Institute of Information Systems and Applications
Speaker:	Dr. Joe Yeh（葉肇元）, 創辦人,  aetherAI 雲象科技
Topic:	運用人工智慧提升醫療影像診斷品質
Date:	13:20-15:00 Wednesday 26-Apr-2023
QR Code:
Link:	https://meet.google.com/xcx-exzx-nod
Location:	Delta 103
Hosted by:	Prof. Chun-Yi Lee

Abstract

自從2015年深層神經網絡在圖像識別任務上勝過人類以來，人們一直熱衷於尋找這種強大技術可以產生影響的應用領域。由於易於獲得的數據及其在診斷和治療計劃過程中的關鍵作用，醫學影像吸引了很多關注。值得注意的是，深層卷積神經網絡（CNN）在診斷糖尿病性視網膜病變，分類皮膚疾病和識別淋巴結轉移性癌細胞方面優於人類專家。令人驚訝的是，CNN也顯示出能夠使用彩色眼底圖像預測心臟危險因素，並能夠使用組織病理學圖像區分基因型。 CNN的這些超人能力為醫生提供了許多無法預料的機會，以使他們能夠制定更好的診斷和治療計劃。在本次演講中，我將使用aetherAI的真實案例來說明人工智慧如何將更高的標準引入醫學影像分析，將傳統的定性實踐轉變為定量實踐並最終對臨床決策過程產生影響的潛力。

Dr. Joe Yeh received his M.D degree from National Taiwan University. After medical school, he spent a year working on reconstructing the wiring diagram of fruit fly brain in the Brain Center of National Tsin-Hua University. He later went to University of Southern California for Ph.D. program in pathology, where he discovered his passion for computer programming and image analysis when he learned to apply advanced light microscopy techniques to investigate the cellular dynamics of embryonic development. Inspired by the power of computer algorithms to perform fast and accurate image analysis and its potential to improve healthcare, he left academia and founded aetherAI. aetherAI is now a leading digital pathology AI company in Taiwan, working with top medical centers to develop AI algorithms to help pathologists to achieve faster and more accurate diagnosis. He also lectured on application of deep learning in medical image analysis at National Taiwan University.

Bio.

葉肇元醫師於台灣大學醫學系畢業後，花了一年的時間在台灣清華大學的腦科學中心進行果蠅腦神經連結圖譜重組之研究。之後他前往美國南加州大學攻讀病理學博士。期間除了學習生醫研究的技術之外，也學習建構雷射掃描顯微鏡，並撰寫電腦程式以控制顯微鏡並進行生醫影像分析。因為預見了影像數位化對於臨床病理診斷可能帶來的衝擊，他創立了雲象科技以提供數位病理玻片的雲端解決方案。他對新事物充滿好奇心，也極度熱愛自學。在就讀研究所時，他透過自學開始寫程式解決研究上的問題。在他積極企圖讓公司轉型成為醫療影像人工智慧公司時，更花了無數的時間研究深度神經網路，成功的在公司裡建立了人工智慧技術的基礎。雲象科技是亞洲數位病理AI產業的領導廠商，與美國及台灣許多 醫學中心有合作計畫。葉醫師亦曾以業界教師的身分於成功大學教授深度學習。他對於醫療最遠大的願景，是一個醫師和人工智慧和諧互助，使醫療照護品質不斷提升的未來。

After receiving his medical degree from National Taiwan University, Joe went to University of Southern California for PhD program in pathology. He has insatiable curiosity and is an avid self-learner. During graduate school years, he mastered tissue experiment and optical microscopy. He was also a self-taught computer programmer. He eventually left academia and founded aetherAI, where He’s dedicated to using his cross-disciplinary skills to develop medical image AI systems and solve important medical problems. Under his leadership, aetherAI has captured the largest share of the digital pathology market in Taiwan.

All faculty and students are welcome to join.