Cited 3 times in
Feasibility of Deep Learning-Based Analysis of Auscultation for Screening Significant Stenosis of Native Arteriovenous Fistula for Hemodialysis Requiring Angioplasty
DC Field | Value | Language |
---|---|---|
dc.contributor.author | 김경민 | - |
dc.contributor.author | 김만득 | - |
dc.contributor.author | 문성모 | - |
dc.contributor.author | 한기창 | - |
dc.contributor.author | 권준호 | - |
dc.contributor.author | 원종윤 | - |
dc.contributor.author | 김수진 | - |
dc.contributor.author | 심용식 | - |
dc.date.accessioned | 2022-12-22T04:51:43Z | - |
dc.date.available | 2022-12-22T04:51:43Z | - |
dc.date.issued | 2022-10 | - |
dc.identifier.issn | 1229-6929 | - |
dc.identifier.uri | https://ir.ymlib.yonsei.ac.kr/handle/22282913/192240 | - |
dc.description.abstract | Objective: To investigate the feasibility of using a deep learning-based analysis of auscultation data to predict significant stenosis of arteriovenous fistulas (AVF) in patients undergoing hemodialysis requiring percutaneous transluminal angioplasty (PTA). Materials and methods: Forty patients (24 male and 16 female; median age, 62.5 years) with dysfunctional native AVF were prospectively recruited. Digital sounds from the AVF shunt were recorded using a wireless electronic stethoscope before (pre-PTA) and after PTA (post-PTA), and the audio files were subsequently converted to mel spectrograms, which were used to construct various deep convolutional neural network (DCNN) models (DenseNet201, EfficientNetB5, and ResNet50). The performance of these models for diagnosing ≥ 50% AVF stenosis was assessed and compared. The ground truth for the presence of ≥ 50% AVF stenosis was obtained using digital subtraction angiography. Gradient-weighted class activation mapping (Grad-CAM) was used to produce visual explanations for DCNN model decisions. Results: Eighty audio files were obtained from the 40 recruited patients and pooled for the study. Mel spectrograms of "pre-PTA" shunt sounds showed patterns corresponding to abnormal high-pitched bruits with systolic accentuation observed in patients with stenotic AVF. The ResNet50 and EfficientNetB5 models yielded an area under the receiver operating characteristic curve of 0.99 and 0.98, respectively, at optimized epochs for predicting ≥ 50% AVF stenosis. However, Grad-CAM heatmaps revealed that only ResNet50 highlighted areas relevant to AVF stenosis in the mel spectrogram. Conclusion: Mel spectrogram-based DCNN models, particularly ResNet50, successfully predicted the presence of significant AVF stenosis requiring PTA in this feasibility study and may potentially be used in AVF surveillance. | - |
dc.description.statementOfResponsibility | open | - |
dc.format | application/pdf | - |
dc.language | English | - |
dc.publisher | Korean Society of Radiology | - |
dc.relation.isPartOf | KOREAN JOURNAL OF RADIOLOGY | - |
dc.rights | CC BY-NC-ND 2.0 KR | - |
dc.subject.MESH | Angioplasty | - |
dc.subject.MESH | Arteriovenous Fistula* | - |
dc.subject.MESH | Auscultation | - |
dc.subject.MESH | Constriction, Pathologic | - |
dc.subject.MESH | Deep Learning* | - |
dc.subject.MESH | Feasibility Studies | - |
dc.subject.MESH | Female | - |
dc.subject.MESH | Humans | - |
dc.subject.MESH | Male | - |
dc.subject.MESH | Middle Aged | - |
dc.subject.MESH | Renal Dialysis | - |
dc.title | Feasibility of Deep Learning-Based Analysis of Auscultation for Screening Significant Stenosis of Native Arteriovenous Fistula for Hemodialysis Requiring Angioplasty | - |
dc.type | Article | - |
dc.contributor.college | College of Medicine (의과대학) | - |
dc.contributor.department | Dept. of Radiology (영상의학교실) | - |
dc.contributor.googleauthor | Jae Hyon Park | - |
dc.contributor.googleauthor | Insun Park | - |
dc.contributor.googleauthor | Kichang Han | - |
dc.contributor.googleauthor | Jongjin Yoon | - |
dc.contributor.googleauthor | Yongsik Sim | - |
dc.contributor.googleauthor | Soo Jin Kim | - |
dc.contributor.googleauthor | Jong Yun Won | - |
dc.contributor.googleauthor | Shina Lee | - |
dc.contributor.googleauthor | Joon Ho Kwon | - |
dc.contributor.googleauthor | Sungmo Moon | - |
dc.contributor.googleauthor | Gyoung Min Kim | - |
dc.contributor.googleauthor | Man-Deuk Kim | - |
dc.identifier.doi | 10.3348/kjr.2022.0364 | - |
dc.contributor.localId | A00296 | - |
dc.contributor.localId | A00420 | - |
dc.contributor.localId | A06146 | - |
dc.contributor.localId | A05062 | - |
dc.contributor.localId | A05085 | - |
dc.contributor.localId | A02443 | - |
dc.contributor.localId | A00636 | - |
dc.relation.journalcode | J02884 | - |
dc.identifier.eissn | 2005-8330 | - |
dc.identifier.pmid | 36174999 | - |
dc.subject.keyword | Angioplasty | - |
dc.subject.keyword | Arteriovenous fistula | - |
dc.subject.keyword | Auscultation | - |
dc.subject.keyword | Deep learning | - |
dc.subject.keyword | Renal dialysis | - |
dc.contributor.alternativeName | Kim, Gyoung Min | - |
dc.contributor.affiliatedAuthor | 김경민 | - |
dc.contributor.affiliatedAuthor | 김만득 | - |
dc.contributor.affiliatedAuthor | 문성모 | - |
dc.contributor.affiliatedAuthor | 한기창 | - |
dc.contributor.affiliatedAuthor | 권준호 | - |
dc.contributor.affiliatedAuthor | 원종윤 | - |
dc.contributor.affiliatedAuthor | 김수진 | - |
dc.citation.volume | 23 | - |
dc.citation.number | 10 | - |
dc.citation.startPage | 949 | - |
dc.citation.endPage | 958 | - |
dc.identifier.bibliographicCitation | KOREAN JOURNAL OF RADIOLOGY, Vol.23(10) : 949-958, 2022-10 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.