{"id":"https://openalex.org/W7161776850","doi":"https://doi.org/10.1109/isbi61048.2026.11515370","title":"Fine-Tuning of Diffusion Models Using Perceptual-DPO for Generation of Biologically Accurate Medical Image","display_name":"Fine-Tuning of Diffusion Models Using Perceptual-DPO for Generation of Biologically Accurate Medical Image","publication_year":2026,"publication_date":"2026-04-08","ids":{"openalex":"https://openalex.org/W7161776850","doi":"https://doi.org/10.1109/isbi61048.2026.11515370"},"language":null,"primary_location":{"id":"doi:10.1109/isbi61048.2026.11515370","is_oa":false,"landing_page_url":"https://doi.org/10.1109/isbi61048.2026.11515370","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE 23rd International Symposium on Biomedical Imaging (ISBI)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"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/A5103262871","display_name":"P. Venkata Hari Prasad","orcid":"https://orcid.org/0009-0008-6428-0677"},"institutions":[{"id":"https://openalex.org/I162827531","display_name":"Indian Institute of Technology Bombay","ror":"https://ror.org/02qyf5152","country_code":"IN","type":"education","lineage":["https://openalex.org/I162827531"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Pratyush Prasad","raw_affiliation_strings":["Koita Centre for Digital Health, Indian Institute of Technology Bombay,India"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Koita Centre for Digital Health, Indian Institute of Technology Bombay,India","institution_ids":["https://openalex.org/I162827531"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113361926","display_name":"Swapnil Bhat","orcid":null},"institutions":[{"id":"https://openalex.org/I162827531","display_name":"Indian Institute of Technology Bombay","ror":"https://ror.org/02qyf5152","country_code":"IN","type":"education","lineage":["https://openalex.org/I162827531"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Swapnil Bhat","raw_affiliation_strings":["Koita Centre for Digital Health, Indian Institute of Technology Bombay,India"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Koita Centre for Digital Health, Indian Institute of Technology Bombay,India","institution_ids":["https://openalex.org/I162827531"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136573644","display_name":"Sadaf Siddiqui","orcid":null},"institutions":[{"id":"https://openalex.org/I162827531","display_name":"Indian Institute of Technology Bombay","ror":"https://ror.org/02qyf5152","country_code":"IN","type":"education","lineage":["https://openalex.org/I162827531"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Sadaf Siddiqui","raw_affiliation_strings":["Koita Centre for Digital Health, Indian Institute of Technology Bombay,India"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Koita Centre for Digital Health, Indian Institute of Technology Bombay,India","institution_ids":["https://openalex.org/I162827531"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108671840","display_name":"Siddharth Bhatt","orcid":null},"institutions":[{"id":"https://openalex.org/I4210113533","display_name":"INHS Asvini","ror":"https://ror.org/01hk19p23","country_code":"IN","type":"healthcare","lineage":["https://openalex.org/I4210113533"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Siddharth Bhatt","raw_affiliation_strings":["INHS Asvini,Department of Dermatology,Mumbai,India"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"INHS Asvini,Department of Dermatology,Mumbai,India","institution_ids":["https://openalex.org/I4210113533"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5115489695","display_name":"Mrinal Mallya","orcid":null},"institutions":[{"id":"https://openalex.org/I2800474033","display_name":"P. D. Hinduja Hospital and Medical Research Centre","ror":"https://ror.org/00a6fbp85","country_code":"IN","type":"healthcare","lineage":["https://openalex.org/I2800474033"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Mrinal Mallya","raw_affiliation_strings":["P.D. Hinduja Hospital & Medical Research Center,Mumbai,India"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"P.D. Hinduja Hospital & Medical Research Center,Mumbai,India","institution_ids":["https://openalex.org/I2800474033"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136528203","display_name":"Ankush K Somvanshi","orcid":null},"institutions":[{"id":"https://openalex.org/I4210113533","display_name":"INHS Asvini","ror":"https://ror.org/01hk19p23","country_code":"IN","type":"healthcare","lineage":["https://openalex.org/I4210113533"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Ankush K Somvanshi","raw_affiliation_strings":["INHS Asvini,Department of Dermatology,Mumbai,India"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"INHS Asvini,Department of Dermatology,Mumbai,India","institution_ids":["https://openalex.org/I4210113533"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136564867","display_name":"Deval Mehta","orcid":null},"institutions":[{"id":"https://openalex.org/I56590836","display_name":"Monash University","ror":"https://ror.org/02bfwt286","country_code":"AU","type":"education","lineage":["https://openalex.org/I56590836"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Deval Mehta","raw_affiliation_strings":["Monash University,Faculty of IT,Department of Data Science & AI,Melbourne,Australia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Monash University,Faculty of IT,Department of Data Science & AI,Melbourne,Australia","institution_ids":["https://openalex.org/I56590836"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5136607469","display_name":"Kshitij Jadhav","orcid":null},"institutions":[{"id":"https://openalex.org/I162827531","display_name":"Indian Institute of Technology Bombay","ror":"https://ror.org/02qyf5152","country_code":"IN","type":"education","lineage":["https://openalex.org/I162827531"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Kshitij Jadhav","raw_affiliation_strings":["Koita Centre for Digital Health, Indian Institute of Technology Bombay,India"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Koita Centre for Digital Health, Indian Institute of Technology Bombay,India","institution_ids":["https://openalex.org/I162827531"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.61018049,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.6717000007629395,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.6717000007629395,"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/T11019","display_name":"Image Enhancement Techniques","score":0.08340000361204147,"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/T11105","display_name":"Advanced Image Processing Techniques","score":0.04610000178217888,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/image","display_name":"Image (mathematics)","score":0.41339999437332153},{"id":"https://openalex.org/keywords/diffusion","display_name":"Diffusion","score":0.40639999508857727},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.365200012922287},{"id":"https://openalex.org/keywords/image-processing","display_name":"Image processing","score":0.3564999997615814},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.323199987411499}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5103999972343445},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5026000142097473},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.48240000009536743},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.41339999437332153},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.40639999508857727},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.365200012922287},{"id":"https://openalex.org/C9417928","wikidata":"https://www.wikidata.org/wiki/Q1070689","display_name":"Image processing","level":3,"score":0.3564999997615814},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3361000120639801},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.323199987411499},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.29910001158714294},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2597000002861023},{"id":"https://openalex.org/C124504099","wikidata":"https://www.wikidata.org/wiki/Q56933","display_name":"Image segmentation","level":3,"score":0.2556000053882599},{"id":"https://openalex.org/C31601959","wikidata":"https://www.wikidata.org/wiki/Q931309","display_name":"Medical imaging","level":2,"score":0.2551000118255615}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/isbi61048.2026.11515370","is_oa":false,"landing_page_url":"https://doi.org/10.1109/isbi61048.2026.11515370","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE 23rd International Symposium on Biomedical Imaging (ISBI)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":11,"referenced_works":["https://openalex.org/W2133665775","https://openalex.org/W2794825826","https://openalex.org/W3194282450","https://openalex.org/W4283782831","https://openalex.org/W4387757480","https://openalex.org/W4404901776","https://openalex.org/W4410216381","https://openalex.org/W4410297066","https://openalex.org/W4413003229","https://openalex.org/W7133185848","https://openalex.org/W7133208539"],"related_works":[],"abstract_inverted_index":{"Denoising":[0],"Diffusion":[1],"Probabilistic":[2],"Models":[3],"(DDPMs)":[4],"have":[5],"recently":[6],"gained":[7],"prominence":[8],"in":[9,25,48],"the":[10,69,114,173],"research":[11],"community":[12],"for":[13,108,160,167],"generating":[14],"synthetic":[15],"samples":[16],"of":[17,65,113,175],"rare":[18],"disease":[19],"types":[20],"to":[21,39,158,165],"mitigate":[22],"class":[23],"imbalance":[24],"medical":[26,34,62],"datasets.":[27],"While":[28],"DDPMs":[29],"can":[30],"produce":[31],"visually":[32],"convincing":[33],"images,":[35],"they":[36],"often":[37],"fail":[38],"preserve":[40],"accurate":[41],"anatomical":[42],"structure":[43],"and":[44,83,101,110,139,163],"clinical":[45,140],"fidelity,":[46],"resulting":[47],"limited":[49],"biological":[50,152],"accuracy.":[51],"To":[52,116],"overcome":[53],"these":[54,118],"limitations,":[55,119],"there":[56,96],"are":[57],"fine-tuning":[58],"frameworks":[59],"that":[60,135],"integrate":[61],"experts'":[63],"evaluations":[64],"generated":[66],"images":[67],"into":[68],"DDPM":[70],"training":[71],"process":[72],"using":[73],"human-in-theloop":[74],"reinforcement":[75,178],"learning":[76],"techniques;":[77],"specifically,":[78],"Direct":[79],"Preference":[80],"Optimization":[81],"(DPO)":[82],"Reinforcement":[84],"Learning":[85],"with":[86,131],"Human":[87],"Feedback":[88],"(RLHF).":[89],"Even":[90],"though,":[91],"RLHF":[92],"is":[93,97,182],"widely":[94],"adopted,":[95],"substantial":[98],"computational":[99],"overhead":[100],"classical":[102],"DPO":[103],"does":[104],"not":[105],"explicitly":[106],"account":[107],"structural":[109,138],"perceptual":[111,125,132],"aspects":[112],"images.":[115],"address":[117],"we":[120],"propose":[121],"Perceptual-DPO":[122,143],"which":[123],"introduces":[124],"preference":[126],"optimization,":[127],"replacing":[128],"log-likelihood":[129],"terms":[130],"similarity":[133],"scores":[134,154],"better":[136],"capture":[137],"fidelity.":[141],"Our":[142,180],"framework":[144],"significantly":[145],"improved":[146],"generation":[147],"performance,":[148],"increasing":[149],"expert":[150],"verified":[151],"accuracy":[153],"from":[155],"18.54":[156],"%":[157],"30.71%":[159],"HAM10000":[161],"datasets":[162],"23.54%":[164],"58.52%":[166],"Bone":[168],"Marrow":[169],"Cytomorphology":[170],"dataset":[171],"demonstrating":[172],"effectiveness":[174],"incorporating":[176],"expert-guided":[177],"learning.":[179],"code":[181],"available":[183],"at":[184],"https://github.com/aidelab-iitbombay/Perceptual-DPO.":[185]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-05-21T00:00:00"}