16 Sept. 2021
aetherAI, Asia’s leading medical image AI solution provider, Novartis Taiwan, a leader in the pharmaceutical industry, and Linkou Chang Gung Memorial Hospital (Linkou CGMH), the largest medical center in Taiwan, recently established a strategic partnership to introduce AI to diagnostic hematopathology, specific for myeloproliferative neoplasm, MPN.
Impressed by aetherAI’s digital pathology solutions and its AI capabilities, Novartis Taiwan invited aetherAI to collaborate on MPN diagnostic support solution.
Following the success of aetherAI Hema, the world’s first bone marrow differential AI system, the latest MPN project with Novartis Taiwan will further explore the power of AI, pushing the boundaries of AI-powered hematology solutions. Trained on the world’s largest dataset of over 1 million carefully curated cells, the plug-and-play aetherAI Hema can provide 15 subtypes of differential count with accuracy as high as 94%.
Since 2018, aetherAI has been collaborating with National Taiwan University Hospital, Linkou CGMH, and Chi Mei Medical Center on AI-powered hematology solutions and has optimized procedures such as bone marrow smear AI differential counting and lymph node metastases AI detection. Joe Yeh, M.D., aetherAI Co-founder and CEO points out that all these steps were part of the journey to this joint MPN project with Novartis.
Dr. Yeh adds that this partnership with Novartis Taiwan serves as a great recognition of aetherAI’s world-class strength, covering digital pathology transformation, medical image AI adoption, and advanced AI research capabilities, and so forth. Leveraging its leading diagnostic imaging AI technologies, aetherAI not only supports medical centers and hospitals but also reaches out to pharmaceutical companies, streamlining workflow with Artificial Intelligence.
Founded in October of 2015, aetherAI is ranked no. 1 in the digital pathology and AI market in Taiwan. Since 2017, aetherAI has been tripling its annual revenue for three years in a row, actively exploring the global markets, including Japan, the Middle East, and the United States. aetherAI-supported researches have been published in prestigious international journals, to name a few: <An Annotation-free Whole-slide Training Approach to Pathological Classification of Lung Cancer Types by Deep Neural Network> by Nature Communications, <Identification of nodal micrometastasis in colorectal cancer using deep learning on annotation-free whole-slide images> by Modern Pathology, etc.
伴隨AI技術成熟、新冠肺炎加速數位轉型，林口長庚紀念醫院、雲象科技、台灣諾華三方跨界攜手，奠基於林口長庚醫院病理數位化累積的龐大資料庫，結合雲象科技AI技術，進行深度學習、訓練深度神經網路來辨識骨髓細胞的形態、特徵與空間分布情形，打造「血液病理AI輔助判讀應用」，以提供客觀且量化的數據，輔助病理醫師作出高效、精準的「骨髓增生性腫瘤」(myeloproliferative neoplasm，簡稱MPN)診斷；而台灣諾華則長期投入血液腫瘤研發治療，促成策略合作。希冀藉此提升國內血液腫瘤篩檢量能，幫助病患獲得即時診斷及治療。
疫情加速智慧醫療。林口長庚完成病理玻片數位化，奠基AI應用基礎
研調機構Research and Markets今夏報告指出，全球數位病理市場規模持續攀升，特別在AI輔助影像分析的應用上，成長動能強勁。林口長庚紀念醫院解剖病理部陳澤卿主任指出，林口長庚每月有近萬個案例、高達上萬筆的病理玻片需要判讀，病理團隊每日皆須面臨數量龐大且急迫的病例。為了能及早且精準幫助病患確診，長庚醫院已將病理玻片數位化，為全台少數完成跨院區病理科數位化的醫療院所，大幅提升判讀方便性。長庚醫院五個院區採用雲象科技數位病理系統，目前林口長庚數位化玻片已累積超過38萬片。
血液腫瘤診斷異常複雜，高品質資料、AI技術、計算能量缺一不可
雲象科技創辦人暨執行長葉肇元醫師指出，血液疾病的診斷與治療相當困難，然而因病患數不如其他器官的癌症，故新技術如AI較不會在第一時間被應用在血液疾病上；不過，對血液疾病來說，以形態學為基礎的病理診斷扮演關鍵角色，而形態辨識正是AI在醫療上能有最大發揮空間的面向。雲象科技自2018年起，陸續與台大醫院、林口長庚醫院、奇美醫院，分別進行骨髓抹片細胞型態辨識、淋巴瘤型態診斷及預後分析等血液疾病AI的應用，所累積的技術與應用開發經驗，挹注於開發「骨髓增生性腫瘤」AI輔助判讀與病理診斷。展望血液病理的發展，葉肇元表示，相信在AI輔助下，形態診斷的重要性會再次提升，和近年備受矚目的分子及基因診斷相輔相成，進一步強化血液疾病診斷以及治療的品質。
運用AI於血液病理診斷，可望提升血液癌症判讀精準度，有助及早治療
林口長庚紀念醫院血液科郭明宗醫師表示，骨髓增生性腫瘤臨床上常見的三種類別，從診斷到治療都是挑戰。就病患臨床表現而言，因為沒有可觸及的腫塊，或是因疾病引起的其他症狀，像是出血、中風等，正確的診斷成為一大挑戰，必須仰賴骨髓切片。針對切片的判讀，林口長庚紀念醫院解剖病理部莊文郁副主任解釋，病理醫師必須在顯微鏡下仔細評估各種造血細胞的數量及形態，特別是巨核細胞的形態特徵、數量及空間分布，才能得到精準的診斷。人工判讀難取得客觀量化的結果，且會存在不同診斷者間的差異，特別是對於該疾病較不熟悉或經驗較少的醫師更加困難。透過高品質、經過專家標註資料的訓練，AI輔助影像分析可以讓診斷流程有更客觀一致的量化標準，提升診斷的準確率。
改善患者生活品質與延長存活期是各界對於癌症治療的共同目標，諾華腫瘤(台灣)總經理陳喬松表示，身為全球製藥領導者，自第一代標靶治療到目前最創新的細胞基因療法，諾華持續為癌症治療創新里程碑。同時，亦致力運用資料科學以發展先進藥品，運用大數據分析及AI數位科技是諾華重要的策略方向。目前血液腫瘤的早期診斷仍有未被滿足的需求，此次攜手雲象科技及長庚醫院，正是諾華企業宗旨「重新創想醫藥未來以改善並延長人們的生命」的實踐。期望結合三方優勢，能幫助更多血液腫瘤病患及早診斷並接受治療，降低疾病惡化的風險，延續病患的生命並提升生活品質。
部分相關報導