『您的瀏覽器不支援JavaScript功能,若網頁功能無法正常使用時,請開啟瀏覽器JavaScript狀態』

跳到主要內容區塊

工業技術研究院

:::

技術名稱: 視力檢測暨眼底影像分析技術 V2

技術簡介

本技術涵蓋一視力健檢創新服務雲端平台(iVision),整合影像品質分析技術、眼底影像縫合、網頁版遠距閱片平台、個人保健護照雲端資料管理系統資料倉儲、即時健檢報告檢視與產出模組與標準化資料擷取單元開發,達到全電子化醫囑操作。增加了可攜式檢測設備整合匝道器、離線資料備援方法,提升服務的可及性與穩定性。在醫療影像輔助判讀技術上,iVision提供包含彩色眼底影像用於糖尿病視網膜病變嚴重程度、青光眼視野缺損與視神經纖維缺損之早期偵測,以及血管阻塞/硬化風險評估等診斷輔助決策支援。本技術整合國內醫療器材(明達、廣和)、健檢中心與眼科診所,將協助國內眼底鏡製造商設備與產品升級;對服務提供者-健檢診所而言,可有效提升醫師執行效率至少20%(檢測時間6min→5min/人)

Abstract

This technology covers an innovative service cloud platform for vision health examination- iVision. iVision was integrated fundus image quality analysis technology, fundus image stitching, web-based remote screening platform, cloud data management system of personal health passport, real-time health examination report review and generation and the standardized data acquisition units module. It achieves fully electronic medical order operations. Furthermore, there is a portable smart gateway for offline data backup to improve service accessibility and stability. In medical imaging-assisted diagnosis, iVision provides diagnosis-assisted decision support including color fundus images for early detection of diabetic retinopathy severity, glaucoma visual field defects and retinal nerve fiber layer (RNFL) defects, and risk assessment of vascular occlusion/sclerosis. Up to now, This technology has been integrated with more than 20 sets of domestic medical equipment (Crystalvue & Kongho) and trial run in the health check centers and ophthalmology clinics. This Service can effectively improve the efficiency of the eye examination. (examination time 6 min-> 5 min / person)

技術規格

依其平台技術功能逐項展開: 眼底影像品質分析模組: 採用C語言與OpenCV函數庫於Windows作業系統開發 眼底影像縫合: 採用C語言與OpenCV函數庫於Windows作業系統開發;影像縫合誤差率<0.3pixels、縫合時間為1.8秒 視力健檢創新服務雲端平台(iVision)服務部署支援規格: 後台: IIS(Internet Information Server)、Java Servelet、DB: Microsoft SQL Server 2008 R2;前台: iOS, Windows Application 醫療影像輔助判讀技術: 糖尿病視網膜眼底病變嚴重程度偵測: 準確度92.06%、敏感度: 87.05%、特異度: 92.37%;五份類Kappa Score: 84.15% 青光眼視野缺損之早期偵測: 準確度93.84%、敏感度: 92.00%、特異度: 95.00% 青光眼視神經纖維缺損之早期偵測: 準確度94.19%、敏感度: 93.92%、特異度: 94.57%

Technical Specification

Expand item according to its platform technical functions: Fundus image quality analysis module: Using C language and OpenCV library to develop on Windows operating system Fundus image stitching: Using C language and OpenCV function library to develop on Windows operating system; image stitching error rate <0.3pixels, stitching time 1.8 seconds Innovative service cloud platform for vision health examination- iVision, Service Deployment Support Specifications: Back-end: IIS (Internet Information Server), Java Server, DB: Microsoft SQL Server 2008 R2; Front-end: iOS, Windows Application Medical imaging assisted diagnosis technology Detection of the severity of diabetic retinopathy: accuracy 92.06%, sensitivity: 87.05%, specificity: 92.37%; Kappa Score: 84.15% Early detection of glaucoma visual field defects: accuracy: 93.84%, sensitivity: 92.00%, specificity: 95.00% Early detection of RNFL defects in glaucoma: accuracy: 94.19%, sensitivity: 93.92%, specificity: 94.57%

技術特色

視力健檢創新服務雲端平台(iVision),資料串接整合超過5個桌上型視力檢查設備廠商15種檢測設備,以及3家手持式檢測設備,並提供多種資料備援方案,減少檢查資料遺失可能。

應用範圍

醫療診所/健檢中心 智慧醫療系統商

接受技術者具備基礎建議(設備)

Windows OS PC, Digital Ophthalmology Scope

接受技術者具備基礎建議(專業)

數位眼底鏡 醫療資訊系統(HIS,CPOE) 嵌入式系統開發

技術分類 H-智慧醫療

聯絡資訊

聯絡人:陳建任 智慧醫療與照護服務組

電話:+886-3-5913416 或 Email:cjr@itri.org.tw

客服專線:+886-800-45-8899

傳真:+886-3-5910212

舊工業技術研究院圖示