{"id":"https://openalex.org/W7165032298","doi":"https://doi.org/10.48550/arxiv.2606.17213","title":"Revisiting LLM Adaptation for 3D CT Report Generation: A Study of Scaling and Diagnostic Priors","display_name":"Revisiting LLM Adaptation for 3D CT Report Generation: A Study of Scaling and Diagnostic Priors","publication_year":2026,"publication_date":"2026-06-15","ids":{"openalex":"https://openalex.org/W7165032298","doi":"https://doi.org/10.48550/arxiv.2606.17213"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.17213","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.17213","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"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.2606.17213","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5043592090","display_name":"Vanshali Sharma","orcid":"https://orcid.org/0000-0003-0008-1579"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sharma, Vanshali","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090774711","display_name":"Andrea Bejar","orcid":"https://orcid.org/0000-0003-2089-5983"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bejar, Andrea M.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114601749","display_name":"Halil Ertugrul Aktas","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Aktas, Halil Ertugrul","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110698580","display_name":"Q.D. Trinh","orcid":"https://orcid.org/0000-0001-5835-861X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Trinh, Quoc-Huy","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5044673103","display_name":"Debesh Jha","orcid":"https://orcid.org/0000-0002-8078-6730"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jha, Debesh","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033826718","display_name":"G\u00f6rkem Durak","orcid":"https://orcid.org/0000-0002-1608-1955"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Durak, Gorkem","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5138797185","display_name":"Ulas Bagci","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bagci, Ulas","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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.659500002861023,"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":0.659500002861023,"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/T11775","display_name":"COVID-19 diagnosis using AI","score":0.094200000166893,"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/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.05220000073313713,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/overfitting","display_name":"Overfitting","score":0.8842999935150146},{"id":"https://openalex.org/keywords/bridging","display_name":"Bridging (networking)","score":0.599399983882904},{"id":"https://openalex.org/keywords/prior-probability","display_name":"Prior probability","score":0.5587000250816345},{"id":"https://openalex.org/keywords/adaptability","display_name":"Adaptability","score":0.5446000099182129},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.5309000015258789},{"id":"https://openalex.org/keywords/phenomics","display_name":"Phenomics","score":0.5023000240325928},{"id":"https://openalex.org/keywords/domain-adaptation","display_name":"Domain adaptation","score":0.4318000078201294},{"id":"https://openalex.org/keywords/medical-imaging","display_name":"Medical imaging","score":0.3874000012874603},{"id":"https://openalex.org/keywords/fluency","display_name":"Fluency","score":0.373199999332428}],"concepts":[{"id":"https://openalex.org/C22019652","wikidata":"https://www.wikidata.org/wiki/Q331309","display_name":"Overfitting","level":3,"score":0.8842999935150146},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6600000262260437},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6449999809265137},{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.599399983882904},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5735999941825867},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.5587000250816345},{"id":"https://openalex.org/C177606310","wikidata":"https://www.wikidata.org/wiki/Q5674297","display_name":"Adaptability","level":2,"score":0.5446000099182129},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.5309000015258789},{"id":"https://openalex.org/C98108635","wikidata":"https://www.wikidata.org/wiki/Q6497275","display_name":"Phenomics","level":5,"score":0.5023000240325928},{"id":"https://openalex.org/C2776434776","wikidata":"https://www.wikidata.org/wiki/Q19246213","display_name":"Domain adaptation","level":3,"score":0.4318000078201294},{"id":"https://openalex.org/C31601959","wikidata":"https://www.wikidata.org/wiki/Q931309","display_name":"Medical imaging","level":2,"score":0.3874000012874603},{"id":"https://openalex.org/C2777413886","wikidata":"https://www.wikidata.org/wiki/Q3276013","display_name":"Fluency","level":2,"score":0.373199999332428},{"id":"https://openalex.org/C57493831","wikidata":"https://www.wikidata.org/wiki/Q3134666","display_name":"Projection (relational algebra)","level":2,"score":0.3675999939441681},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.35350000858306885},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.3499999940395355},{"id":"https://openalex.org/C2983449737","wikidata":"https://www.wikidata.org/wiki/Q177719","display_name":"Clinical diagnosis","level":2,"score":0.34850001335144043},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.34439998865127563},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.33399999141693115},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.31290000677108765},{"id":"https://openalex.org/C89600930","wikidata":"https://www.wikidata.org/wiki/Q1423946","display_name":"Segmentation","level":2,"score":0.3118000030517578},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.28690001368522644},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.273499995470047},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.2689000070095062},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.26109999418258667},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.2529999911785126}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.17213","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.17213","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.17213","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.17213","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Recent":[0],"advances":[1],"in":[2],"multimodal":[3],"learning,":[4],"including":[5],"large":[6],"language":[7],"models":[8,12],"(LLMs)":[9],"and":[10,43,50,64,88,134,170,183,190,200],"vision-language":[11],"(VLMs),":[13],"have":[14],"demonstrated":[15],"strong":[16,202],"adaptability":[17],"to":[18,25,37,62,152],"natural":[19],"images.":[20],"However,":[21],"extending":[22],"their":[23],"use":[24],"the":[26,44,98,120,125,136],"medical":[27,58],"domain,":[28],"particularly":[29],"for":[30,83,100,162],"volumetric":[31,41,84],"(3D)":[32],"images,":[33],"is":[34,70,159],"challenging":[35],"due":[36],"high":[38],"computational":[39,184],"complexity,":[40],"dependencies":[42],"semantic":[45,121],"gap":[46],"between":[47,180],"visual":[48],"features":[49],"clinical":[51,65,73,116,192],"terminology.":[52],"Naively":[53],"fine-tuning":[54,158],"LLMs":[55,149,169],"on":[56,140],"limited":[57],"data":[59],"often":[60],"leads":[61],"overfitting":[63,139],"hallucination,":[66],"where":[67],"linguistic":[68],"fluency":[69],"prioritized":[71],"over":[72],"factuality.":[74],"In":[75],"this":[76],"study,":[77,194],"we":[78,155],"investigate":[79],"parameter-efficient":[80,198],"adaptation":[81],"strategies":[82],"CT":[85],"report":[86],"generation":[87],"introduce":[89],"RAD3D-Prefix,":[90],"a":[91,145,177,191],"lightweight":[92,173],"diagnostic-prior":[93],"conditioning":[94],"framework":[95],"that":[96,157],"minimizes":[97],"need":[99],"extensive":[101],"parameter":[102],"training.":[103],"This":[104],"module":[105],"integrates":[106],"image":[107],"embeddings":[108],"with":[109],"multi-label":[110],"diagnostic":[111],"classification":[112],"logits,":[113],"preserving":[114],"critical":[115],"details":[117],"while":[118,205],"bridging":[119],"gap.":[122],"By":[123],"keeping":[124],"LLM":[126],"frozen,":[127],"our":[128],"method":[129],"requires":[130],"minimal":[131],"trainable":[132,209],"parameters":[133,210],"mitigates":[135],"risk":[137],"of":[138],"small,":[141],"domain-specific":[142],"datasets.":[143],"Through":[144],"systematic":[146],"study":[147],"spanning":[148],"from":[150],"96.1M":[151],"1.6B":[153],"parameters,":[154],"find":[156],"most":[160],"beneficial":[161],"smaller":[163],"LLMs,":[164],"whereas":[165],"freezing":[166],"larger":[167],"(~1B+":[168],"training":[171],"only":[172],"projection":[174],"layers":[175],"provides":[176],"superior":[178],"trade-off":[179],"performance,":[181],"generalization,":[182],"efficiency.":[185],"Across":[186],"multiple":[187],"automatic":[188],"metrics":[189],"reader":[193],"RAD3D-Prefix":[195],"outperforms":[196],"comparable":[197],"baselines":[199],"demonstrates":[201],"out-of-domain":[203],"generalization":[204],"using":[206],"substantially":[207],"fewer":[208],"than":[211],"fully":[212],"fine-tuned":[213],"alternatives.":[214]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-06-18T00:00:00"}