計畫團隊成員

 
圖片1-2

總計畫暨子計畫一
Main Project & Subproject 1

簡韶逸  教授

國立臺灣大學電子所

圖片2-2

子計畫二
Subproject 2

楊家驤  教授

國立臺灣大學電子所

圖片3-2

子計畫三
Subproject 3

劉宗德  助理教授

國立臺灣大學電子所

圖片4-2

子計畫四
Subproject 4

施吉昇  教授

國立臺灣大學網媒所

圖片5-2

子計畫五
Subproject 5

陳炳宇 教授

國立臺灣大學創新研究中心&網媒所

 

技術亮點 Technical Highlights


這項跨領域的計畫將讓頂尖的人工智慧、電腦視覺與電子系統設計技術走進醫學領域,協助外科醫師做手術的影像導航,以提高手術成功度。本計畫的技術特點在於即時分析辨識及動態三維模型貼合的擴增實境外科手術系統。本計畫將大量使用人工智慧技術做手術影像的辨識以及人體和手術器械的定位,以進行外科手術的影像導航,本計畫也將建置高適應性、能和醫學影像結合之動態人體模型,並開發相關的低耗電人工智慧運算及定位加速晶片。

This cross-disciplined project will bring state-of-the-art artificial intelligence (AI), computer vision and electronic system design technologies into the medical field to assist the surgeon with image guided surgery, thereby increasing the efficiency and quality of surgery. The feature of this project is the surgical auxiliary system powered by augmented reality with real-time recognition and dynamic body model. In this image guided surgery system, AI techniques will be developed for surgical image recognition and pose estimation for body/organ and surgical equipments. Adaptive human model will also be developed, which can integrate the data from medical imaging systems, such as CT and MRI. Moreover, we will also develop the associated low-power deep learning accelerator pose estimation hardware engine.

 

應用情境 Applications


本計畫之目的為完成新一代可提昇可視性的智慧輔助手術系統。隨著電腦視覺、人工智慧及擴增實境技術的進步,我們認為有可能可以協助外科醫師做手術的影像導航,將事前拍攝好的三維的醫學影像即時顯示在一頭戴式顯示器中,且和醫師看到的現場影像直接做疊合,並依據人體的動態模型做及時的更新,如此一來醫生可以即時知道手術目標周圍的血管、神經等重要組織的深度和相關位置,以及預先規劃的下刀路徑和切除範圍,對於避免傷害這些重要組織會有很大的助益,也能提高手術成功度以及減少後遺症。這項跨領域的計畫將讓頂尖的人工智慧、電腦視覺與電子系統設計技術走進醫學領域,進而提升手術醫療的效率與品質。

The purpose of this project is to accomplish a new-generation intelligent surgical auxiliary system which provides indirect surgical visualization with artificial intelligence and augmented reality techniques. With advances in artificial intelligence, computer vision and augmented reality technologies, we believe that it is possible to assist the surgeon with image guided surgery. The proposed system will real-time display the 3D medical image in a head-mounted device and overlay the image seen directly by the surgeon. This allows the surgeon to immediately know the depth and relative location of important tissues such as blood vessels and nerves around the target, as well as the planned knife path and the resection range. This can be of great help to avoid harming these important tissues and also improve the success of surgery while reducing sequelae. This cross-disciplined project will bring state-of-the-art artificial intelligence, computer vision and electronic system design technologies into the medical field, thereby increasing the efficiency and quality of surgery.