Medical AI · Active Health · Governance · Decision Support

Medical AI Portfolio for Active Health and Governance.

A curated portfolio of projects translating health data, AI models and compliance requirements into interpretable, auditable and decision-oriented tools.

医疗 AI、主动健康与合规转化型决策工具作品集。聚焦多源健康数据建模、模型解释、风险评估、伦理合规与智能决策支持。

Projects built around evidence, not slogans.

Each work is designed as a concrete project asset: a demo, a report, a method pipeline or a governance-ready documentation framework.

01 Core Project

Active Health AI Governance Toolkit

主动健康医疗 AI 合规评估工具链

A toolkit for turning medical AI prototypes into reviewable, explainable and governance-ready project assets.

Problem · Medical AI prototypes often lack data authorization, ethical review, model explanation and risk-control materials.
Output · Model card, data sheet, risk matrix, evaluation report and user-facing risk notice.
02 Demo Design

Fall Risk Assessment Demo

老年跌倒风险评估 Demo

A lightweight active-health demo that estimates fall risk and explains key contributing factors using structured health inputs.

Input · Age, fall history, gait, sleep, chronic disease count, medication count and activity level.
Output · Risk level, explanation, intervention suggestions and compliance notice.
03 Data System

Health Data Pipeline Dashboard

健康数据 ETL 与指标看板

A compact data workflow for cleaning, storing and visualizing multi-source health data with clear metric definitions.

Method · CSV / device logs → cleaning → PostgreSQL → quality checks → dashboard.
Stack · Python, Pandas, PostgreSQL, Docker, Metabase / Plotly.
04 Report Template

Medical AI Evaluation Report

医疗 AI 模型评估报告模板

A structured template for documenting model performance, explainability, subgroup behavior, limitations and deployment risk.

Evidence · AUC, sensitivity, specificity, calibration, subgroup analysis and external validation plan.
Use · Research reporting, platform presentation, ethics review and project communication.

A practical bridge from data to decision.

The work focuses on the middle layer between raw medical data and real-world application: modeling, explanation, validation and governance-ready translation.

Health Data Modeling

多源健康数据清洗、特征工程、风险建模与状态评估。

Explainable AI Evaluation

面向研发、临床和用户的分层模型解释与误差分析。

Governance-Ready Documentation

数据授权、伦理审查、隐私保护、风险矩阵与责任边界说明。

Decision Support Systems

将模型输出转化为可读报告、干预建议和平台展示材料。

Concrete assets for research, teaching and collaboration.

The site will gradually collect reusable templates, demo interfaces, sample reports and evaluation materials for medical AI projects.

Model Card Data Sheet Risk Matrix Evaluation Report Ethics Review Template Privacy Impact Checklist Demo Interface Dashboard Prototype

Built from a cross-disciplinary background.

Research Position

Bai MedAI Lab focuses on medical AI evaluation, active health risk modeling, explainable decision support and governance-ready project translation.

Background

Cross-disciplinary training in communication engineering, economic law, management science and medical AI research. The core strength is translating complex multi-source data into interpretable and decision-oriented outcomes.

For collaboration, review or academic communication.

This portfolio is currently under active development. The first stage focuses on selected project assets, demo prototypes and governance-ready documentation templates.