Khyatimirani/pcos-clinical-guideline-101

--- license: apache-2.0 task_categories: - question-answering language: - en tags: - medical - fertility - hormonal-health - pcos - women-health size_categories: - n<1K --- ### PCOS Clinical Q&A Dataset (Dr. Thaïs Aliabadi Guideline-Based) ## 📌 Overview This dataset contains 500+ clinically grounded conversational Q&A pairs simulating interactions between a PCOS patient and a clinically informed assistant. The dataset is derived from the following primary source: Female Hormone Health, PCOS, Endometriosis, Fertility & Breast Cancer | Dr. Thaïs Aliabadi This conversation features Dr. Thaïs Aliabadi (board-certified OB/GYN) explaining symptoms, diagnosis criteria, root causes, and treatment approaches for PCOS and related conditions. ## 🎯 Objective To build a clinically reliable conversational dataset that: Reflects real-world patient questions Provides clear, medically grounded explanations Includes statistics and prevalence data where available Helps users understand the “why” behind symptoms and treatments Encourages safe escalation to healthcare providers when needed. ## 🧠 Dataset Design Philosophy Single source of truth: Dr. Thaïs Aliabadi’s clinical explanations No hallucinated or external medical knowledge added Questions are reverse-engineered from real patient behavior Answers are: Clinically accurate Empathetic but not fluffy Educational and explanatory Safety-aware ## 📊 Key Medical Grounding (from source) PCOS affects ~15–20% of women of reproductive age Diagnosis is based on 2 out of 3 criteria: Androgen excess (acne, hair growth, hair thinning) Ovulatory dysfunction (irregular periods) Polycystic ovaries / high AMH PCOS is often underdiagnosed or misdiagnosed despite being highly prevalent Insulin resistance and inflammation are key underlying drivers ## 📂 Dataset Structure The dataset is provided in JSONL format: `{ "id": "pcos_001", "section": "Diagnosis", "tags": ["irregular_periods", "androgens"], "messages": [ { "role": "user", "content": "My periods are very irregular. Could this be PCOS?" }, { "role": "assistant", "content": "Irregular periods are one of the key indicators of ovulatory dysfunction, which is one of the three diagnostic criteria for PCOS. Typically, a diagnosis requires meeting at least two of the three criteria, so it would be important to evaluate for other features as well." } ] }` ## 🧩 Topics Covered PCOS symptoms (cycles, acne, hair growth, weight changes) Diagnosis & clinical criteria Hormonal imbalance & androgens Insulin resistance & inflammation Fertility & ovulation Lifestyle interventions Supplements (evidence-based discussion from source) Mental health considerations Misdiagnosis & delayed diagnosis When to seek medical help ## ⚠️ Safety & Intended Use This dataset is for: Education AI assistant training Research The assistant: Does not prescribe medication Does not replace a doctor Encourages clinical consultation when necessary ## 🚀 Use Cases Fine-tuning LLMs for women’s health assistants Building PCOS-specific AI companions Clinical conversational AI research Patient education tools Preventive care applications ## 🧪 Limitations Based on a single expert source Not a substitute for: Clinical diagnosis Personalized treatment Some clinical nuances may require multi-source validation ## 🤝 Contributions We welcome: Multi-source clinical expansion Localization (Hindi / voice-first) IVF, fertility, and menstrual health extensions ## 📜 License (Choose one based on your preference) CC BY-NC 4.0 (recommended for healthcare datasets) or custom research-use license ## 💬 Contact If you are working in: Women’s health PCOS / fertility Healthcare AI We’d love to collaborate and build meaningful solutions in this space. email: [e-mail]--- 许可证：Apache 2.0 任务类别：问答任务 语言：英语 标签：医疗、生育健康、激素健康、多囊卵巢综合征（Polycystic Ovary Syndrome，PCOS）、女性健康 数据规模：样本量小于1000 --- ### 基于Thaïs Aliabadi医生临床指南的多囊卵巢综合征（Polycystic Ovary Syndrome，PCOS）临床问答数据集 ## 📌 概述 本数据集包含500余条基于临床场景的对话式问答对，模拟多囊卵巢综合征（PCOS）患者与具备临床专业知识的助手之间的交互。 本数据集源自以下核心来源： 《女性激素健康、多囊卵巢综合征、子宫内膜异位症、生育与乳腺癌 | Thaïs Aliabadi医生》 本对话内容由认证妇产科医师（OB/GYN）Thaïs Aliabadi医生讲解多囊卵巢综合征及相关病症的症状、诊断标准、致病根源与治疗方案。 ## 🎯 目标 构建具备临床可靠性的对话式数据集，旨在达成以下目标： 1. 还原真实患者的临床问询场景 2. 提供清晰且基于医学依据的解释说明 3. 酌情纳入统计数据与患病率信息 4. 帮助用户理解症状与治疗方案背后的医学逻辑 5. 提示用户在必要时及时寻求医疗专业人员的帮助 ## 🧠 数据集设计理念 采用单一可信来源：仅使用Thaïs Aliabadi医生的临床讲解内容，未添加任何虚构或外部医学知识。 问题均基于真实患者的问诊行为反向构建，答案需满足以下要求： - 临床准确性 - 共情且不失专业严谨性 - 具备教育性与解释性 - 符合医疗安全规范 ## 📊 核心医学依据（源自原始来源） - 多囊卵巢综合征（PCOS）影响约15%~20%的育龄女性 - 诊断需满足三项标准中的至少两项： 1. 雄激素过量（痤疮、多毛、毛发稀疏） 2. 排卵功能障碍（月经不规律） 3. 多囊卵巢/抗缪勒管激素（Anti-Müllerian Hormone，AMH）水平升高 - 尽管多囊卵巢综合征（PCOS）患病率较高，但常存在漏诊或误诊情况 - 胰岛素抵抗与炎症反应是核心致病驱动因素 ## 📂 数据集结构 本数据集以JSONL格式提供，示例如下： json { "id": "pcos_001", "section": "诊断", "tags": ["irregular_periods", "androgens"], "messages": [ { "role": "user", "content": "我的月经非常不规律，这可能是多囊卵巢综合征（PCOS）吗？" }, { "role": "assistant", "content": "月经不规律是排卵功能障碍的关键指征之一，而排卵功能障碍是多囊卵巢综合征（PCOS）三项诊断标准中的一项。通常来说，诊断需满足至少两项标准，因此还需对其他特征进行评估。" } ] } ## 🧩 涵盖主题 多囊卵巢综合征（PCOS）症状（月经周期、痤疮、毛发生长、体重变化） 诊断与临床标准 激素失衡与雄激素异常 胰岛素抵抗与炎症反应 生育与排卵问题 生活方式干预 营养补充剂（基于原始来源的循证讨论） 心理健康考量 误诊与延迟诊断 何时需寻求医疗帮助 ## ⚠️ 安全须知与适用场景 本数据集适用于以下场景： - 医学教育 - AI助手训练 - 学术研究 本助手需注意： - 不提供药物处方建议 - 不可替代执业医师的诊疗服务 - 提示用户在必要时寻求临床咨询 ## 🚀 应用场景 针对女性健康助手的大语言模型（Large Language Model，LLM）微调 构建专属多囊卵巢综合征（PCOS）的AI陪护工具 临床对话式AI研究 患者教育工具 预防性医疗应用 ## 🧪 局限性 本数据集仅基于单一专家来源构建，不可替代： - 临床诊断 - 个性化治疗方案 部分临床细节需通过多源数据进行验证 ## 🤝 贡献与合作 我们欢迎以下形式的贡献： - 多源临床内容扩展 - 本地化适配（印地语/语音优先） - 试管婴儿（In Vitro Fertilization，IVF）、生育与月经健康相关内容的拓展 ## 📜 许可证 （可根据需求二选一） CC BY-NC 4.0（医疗数据集推荐使用） 或自定义研究专用许可证 ## 💬 联系方式 如果您从事以下领域： - 女性健康 - 多囊卵巢综合征（PCOS）/生育健康 - 医疗AI 我们期待与您合作，共同在该领域构建有意义的解决方案。 邮箱：[e-mail]