2025年6月28日 星期六

健康照護科技專題(二) Syllabus 2025 Fall

 

健康照護科技專題(二) 課程綱要 (Syllabus)

Special Topics in Healthcare Technology (II)
2025 Fall Semester

一、課程基本資訊 (Course Information)

  • 學分數： 3學分

  • 授課對象： 研究所碩專碩士學生

二、課程簡介 (Course Description)

本課程為「健康照護科技專題」的進階課程，聚焦於運用尖端人工智慧（AI）與電腦視覺（Computer Vision）技術，解決特定的健康照護挑戰。課程將從環境感知與空間識別出發，深入探討視覺障礙者的導航輔助技術，涵蓋障礙物偵測與智慧避障演算法。接著，課程將轉向心智健康與認知輔助領域，探索如何利用科技工具進行多工任務訓練以延緩失智，並導入AI於情感分析與心理諮商的應用。最後，課程將介紹一種專門的數據編碼技術 (Vibe Coding)，並展望未來AI在醫療、復健及特殊需求照護領域的革命性發展。

本課程強調理論與實踐並重，學生將透過多次實作練習(Homework)與專題設計，將所學演算法與概念應用於真實世界的健康照護情境中。

三、課程目標 (Learning Objectives)

修習本課程後，學生應能：

  1. 分析與比較 視覺語言模型(VLM)與傳統電腦視覺在環境理解上的差異與應用。

  2. 設計與評估 一個針對視覺障礙者的安全路徑規劃系統原型。

  3. 實作與驗證 基於電腦視覺的動態與靜態障礙物偵測演算法。

  4. 理解並應用 Voronoi圖等核心演算法於無人載具或輔具的避障路徑規劃。

  5. 設計 針對心智障礙者或失智症患者的科技輔具或互動訓練遊戲。

  6. 探索與批判性評估 AI在情感分析與心理健康輔助應用中的潛力、可行性與倫理界線。

  7. 掌握 特定數據處理技術(Vibe Coding)，並應用於不完整數據的分析與逆向工程。

  8. 整合與前瞻 AI、VR等新興科技在未來長期照護、復健醫療與特殊教育領域的應用趨勢。

四、授課方式 (Teaching Methods)

  • 主題式講授 (Thematic Lectures): 深入講解各單元核心概念與技術。

  • 案例分析 (Case Studies): 剖析國內外最新的健康照護科技產品與研究。

  • 實作工作坊 (Hands-on Workshops): 程式設計練習、演算法實作、輔具設計練習。

  • 分組討論與報告 (Group Discussion & Presentation): 針對特定議題進行探討與分享。

  • 專題導向學習 (Project-Based Learning): 學生將完成數個小型作業與一項期末專題報告。

五、評分方式 (Grading Policy)

  • 實作作業 (Homework Assignments): 

  • 課堂練習與參與 (In-class Exercises & Participation): 

  • 期末專題報告 (Final Project & Presentation): 

六、課程進度與內容 (Course Schedule & Content)


This course is structured thematically over 18 weeks. Each unit will cover specific concepts, theories, and practical applications, supported by in-class activities and homework assignments.

Weeks 1-2: Unit 1 - Understanding Environments (環境理解)
This introductory unit establishes the foundation for the course, focusing on how AI perceives and interprets physical spaces for healthcare applications.

  • 課程開場 (Course Introduction): Overview of syllabus, learning objectives, and grading policy.

  • 核心技術比較 (Core Technology Comparison):

    • Visual Language Models (VLM) vs. traditional AI Computer Vision.

    • Illustrative examples of how each technology interprets a scene.

  • 空間感知應用 (Spatial Awareness Applications):

    • 空間資訊擷取與識別輔助 (Acquiring and using spatial information for assistance).

    • 環境危害偵測 (Hazard Detection), e.g., identifying spills, obstacles, or unsafe conditions.

  • 前瞻性分析 (Forward-looking Analysis):

    • 意外事件預測與模擬 (Predicting and simulating potential accidents).

Weeks 3-4: Unit 2 - 盲人導航影像辨識 (Image Recognition for Visually Impaired Navigation)
This unit dives into a specific application: creating assistive navigation technology for individuals with visual impairments.

  • 場景辨識技術 (Scene Recognition Techniques): For both indoor and outdoor environments.

  • 3D導航的必要性 (The Need for 3D Navigation):

    • Discussion: Why is 2D information insufficient?

    • The role of depth information in constructing a safe and accurate spatial map.

  • 實作任務 (Practical Assignment):

    • 作業一 (HW#1) 發佈：盲人導航影像辨識 (Safe Route Planning). Students will be tasked with developing a concept or basic algorithm for identifying a safe path from visual data.

Weeks 5-6: Unit 3 - 障礙偵測演算法 (Obstacle Detection Algorithms)
Building on navigation, this unit focuses on the critical sub-problem of detecting potential obstacles in real-time.

  • 障礙物分類 (Obstacle Categorization):

    • Algorithms for static obstacle detection.

    • Techniques for dynamic obstacle detection and motion tracking.

  • 語意理解 (Semantic Understanding):

    • Is it a barrier or not? Applying semantic segmentation to differentiate between true obstacles and benign objects.

  • 進度考核 (Milestone & New Task):

    • 作業一 (HW#1) 繳交 (Submission).

    • 作業二 (HW#2) 發佈：障礙偵測演算法實作 (Obstacle Detection Implementation).

Weeks 7-8: Unit 4 - 避障演算法 (Obstacle Avoidance Algorithms)
Once an obstacle is detected, the system must intelligently navigate around it. This unit explores the algorithms that make this possible.

  • 核心演算法介紹 (Introduction to Core Algorithms):

    • An introduction to the Voronoi diagram and its application in pathfinding.

  • 應用案例分析 (Application Case Studies):

    • General strategies in autonomous vehicle obstacle avoidance.

    • Deep Dive: The OmniSafe Project

      • Following Voronoi on rural roads.

      • Following Voronoi in country alleys.

      • Voronoi-based safe routing planning.

Week 9: 期中回顧與專題討論 (Midterm Review & Project Discussion)
This week serves as a checkpoint to consolidate learning and begin planning for the final project.

  • 課程回顧 (Course Review): Synthesizing concepts from Units 1-4.

  • 進度考核 (Milestone):

    • 作業二 (HW#2) 繳交 (Submission).

  • 期末專題腦力激盪 (Final Project Brainstorm): Discussion of potential topics and formation of project groups.

Weeks 10-11: Unit 5 - 訊息處理與心智障礙生活輔助 (Information Processing for Cognitive Disability Assistance)
The course focus shifts from physical to cognitive and mental health assistance.

  • 認知訓練科技 (Technology for Cognitive Training):

    • Designing multi-tasking training programs to potentially delay dementia.

    • Applying gamification principles to enhance engagement.

  • 互動設計 (Interaction Design):

    • Principles of Human-Computer Interaction (HCI) for users with cognitive impairments.

  • 課堂實作練習 (In-Class Activity):

    • 練習：輔具設計腦力激盪 (Exercise: Assistive Device Design Brainstorm).

Weeks 12-13: Unit 6 - AI輔助情感分析 (AI-Powered Emotion Analysis)
This unit explores the sensitive and rapidly evolving field of using AI to understand and respond to human emotions.

  • 情感與AI (Emotion & AI):

    • 課堂實作練習：情感、情緒、人工智慧 (Exercise: Emotion, Mood, and AI).

    • Technical approaches to emotion detection (e.g., from text, voice, facial expressions).

  • AI與心理諮商 (AI in Counseling):

    • Discussion: The feasibility, potential, and ethical boundaries of AI-assisted psychotherapy.

  • 實作任務 (Practical Assignment):

    • 作業三 (HW#3) 發佈：諮商輔助App概念設計 (Counseling Assistance App Concept Design).

Week 14: Unit 6 (cont.) - AI輔助情感分析(II) (Emotion Analysis II)
This session continues the exploration of mental health tech with a focus on practical design and application.

  • 使用者介面設計 (User Interface Design):

    • Best practices for designing interfaces for mental health applications.

  • 專題應用探討 (Thematic Application Deep Dive):

    • Designing an app to help users establish healthy personal "boundaries."

  • 案例分析 (Case Study):

    • The process and importance of clinical validation for Mental Health Apps.

  • 進度考核 (Milestone):

    • 作業三 (HW#3) 繳交 (Submission).

Weeks 15-16: Unit 7 - Vibe Coding (特定數據編碼技術)
This unit introduces a specialized data processing methodology, exploring its principles and applications.

  • 從零開始 (From Scratch):

    • Understanding the foundational principles and implementing Vibe Coding.

  • 處理不完整數據 (Handling Imperfect Data):

    • Applying Vibe Coding to datasets with missing or incomplete information.

  • 逆向工程應用 (Reverse Engineering Application):

    • Using Vibe Coding as a tool for reverse engineering data formats or systems.

Week 17: Unit 8 - 展望未來 (Future Outlook) & Final Presentations (Part 1)
The final unit looks at the convergence of technologies and reviews groundbreaking achievements, while students begin to present their final projects.

  • 科技整合趨勢 (Technological Convergence):

    • The future of AI + VR + Gaming in healthcare and rehabilitation.

  • 關鍵突破回顧 (Significant Breakthrough Review - SVG):

    • The AI revolution in medicine (2022-2025).

    • Breakthroughs in AI for long-term care and geriatric rehabilitation.

    • Breakthroughs in AI applications for autism spectrum disorder.

  • 期末成果發表 (Final Presentations):

    • Presentations from the first half of the class.

Week 18: 期末成果發表 (Final Presentations) & 課程總結
The course concludes with the remaining project presentations and a final summary.

  • 期末成果發表 (Final Presentations):

    • Presentations from the second half of the class.

  • 課程總結 (Course Wrap-up):

    • Review of key takeaways from the semester.

    • Discussion of future learning paths and career opportunities in healthcare technology.


七、指定用書與參考資料 (Textbooks & References)

  • 指定用書： 無。本課程將提供每週主題所需之學術論文、技術文件、線上文章與程式碼範例。

  • 參考資料：

    • 學術期刊：IEEE Transactions on Affective Computing, IEEE Transactions on Neural Systems and Rehabilitation Engineering, JMIR (Journal of Medical Internet Research), etc.

    • 技術文獻：OpenCV, TensorFlow, PyTorch 官方文件。

    • 線上資源：arXiv, Google Scholar, Medium (Towards Data Science)。

八、學術誠信 (Academic Integrity)

所有繳交的作業與報告皆須為學生個人或該組別的原創成果。嚴禁任何形式的抄襲或舞弊行為。若引用他人資料，必須註明出處。違反學術倫理之規定，將依校規處理，該次作業或報告以零分計算。



線上資源


Unit 1: Understanding environments

VLM vs. AI  Vision (illustrate)


ＡＩ輔助情感分析(II)  　（界線主題app )


Mental Health App 心理健康遊戲應用程式的臨床驗證





Unit 7:


Vibe Coding from scratch

張貼者： 於

沒有留言:

張貼留言

訂閱： 張貼留言 (Atom)