Raniere Gaia Costa da Silva, Ambika Prasad Mishra, Christopher Michael Riggs, Michael Doube
{"title":"利用深度卷积神经网络对赛马肢体X光片进行分类。","authors":"Raniere Gaia Costa da Silva, Ambika Prasad Mishra, Christopher Michael Riggs, Michael Doube","doi":"10.1002/vro2.55","DOIUrl":null,"url":null,"abstract":"<p><strong>Purpose: </strong>To assess the capability of deep convolutional neural networks to classify anatomical location and projection from a series of 48 standard views of racehorse limbs.</p><p><strong>Materials and methods: </strong>Radiographs (<i>N</i> = 9504) of horse limbs from image sets made for veterinary inspections by 10 independent veterinary clinics were used to train, validate and test (116, 40 and 42 radiographs, respectively) six deep learning architectures available as part of the open source machine learning framework PyTorch. The deep learning architectures with the best top-1 accuracy had the batch size further investigated.</p><p><strong>Results: </strong>Top-1 accuracy of six deep learning architectures ranged from 0.737 to 0.841. Top-1 accuracy of the best deep learning architecture (ResNet-34) ranged from 0.809 to 0.878, depending on batch size. ResNet-34 (batch size = 8) achieved the highest top-1 accuracy (0.878) and the majority (91.8%) of misclassification was due to laterality error. Class activation maps indicated that joint morphology, not side markers or other non-anatomical image regions, drove the model decision.</p><p><strong>Conclusions: </strong>Deep convolutional neural networks can classify equine pre-import radiographs into the 48 standard views including moderate discrimination of laterality, independent of side marker presence.</p>","PeriodicalId":23565,"journal":{"name":"Veterinary Record Open","volume":"10 1","pages":"e55"},"PeriodicalIF":1.3000,"publicationDate":"2023-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9884469/pdf/","citationCount":"0","resultStr":"{\"title\":\"Classification of racehorse limb radiographs using deep convolutional neural networks.\",\"authors\":\"Raniere Gaia Costa da Silva, Ambika Prasad Mishra, Christopher Michael Riggs, Michael Doube\",\"doi\":\"10.1002/vro2.55\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Purpose: </strong>To assess the capability of deep convolutional neural networks to classify anatomical location and projection from a series of 48 standard views of racehorse limbs.</p><p><strong>Materials and methods: </strong>Radiographs (<i>N</i> = 9504) of horse limbs from image sets made for veterinary inspections by 10 independent veterinary clinics were used to train, validate and test (116, 40 and 42 radiographs, respectively) six deep learning architectures available as part of the open source machine learning framework PyTorch. The deep learning architectures with the best top-1 accuracy had the batch size further investigated.</p><p><strong>Results: </strong>Top-1 accuracy of six deep learning architectures ranged from 0.737 to 0.841. Top-1 accuracy of the best deep learning architecture (ResNet-34) ranged from 0.809 to 0.878, depending on batch size. ResNet-34 (batch size = 8) achieved the highest top-1 accuracy (0.878) and the majority (91.8%) of misclassification was due to laterality error. Class activation maps indicated that joint morphology, not side markers or other non-anatomical image regions, drove the model decision.</p><p><strong>Conclusions: </strong>Deep convolutional neural networks can classify equine pre-import radiographs into the 48 standard views including moderate discrimination of laterality, independent of side marker presence.</p>\",\"PeriodicalId\":23565,\"journal\":{\"name\":\"Veterinary Record Open\",\"volume\":\"10 1\",\"pages\":\"e55\"},\"PeriodicalIF\":1.3000,\"publicationDate\":\"2023-01-29\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9884469/pdf/\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Veterinary Record Open\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1002/vro2.55\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"2023/6/1 0:00:00\",\"PubModel\":\"eCollection\",\"JCR\":\"Q2\",\"JCRName\":\"VETERINARY SCIENCES\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Veterinary Record Open","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1002/vro2.55","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2023/6/1 0:00:00","PubModel":"eCollection","JCR":"Q2","JCRName":"VETERINARY SCIENCES","Score":null,"Total":0}
Classification of racehorse limb radiographs using deep convolutional neural networks.
Purpose: To assess the capability of deep convolutional neural networks to classify anatomical location and projection from a series of 48 standard views of racehorse limbs.
Materials and methods: Radiographs (N = 9504) of horse limbs from image sets made for veterinary inspections by 10 independent veterinary clinics were used to train, validate and test (116, 40 and 42 radiographs, respectively) six deep learning architectures available as part of the open source machine learning framework PyTorch. The deep learning architectures with the best top-1 accuracy had the batch size further investigated.
Results: Top-1 accuracy of six deep learning architectures ranged from 0.737 to 0.841. Top-1 accuracy of the best deep learning architecture (ResNet-34) ranged from 0.809 to 0.878, depending on batch size. ResNet-34 (batch size = 8) achieved the highest top-1 accuracy (0.878) and the majority (91.8%) of misclassification was due to laterality error. Class activation maps indicated that joint morphology, not side markers or other non-anatomical image regions, drove the model decision.
Conclusions: Deep convolutional neural networks can classify equine pre-import radiographs into the 48 standard views including moderate discrimination of laterality, independent of side marker presence.
期刊介绍:
Veterinary Record Open is a journal dedicated to publishing specialist veterinary research across a range of topic areas including those of a more niche and specialist nature to that considered in the weekly Vet Record. Research from all disciplines of veterinary interest will be considered. It is an Open Access journal of the British Veterinary Association.