1
|
Lee H, Eun Y, Hwang JY, Eun LY. Explainable deep learning algorithm for distinguishing incomplete Kawasaki disease by coronary artery lesions on echocardiographic imaging. COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE 2022; 223:106970. [PMID: 35772231 PMCID: PMC9214709 DOI: 10.1016/j.cmpb.2022.106970] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 08/28/2021] [Revised: 04/30/2022] [Accepted: 06/19/2022] [Indexed: 06/15/2023]
Abstract
BACKGROUND AND OBJECTIVE Incomplete Kawasaki disease (KD) has often been misdiagnosed due to a lack of the clinical manifestations of classic KD. However, it is associated with a markedly higher prevalence of coronary artery lesions. Identifying coronary artery lesions by echocardiography is important for the timely diagnosis of and favorable outcomes in KD. Moreover, similar to KD, coronavirus disease 2019, currently causing a worldwide pandemic, also manifests with fever; therefore, it is crucial at this moment that KD should be distinguished clearly among the febrile diseases in children. In this study, we aimed to validate a deep learning algorithm for classification of KD and other acute febrile diseases. METHODS We obtained coronary artery images by echocardiography of children (n = 138 for KD; n = 65 for pneumonia). We trained six deep learning networks (VGG19, Xception, ResNet50, ResNext50, SE-ResNet50, and SE-ResNext50) using the collected data. RESULTS SE-ResNext50 showed the best performance in terms of accuracy, specificity, and precision in the classification. SE-ResNext50 offered a precision of 81.12%, a sensitivity of 84.06%, and a specificity of 58.46%. CONCLUSIONS The results of our study suggested that deep learning algorithms have similar performance to an experienced cardiologist in detecting coronary artery lesions to facilitate the diagnosis of KD.
Collapse
Affiliation(s)
- Haeyun Lee
- Department of Electrical Engineering and Computer Science
| | - Yongsoon Eun
- Department of Electrical Engineering and Computer Science; The Interdisciplinary Studies of Artificial Intelligence
| | - Jae Youn Hwang
- Department of Electrical Engineering and Computer Science; The Interdisciplinary Studies of Artificial Intelligence.
| | - Lucy Youngmin Eun
- Division of Pediatric Cardiology, Department of Pediatrics, Yonsei University College of Medicine, Seoul 03722, South Korea.
| |
Collapse
|
2
|
Guo C, Tan C, Xia X, Yuan Y, Zhao M, Yuan Z, Wang Y, Deng Z, Chen J, Zhou Y, Huang Y. Tumour necrosis factor-α and myoglobin associated with the recovery time of coronary artery lesions in Kawasaki disease patients. J Paediatr Child Health 2020; 56:1382-1387. [PMID: 32479688 DOI: 10.1111/jpc.14942] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.2] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 12/14/2018] [Revised: 01/20/2020] [Accepted: 04/12/2020] [Indexed: 12/12/2022]
Abstract
AIM To assess the relationship between clinical parameters and medium term recovery time of coronary artery lesions (CALs). METHODS In total, 344 Kawasaki disease patients were screened and 311 Kawasaki disease patients were included and followed-up for the next 2 years. Clinical records, clinical parameters and inflammatory biomarkers were collected for all subjects. RESULTS Tumour necrosis factor (TNF)-α and myoglobin (MYO) levels in patients without recovery from CALs were significantly higher than those without CALs and with recovery from CALs. Kaplan-Meier survival analysis showed that in the high-TNF-α group, the estimated median time to recovery (5.0 months, 95% confidence interval (CI) 1.436-8.564) is significantly longer than the low-TNF-α group (2.00 months, 95% CI: 0.633-3.367, P = 0.044). Also, the estimated median time (5.0 months, 95% CI: 1.836-8.164) in the high-MYO group is significantly longer than the low-MYO group (2.00 months, 95% CI: 0.405-3.595, P = 0.002). Cox regression analysis showed independent factors for recovery of CALs included age, left coronary artery to aortic annulus ratio, TNF-α and MYO levels. CONCLUSIONS These findings suggest that clinical parameters such as age, left coronary artery to aortic annulus ratio, TNF-α and MYO levels associate with medium term recovery time of CALs and could help in the design of a clinical strategy for the surveillance and prevention of late cardiovascular events.
Collapse
Affiliation(s)
- Chun Guo
- Children's Medical Center, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| | - Chaochao Tan
- Department of Clinical Laboratory, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| | - Xiaohui Xia
- Ultrasound Department, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| | - Yonghua Yuan
- Children's Medical Center, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| | - Menghua Zhao
- Children's Medical Center, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| | - Zhijie Yuan
- Children's Medical Center, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| | - Yupeng Wang
- Department of Clinical Laboratory, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| | - Zhonghua Deng
- Department of Clinical Laboratory, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| | - Jie Chen
- Department of Clinical Laboratory, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| | - Yujie Zhou
- Department of Clinical Laboratory, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| | - Ying Huang
- Department of Emergency, Hunan Provincial People's Hospital, the First Affiliated Hospital, Hunan Normal University, Changsha, China
| |
Collapse
|
3
|
Lazea C, Man O, Sur LM, Serban R, Lazar C. Unusual Presentation Of Kawasaki Disease With Gastrointestinal And Renal Manifestations. Ther Clin Risk Manag 2019; 15:1411-1416. [PMID: 31824164 PMCID: PMC6901056 DOI: 10.2147/tcrm.s226624] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/12/2019] [Accepted: 10/17/2019] [Indexed: 12/18/2022] Open
Abstract
Diagnosis of Kawasaki disease (KD) is based on well-established clinical criteria. In incomplete or atypical KD, the diagnosis is challenging, because of the paucity of clinical signs or because of the presence of clinical manifestations that generally are not seen in KD. We describe the case of a 3-year-old female patient with persistent high fever, vomiting, watery diarrhea, metabolic acidosis and severe hypopotassemia. On the fourth day of fever, bilateral conjunctivitis, mucous and extremity changes were registered. Urine changes as glycosuria and proteinuria were also noticed. Echocardiography revealed ectasia of the left anterior descending coronary artery, and diagnosis of KD was established. The treatment consisted of intravenous immunoglobulin (IVIG) and oral aspirin. Recurrence of disease was recorded on the 23rd day of the disease, with favorable evolution after the second dose of IVIG was infused.
Collapse
Affiliation(s)
- Cecilia Lazea
- University of Medicine and Pharmacy, Emergency Clinic Hospital for Children, Department Pediatrics I, Cluj-Napoca, Romania
| | - Oana Man
- Emergency Clinic Hospital for Children, Department Pediatrics I, Cluj-Napoca, Romania
| | - Lucia Maria Sur
- University of Medicine and Pharmacy, Emergency Clinic Hospital for Children, Department Pediatrics I, Cluj-Napoca, Romania
| | - Radu Serban
- University of Medicine and Pharmacy, Emergency Clinic Hospital for Children, Department Pediatrics I, Cluj-Napoca, Romania
| | - Calin Lazar
- University of Medicine and Pharmacy, Emergency Clinic Hospital for Children, Department Pediatrics I, Cluj-Napoca, Romania
| |
Collapse
|