220122 Jupyter Meets the Earth：量身打造地科研究的軟體生態系

Title:
Jupyter Meets the Earth：量身打造地科研究的軟體生態系

Speaker:
鄭懷傑(Whyjay Zheng), PhD, University of California Berkeley

Time:
01/22 (Sat.) 6 pm PST, 7 pm MST, 8 pm CST, 9 pm EST
01/23 (Sun.) 3 am CET, 10 am Taiwan
Time zone conversion tool

Keywords:
Earth science, data science, cryosphere science, remote sensing, atmospheric science, Jupyter, Earth science, science communication, research, education

Abstract:
隨著地球科學的觀測資料與速度日漸提升，地科研究的社群正面臨許多新挑戰，包括如何在有限的時間內獲取並分析不同尺度的資料、交流最新的研究方法、重現研究成果，以及如何讓論文讀者更有效率的吸收知識等等。Jupyter meets the Earth 計畫團隊包含了地球科學研究者和資料科學工作者，我們以軟體開發的角度解決上述問題，採用研究者的即時回饋設計基於 Jupyter 的開源研究工具。我會從幾個活躍的使用例出發，包括大氣科學、遙測與冰雪圈研究，分享新的 Jypyter 生態系如何改善地科研究的效率，並增強科研與教育工作者、產業與社會大眾的交流。

220108 Re-Examine Musculoskeletal Disease by Deep Learning

Title:
Re-Examine Musculoskeletal Disease by Deep Learning

Speaker:
張瀚 (Han Chang), PhD, Boston University

Time:
01/08 (Sat.) 8 pm PST, 9 pm MST, 10 pm CST, 11 pm EST
01/09 (Sun.) 5 am CET, 12 pm Taiwan
Time zone conversion tool

Keywords:
Medical Images, Deep Learning, Osteoarthritis, MRI, Deep generative Models

  為配合本研究發表時程，本次錄影將延遲上傳

Abstract:
The management of degenerative and age-related diseases are expected to be one of the prominent challenges to Taiwan’s medical system as the population ages. Osteoarthritis (OA) is one of the most prevalent and debilitating degenerative and age-related diseases, with its burden expected to increase by several folds around the globe due to population aging and risks factors such as increasing rates of obesity. Deep learning-based imaging biomarkers provide great promise in specifying novel treatments target, facilitating new strategies of early detection and prevention techniques, and identifying population that is vulnerable to disease progression and disability. It is therefore of great importance to explore these new opportunities to alleviate the burden of OA and other age-related musculoskeletal diseases, especially in an aging society of Taiwan. In this talk, we will cover our efforts to study morphological biomarkers of early OA and onset of OA by performing automatic knee MRI segmentation across various population and development of the explainable AI models to identify potential target of treatment of reducing osteoarthritis (OA) pain and minimizing likelihood of surgical intervention.