{"id":"https://openalex.org/W6991127486","doi":null,"title":"Finding the Trade-offs between Model Size and Performance in Lightweight ResNet-18 architecture for ECG Classification","display_name":"Finding the Trade-offs between Model Size and Performance in Lightweight ResNet-18 architecture for ECG Classification","alternate_titles":[],"publication_year":2025,"publication_date":"2025-06-25","ids":{"openalex":"https://openalex.org/W6991127486"},"language":"en","primary_location":{"id":"pmh:oai:utupub.fi:11111/12005","doi":null,"is_oa":true,"landing_page_url":"https://www.utupub.fi/handle/11111/12005","pdf_url":"https://www.utupub.fi/bitstreams/f0f68a19-e6ee-4f9a-8d02-42e8dd1e35e5/download","source":{"id":"https://openalex.org/S4306402470","display_name":"UTUPub (University of Turku)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I155660961","host_organization_name":"University of Turku","host_organization_lineage":["https://openalex.org/I155660961"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"https://www.utupub.fi/handle/10024/182534","raw_type":"fi=Pro gradu -tutkielma|en=Master&apos;s thesis|"},"parent_works":[],"lineage":["https://openalex.org/W6991127486"],"child_works_count":0,"type":"dissertation","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://www.utupub.fi/bitstreams/f0f68a19-e6ee-4f9a-8d02-42e8dd1e35e5/download","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Ahmed, Faizan","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I155660961","display_name":"University of Turku","ror":"https://ror.org/05vghhr25","country_code":"FI","type":"education","lineage":["https://openalex.org/I155660961"]}],"countries":["FI"],"is_corresponding":true,"contributor_roles":[],"raw_author_name":"Ahmed, Faizan","raw_affiliation_strings":["Health Technology","UNIVERSITY OF TURKU Department of Computing"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Health Technology","institution_ids":[]},{"raw_affiliation_string":"UNIVERSITY OF TURKU Department of Computing","institution_ids":["https://openalex.org/I155660961"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I155660961"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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_pulled":false,"is_paratext":false,"is_xpac":true,"primary_topic":{"id":"https://openalex.org/T11021","display_name":"ECG Monitoring and Analysis","score":0.9990000128746033,"subfield":{"id":"https://openalex.org/subfields/2705","display_name":"Cardiology and Cardiovascular Medicine"},"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/T11021","display_name":"ECG Monitoring and Analysis","score":0.9990000128746033,"subfield":{"id":"https://openalex.org/subfields/2705","display_name":"Cardiology and Cardiovascular Medicine"},"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/T10036","display_name":"Advanced Neural Network Applications","score":0.00019999999494757503,"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/T11396","display_name":"Artificial Intelligence in Healthcare","score":9.999999747378752e-05,"subfield":{"id":"https://openalex.org/subfields/3605","display_name":"Health Information Management"},"field":{"id":"https://openalex.org/fields/36","display_name":"Health Professions"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/ecg-classification","display_name":"ECG classification","score":0.9860000014305115},{"id":"https://openalex.org/keywords/resnet-18","display_name":"ResNet-18","score":0.9020000100135803},{"id":"https://openalex.org/keywords/model-compression","display_name":"model compression","score":0.8090000152587891},{"id":"https://openalex.org/keywords/lightweight-neural-networks","display_name":"lightweight neural networks","score":0.921999990940094},{"id":"https://openalex.org/keywords/depthwise-separable-convolution","display_name":"depthwise separable convolution","score":0.7900000214576721},{"id":"https://openalex.org/keywords/multi-label-classification","display_name":"multi-label classification","score":0.7509999871253967},{"id":"https://openalex.org/keywords/edge-deployment","display_name":"edge deployment","score":0.6010000109672546},{"id":"https://openalex.org/keywords/mobile-health","display_name":"mobile health","score":0.8009999990463257},{"id":"https://openalex.org/keywords/overfitting","display_name":"overfitting","score":0.6269999742507935}],"concepts":[{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.32440000772476196},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.3028999865055084},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.3012999892234802},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.29679998755455017},{"id":"https://openalex.org/C144251240","wikidata":"https://www.wikidata.org/wiki/Q7068229","display_name":"Nucleofection","level":4,"score":0.26170000433921814},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.24959999322891235},{"id":"https://openalex.org/C188198153","wikidata":"https://www.wikidata.org/wiki/Q1613840","display_name":"Limiting","level":2,"score":0.23690000176429749},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.22939999401569366},{"id":"https://openalex.org/C106131492","wikidata":"https://www.wikidata.org/wiki/Q3072260","display_name":"Filter (signal processing)","level":2,"score":0.22210000455379486},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.22110000252723694}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:utupub.fi:11111/12005","doi":null,"is_oa":true,"landing_page_url":"https://www.utupub.fi/handle/11111/12005","pdf_url":"https://www.utupub.fi/bitstreams/f0f68a19-e6ee-4f9a-8d02-42e8dd1e35e5/download","source":{"id":"https://openalex.org/S4306402470","display_name":"UTUPub (University of Turku)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I155660961","host_organization_name":"University of Turku","host_organization_lineage":["https://openalex.org/I155660961"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"https://www.utupub.fi/handle/10024/182534","raw_type":"fi=Pro gradu -tutkielma|en=Master&apos;s thesis|"},{"id":"pmh:oai:www.utupub.fi:10024/182534","doi":null,"is_oa":false,"landing_page_url":"https://www.utupub.fi/handle/10024/182534","pdf_url":null,"source":{"id":"https://openalex.org/S4306402470","display_name":"UTUPub (University of Turku)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I155660961","host_organization_name":"University of Turku","host_organization_lineage":["https://openalex.org/I155660961"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"fi=Pro gradu -tutkielma|en=Master's thesis|"}],"best_oa_location":{"id":"pmh:oai:utupub.fi:11111/12005","doi":null,"is_oa":true,"landing_page_url":"https://www.utupub.fi/handle/11111/12005","pdf_url":"https://www.utupub.fi/bitstreams/f0f68a19-e6ee-4f9a-8d02-42e8dd1e35e5/download","source":{"id":"https://openalex.org/S4306402470","display_name":"UTUPub (University of Turku)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I155660961","host_organization_name":"University of Turku","host_organization_lineage":["https://openalex.org/I155660961"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"https://www.utupub.fi/handle/10024/182534","raw_type":"fi=Pro gradu -tutkielma|en=Master&apos;s thesis|"},"sustainable_development_goals":[{"display_name":"Good health and well-being","score":0.4043000042438507,"id":"https://metadata.un.org/sdg/3"}],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W6991127486.pdf","grobid_xml":"https://content.openalex.org/works/W6991127486.grobid-xml"},"referenced_works_count":33,"referenced_works":["https://openalex.org/W1821462560","https://openalex.org/W1836465849","https://openalex.org/W1845051632","https://openalex.org/W1849277567","https://openalex.org/W2302255633","https://openalex.org/W2433379750","https://openalex.org/W2515385951","https://openalex.org/W2524428287","https://openalex.org/W2531409750","https://openalex.org/W2612445135","https://openalex.org/W2766839578","https://openalex.org/W2785648721","https://openalex.org/W2804262948","https://openalex.org/W2805003733","https://openalex.org/W2884585870","https://openalex.org/W2902644322","https://openalex.org/W2944779197","https://openalex.org/W2949650786","https://openalex.org/W2965520043","https://openalex.org/W3009460750","https://openalex.org/W3140854437","https://openalex.org/W4205381470","https://openalex.org/W4206135956","https://openalex.org/W4236853429","https://openalex.org/W4281647904","https://openalex.org/W4283786338","https://openalex.org/W4309838302","https://openalex.org/W4324057834","https://openalex.org/W4360955643","https://openalex.org/W4392284156","https://openalex.org/W4394670654","https://openalex.org/W4404115913","https://openalex.org/W4407868488"],"related_works":["https://openalex.org/W7134193111","https://openalex.org/W7135199661","https://openalex.org/W4411570448","https://openalex.org/W4200068023","https://openalex.org/W7128601441","https://openalex.org/W7220867085","https://openalex.org/W3190244855","https://openalex.org/W7160448338","https://openalex.org/W7162651777","https://openalex.org/W7139932346"],"abstract_inverted_index":{"Deep":[0],"learning":[1,294],"approaches":[2],"have":[3],"shown":[4],"remarkable":[5],"success":[6],"in":[7,67,109,198,280,303],"electrocardiogram":[8],"(ECG)":[9],"classification,":[10],"but":[11,151],"their":[12,92],"computational":[13,45],"costs":[14],"are":[15,208,242],"often":[16],"a":[17,52,77,106],"bottleneck":[18],"for":[19,56,245],"deployment":[20],"on":[21,266],"low-resource":[22],"platforms":[23],"such":[24,184],"as":[25,125,127,185],"mobile":[26],"health":[27],"initiatives":[28],"and":[29,44,71,91,159,164,173,194,221,274,299],"edge":[30],"devices.":[31],"The":[32,251,282],"thesis":[33],"addresses":[34],"the":[35,40,61,84,101,138,153,169,199,203,230,276,288],"challenge":[36],"by":[37,117],"investigating":[38],"exhaustively":[39],"trade-offs":[41],"between":[42],"performance":[43],"cost":[46],"through":[47],"architectural":[48,89,189],"design":[49],"modifications":[50],"to":[51,82,113,119,124,168,182,263],"ResNet-18":[53],"base":[54],"model":[55,102,115],"multi-label":[57],"ECG":[58,249,260],"classification.":[59,250],"Using":[60],"extensive":[62,216],"combined":[63],"dataset":[64],"of":[65,79,86,100,137,162,218,223,253,278,287,290,301],"PhysioNet/Computing":[66],"Cardiology":[68],"Challenge":[69],"2020":[70],"Shandong":[72],"Provincial":[73],"Hospital,":[74],"we":[75],"conducted":[76],"series":[78],"18":[80],"experiments":[81],"study":[83,255],"impact":[85],"layer":[87],"ablation,":[88],"simplification,":[90],"combinations.":[93],"Our":[94],"findings":[95],"indicate":[96],"that":[97,232],"extreme":[98],"compression":[99,206],"is":[103,180,256],"achievable":[104],"with":[105,156,212,268],"negligible":[107],"reduction":[108,217],"performance.":[110,134],"We":[111],"managed":[112],"compress":[114],"size":[116],"up":[118],"99.2%":[120],"(from":[121],"33.78":[122],"MB":[123],"low":[126],"0.28":[128],"MB)":[129],"without":[130],"sacrificing":[131],"useful":[132],"diagnostic":[133],"Interestingly,":[135],"most":[136],"smaller":[139,200],"models,":[140],"like":[141,248],"Experiment":[142],"E17":[143],"(2.24":[144],"MB,":[145],"93.4%":[146],"compression),":[147],"not":[148],"only":[149],"survived":[150],"outperformed":[152],"baseline":[154],"model,":[155],"micro":[157,160],"AUROC":[158,172],"precision":[161],"0.88":[163],"0.45,":[165],"respectively,":[166],"compared":[167],"baseline\u2019s":[170],"0.86":[171],"0.39":[174],"precision.":[175],"This":[176],"seemingly":[177],"counterintuitive":[178],"effect":[179],"attributed":[181],"factors":[183],"reduced":[186],"overfitting,":[187],"enhanced":[188],"efficiency,":[190],"superior":[191,195],"signal-to-noise":[192],"ratio,":[193],"optimization":[196],"landscape":[197],"models.":[201],"Among":[202],"major":[204],"effective":[205],"techniques":[207],"substituting":[209],"normal":[210],"convolutions":[211],"depthwise":[213],"separable":[214],"convolutions,":[215],"feature":[219],"maps,":[220],"removal":[222],"redundant":[224],"layers":[225],"selectively.":[226],"These":[227],"results":[228],"counteract":[229],"notion":[231],"larger":[233],"models":[234,262],"equal":[235],"better,":[236],"showing":[237],"how":[238],"well-optimized":[239],"small":[240],"architectures":[241],"more":[243],"efficient":[244,292],"some":[246],"tasks":[247],"contribution":[252],"this":[254],"significant,":[257],"enabling":[258,271,275,296],"high-performance":[259],"analysis":[261],"be":[264],"implemented":[265],"hardware":[267],"limited":[269],"resources,":[270],"real-time":[272],"monitoring,":[273],"sustainability":[277],"AI":[279,302],"healthcare.":[281],"research":[283],"offers":[284],"empirical":[285],"proof":[286],"feasibility":[289],"highly":[291],"deep":[293],"solutions,":[295],"broader":[297],"accessibility":[298],"utilization":[300],"clinical":[304],"settings.":[305]},"cited_by_api_url":"https://api.openalex.org/works?filter=cites:W6991127486","counts_by_year":[],"updated_date":"2026-11-16T12:05:36.265796","created_date":"2025-10-10T00:00:00"}
