Comparison of accuracy of five DL models in old test data, new data, and combined results.
figure
posted on 2023-02-02, 18:47 authored by Jinhyung Rho, Sung-Min Shin, Kyoungsun Jhang, Gwanghee Lee, Keun-Ho Song, Hyunguk Shin, Kiwon Na, Hyo-Jung Kwon, Hwa-Young SonFive DL architectures were used and compared (ResNet50V2, ResNet152, InceptionResNetV2, MobileNetV2, and Xception).
History
Usage metrics
Categories
Keywords
assist veterinary radiologistsamong various methods5 independent institutionslearning system usingfive cnn models22 radiographic images20 radiographic imagesdiagnosing feline hcmtwo voting strategies45 %; resnet15245 %; mobilenetv2combined test dataray imagesmodels showedfeline hcmdeep learning42 imagesvoting strategymajority votingtest datasudden deathresidual architectureprediction analysispeeking phenomenonoptimal architectureheart murmurgold standardsfindings demonstrateautomated deep88 normal143 hcm
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC