Explainable deep learning algorithm for distinguishing incomplete Kawasaki disease by coronary artery lesions on echocardiographic imaging.

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.

The Impact of Onset Age on Eosinophils in Kawasaki Disease.

(1) Background: Kawasaki disease (KD) mainly affects children under the age of 5 years and eosinophilia in KD patients might be associated with the development of allergic diseases. We compared the age-adjusted Z-score (Z) of eosinophils and aimed to evaluate the impact of onset age on eosinophils in KD patients. (2) Methods: We divided 398 KD patients into seven age subgroups. Laboratory data and the age-adjusted Z-score of eosinophils during the phases of Kawasaki disease were analyzed. (3) Results: The absolute eosinophil count among all age groups showed significant differences in the post-intravenous immunoglobulin (IVIG) phase and throughout the course of KD with Z-score adjusted for age. Further analysis showed persistent elevation of the age-adjusted Z-score of eosinophils (Z-eosinophil) especially in the under six-month-old age subgroup. In addition, we divided the Z-eosinophil into two groups to find the relationship with coronary artery lesions (CALs). Patients with a higher eosinophil count than average age values had a higher risk of developing CALs, while those with a lower eosinophil count than average age values had a lower risk of having CALs. (4) Conclusions: These findings may provide information to clinicians to pay attention to allergic diseases during the follow-up of KD, especially for children who are younger than 6 months old at the onset of KD, and eosinophil count could be a crucial focus in KD.