{"id":"https://openalex.org/C158154518","wikidata":"https://www.wikidata.org/wiki/Q7310970","display_name":"Relevance (law)","level":2,"score":0.35929998755455017},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.3447999954223633},{"id":"https://openalex.org/C3020452639","wikidata":"https://www.wikidata.org/wiki/Q454812","display_name":"Clinical judgment","level":2,"score":0.32589998841285706},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.3034000098705292},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3034000098705292},{"id":"https://openalex.org/C126838900","wikidata":"https://www.wikidata.org/wiki/Q77604","display_name":"Radiology","level":1,"score":0.2985000014305115},{"id":"https://openalex.org/C57691317","wikidata":"https://www.wikidata.org/wiki/Q1289248","display_name":"Scalar (mathematics)","level":2,"score":0.27889999747276306},{"id":"https://openalex.org/C106436119","wikidata":"https://www.wikidata.org/wiki/Q836575","display_name":"Quality assurance","level":3,"score":0.27469998598098755},{"id":"https://openalex.org/C2779974597","wikidata":"https://www.wikidata.org/wiki/Q28448986","display_name":"Clinical Practice","level":2,"score":0.2678999900817871},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.25870001316070557},{"id":"https://openalex.org/C56666940","wikidata":"https://www.wikidata.org/wiki/Q788790","display_name":"Documentation","level":2,"score":0.25360000133514404},{"id":"https://openalex.org/C2779473830","wikidata":"https://www.wikidata.org/wiki/Q1540899","display_name":"MEDLINE","level":2,"score":0.251800000667572},{"id":"https://openalex.org/C71924100","wikidata":"https://www.wikidata.org/wiki/Q11190","display_name":"Medicine","level":0,"score":0.250900000333786}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.18797","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.18797","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.18797","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.18797","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"score":0.6429795026779175,"id":"https://metadata.un.org/sdg/10","display_name":"Reduced inequalities"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Reliable":[0],"evaluation":[1],"of":[2],"generated":[3],"radiology":[4],"reports":[5],"requires":[6],"strict":[7],"clinical":[8,73,78,132],"accuracy,":[9],"as":[10,68,169],"omitted":[11],"critical":[12],"findings":[13,164],"or":[14],"mischaracterized":[15],"radiographic":[16],"observations":[17],"can":[18],"directly":[19],"affect":[20],"patient":[21],"care.":[22],"Existing":[23],"metrics":[24,122,168],"obscure":[25],"this":[26,63],"requirement":[27],"by":[28],"reducing":[29],"report":[30,116],"quality":[31],"to":[32,49,152],"a":[33,51,97],"medically":[34],"ungrounded":[35],"scalar.":[36],"Although":[37],"Large":[38],"Language":[39],"Models":[40],"(LLMs)":[41],"possess":[42],"rich":[43],"medical":[44,137],"knowledge,":[45],"they":[46],"likewise":[47],"struggle":[48],"draw":[50],"reliable":[52],"boundary":[53,64],"between":[54],"clinically":[55],"significant":[56],"errors":[57,79,104],"and":[58,70,81,92,118,125,139,157,192],"harmless":[59,108],"variation.":[60],"We":[61,187],"study":[62],"using":[65],"ReEvalMed":[66],"benchmark":[67],"testbed":[69],"evaluate":[71],"metric-level":[72],"significance":[74,133],"from":[75],"detecting":[76],"true":[77],"(\"Discrimination\")":[80],"tolerating":[82],"insignificant":[83],"variations":[84],"(\"Robustness\").":[85],"Across":[86],"8":[87],"LLM":[88],"evaluators":[89],"under":[90],"one-pass":[91,166],"two-pass":[93,149,177],"settings,":[94],"we":[95,113],"identify":[96],"widespread":[98],"discrimination":[99,160],"bias:":[100],"models":[101],"effectively":[102],"detect":[103],"but":[105],"also":[106],"over-penalize":[107],"rephrasings.":[109],"To":[110],"mitigate":[111],"this,":[112],"synthesize":[114],"4k":[115],"pairs":[117],"train":[119],"lightweight":[120],"interpretable":[121],"on":[123],"Qwen3-8B":[124],"MedGemma-4B.":[126],"Our":[127],"trained":[128,167],"metric":[129],"sharpens":[130],"the":[131,146,170,190],"boundary,":[134],"surpassing":[135],"32B-scale":[136],"LLMs":[138],"remaining":[140],"competitive":[141],"with":[142,176],"proprietary":[143],"models.":[144],"Crucially,":[145],"more":[147],"costly":[148],"setting":[150],"fails":[151],"consistently":[153],"improve":[154],"overall":[155],"performance":[156],"mainly":[158],"trades":[159],"for":[161,173,180],"robustness.":[162],"These":[163],"suggest":[165],"practical":[171],"choice":[172],"cost-sensitive":[174],"deployment,":[175],"inference":[178],"reserved":[179],"settings":[181],"where":[182],"D-R":[183],"balance":[184],"is":[185],"critical.":[186],"will":[188],"release":[189],"dataset":[191],"metric.":[193]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-19T00:00:00"}