Enimmune Corporation yesterday (25) announced that its AI-powered Kawasaki disease risk assessment software, PEKAN, has successfully completed clinical performance studies at two major medical centers in Taiwan. Results showed that PEKAN demonstrated stable and highly accurate clinical differentiation, with an accuracy rate reaching 90%, laying a solid foundation for AI-based medical device applications in pediatrics. The company is now preparing for regulatory approval under Taiwan’s Medical Device Act as a Class II medical device, with the goal of securing TFDA registration and launching by the end of 2026. Enimmune also plans to accelerate international expansion, initially targeting Asian markets through clinical collaborations and licensing partnerships.
The clinical study was a multicenter trial involving febrile pediatric patients from 2001 to 2024. By inputting routine blood test data and patient age, PEKAN utilized machine learning algorithms to generate risk assessments, which were then compared with physicians’ actual diagnoses. The results confirmed an accuracy of 90.9%, demonstrating both the robustness of the model and its broad clinical applicability. Professor Huang Ying-Hsien, principal investigator of the study, noted that PEKAN achieved high accuracy across large-scale clinical datasets, meeting clinical demands and significantly improving early identification of Kawasaki disease, thereby reducing the risks associated with delayed diagnosis.
First reported in 1961 by Japanese pediatrician Tomisaku Kawasaki, Kawasaki disease (KD) remains one of the most common acquired heart diseases among children in Asia. Prevalence is particularly high in the region, with Taiwan ranking third worldwide after Japan and South Korea. Statistics indicate that approximately 300,000 children across Asia are affected annually, while in Taiwan, 50,000 to 100,000 febrile pediatric patients undergo evaluation each year for potential KD. Without timely intravenous immunoglobulin (IVIG) treatment during the critical therapeutic window, 15%–25% of children risk developing coronary artery aneurysms or other cardiovascular complications. By integrating AI analysis with routine blood test data, PEKAN provides real-time risk assessment to support clinicians in emergency rooms, pediatric clinics, and resource-limited areas, enabling earlier and more accurate decisions and reducing the likelihood of coronary complications.
Professor Eric Kuo (Kuo, He-Chang), a Kawasaki disease expert, emphasized that diagnosing KD has long been a major challenge in emergency and pediatric care, particularly in its early stages when symptoms overlap with other common childhood illnesses, often leading to misdiagnosis or delayed treatment. Access to AI-driven risk evaluation and diagnostic support at the first point of care would not only improve diagnostic accuracy but also substantially lower the risk of severe complications such as coronary aneurysms. He noted that PEKAN’s successful clinical validation marks not just a technological innovation but also a significant advancement in pediatric care quality and children’s health protection, highlighting Enimmune’s forward-looking strategy in the AI medical device field.
Enimmune stressed its commitment to regulatory progress, with plans to file for TFDA approval as a Class II medical device and launch PEKAN by late 2026. The company is also fast-tracking its international strategy, focusing on Asian markets through collaborations and licensing. Beyond Kawasaki disease, Enimmune intends to expand its AI diagnostic technology to other pediatric and infectious diseases, gradually building a diversified smart healthcare product portfolio. CEO Chang, Che-Wei stated that PEKAN is not just a single-disease AI diagnostic tool but also the starting point of the company’s broader intelligent healthcare platform. By harnessing data-driven insights and AI algorithms, Enimmune aims to deliver long-term value to clinicians, pediatric patients and families, as well as investors.
Resource: 安特羅川崎氏症輔助診斷軟體 拚明年底前上市
