[{"data":1,"prerenderedAt":23},["ShallowReactive",2],{"blog-detail-1005-zh":3},{"id":4,"date":5,"date_gmt":6,"modified":7,"modified_gmt":8,"title":9,"excerpt":11,"content":12,"featured_image":15,"categories":16,"tags":18,"sort_order":19,"focus_keyphrase":20,"seo_title":21,"seo_description":22},1005,"2025-12-02T15:29:08","2025-12-02T07:29:08","2026-08-13T14:29:27","2026-08-13T06:29:27",{"rendered":10},"[易学堂]AI赋能促排卵：一种基于人工智能的促排卵决策支持系统","\u003Cp>本次介绍的这篇论文将可解释AI技术引入IVF临床中的促排卵方案决策过程，提出一个覆盖促排全过程的智能决策系统（\u003C\u002Fp>\n",{"rendered":13,"protected":14},"\u003Cspan style=\"font-size: 17px;font-family: mp-quote, "PingFang SC", system-ui, -apple-system, BlinkMacSystemFont, "Helvetica Neue", "Hiragino Sans GB", "Microsoft YaHei UI", "Microsoft YaHei", Arial, sans-serif;letter-spacing: 0.034em;font-style: normal;font-weight: normal\">\u003Cspan style=\"font-size: 16px\">本次介绍的这篇论文将可解释AI技术引入IVF临床中的促排卵方案决策过程，提出一个覆盖促排全过程的智能决策系统（AACS）。不仅实现了取卵时机判断的量化标准，更提供了可视化的处方推荐依据，为AI在IVF中的应用提供了新的思路。不过该系统模型基于回顾性数据构建，以专家医生的实际操作为标准，未纳入促排是否成功的结果信息，且系统目前仅适用于拮抗剂方案，仍有进一步完善空间。\u003C\u002Fspan>\u003C\u002Fspan>\n\n\n\n\u003Cspan style=\"font-family: BlinkMacSystemFont, -apple-system, "Segoe UI", Roboto, Oxygen, Ubuntu, Cantarell, "Fira Sans", "Droid Sans", "Helvetica Neue", sans-serif;font-size: 16px;font-style: normal;font-weight: 400;letter-spacing: normal;text-align: start;text-indent: 0px;text-transform: none;float: none;display: inline !important\" data-pm-slice=\"0 0 []\">\u003Cspan>\u003Cspan style=\"font-size: 14px\">Asada Y, Shinohara T, Yonezawa S, Kinugawa T, Asano E, Kojima M, Fukunaga N, Hashizume N, Hashiba Y, Inoue D, Mizuno R, Saito M, Kabeya Y. Development of an AI-based support system for controlled ovarian stimulation. Reprod Med Biol. 2024 Sep 1;23(1):e12603. doi: 10.1002\u002Frmb2.12603. PMID: 39224211; PMCID: PMC11366684.\u003C\u002Fspan>\u003C\u002Fspan>\u003C\u002Fspan>\n\n\n\n\u003Ch2>研究背景\u003C\u002Fh2>\n\n\n\n\u003Cspan style=\"font-size: 17px;font-family: mp-quote, "PingFang SC", system-ui, -apple-system, BlinkMacSystemFont, "Helvetica Neue", "Hiragino Sans GB", "Microsoft YaHei UI", "Microsoft YaHei", Arial, sans-serif;line-height: 1.6;letter-spacing: 0.034em;font-style: normal;font-weight: normal\">促排卵（COS）作为体外受精（IVF）治疗的关键环节，其效果直接影响卵母细胞的获取数量和质量。然而，由于COS方案的制定高度依赖医师经验，年轻医师因缺乏临床经验往往导致治疗效果不稳定，加之全球不孕症发病率的持续上升，开发高效、标准化的COS方案显得尤为迫切。为此，本研究开发了一种基于人工智能（AI）的COS支持系统（AACS），旨在通过模拟专家医师的决策逻辑，优化取卵时机判断和药物处方推荐，从而提高治疗的一致性和成功率。\u003C\u002Fspan>\n\n\n\n\u003Cfigure class=\"wp-block-image size-full\">\u003Cimg decoding=\"async\" alt=\"图片\" class=\"rich_pages wxw-img fertsy-content-image fertsy-content-image--wide\" data-backh=\"385\" data-backw=\"578\" data-imgfileid=\"100000403\" data-ratio=\"0.6666666666666666\" data-s=\"300,640\" data-src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-1.png\" data-type=\"png\" data-w=\"1080\" style=\"display:block;max-width:90%;width:90%;height:auto;aspect-ratio:1 \u002F 0.666667;object-fit:contain;margin:24px auto;\" src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-1.png\">\u003C\u002Ffigure>\n\n\n\n\u003Ch2>研究方法\u003C\u002Fh2>\n\n\n\n\u003Cspan style=\"font-size: 17px;font-family: mp-quote, "PingFang SC", system-ui, -apple-system, BlinkMacSystemFont, "Helvetica Neue", "Hiragino Sans GB", "Microsoft YaHei UI", "Microsoft YaHei", Arial, sans-serif;line-height: 1.6;letter-spacing: 0.034em;font-style: normal;font-weight: normal\">研究采用回顾性队列设计，数据来源于日本Asada Ladies Clinic的IVF管理系统（2017年6月至2021年11月），共纳入5,969名患者的7,850个治疗周期，其中专家医师执行的1,068个周期被用于模型训练和验证。研究团队构建了两个核心模型：促排卵决策模型和处方推断模型。前者基于LightGBM算法，通过分析卵泡数量、大小、激素水平（FSH、LH、E2等）、AMH值和患者年龄等特征，输出是否建议取卵的决策；后者则包含四个子模型（hMG、hCG、Cetrorelix、Estradiol预测模型），分别用于推荐药物种类和剂量。模型性能通过5折交叉验证进行评估，主要指标包括AUC和准确率。\u003C\u002Fspan>\n\n\n\n\u003Cfigure class=\"wp-block-image size-full\">\u003Cimg decoding=\"async\" alt=\"图片\" class=\"rich_pages wxw-img fertsy-content-image fertsy-content-image--wide\" data-backh=\"459\" data-backw=\"553\" data-imgfileid=\"100000400\" data-ratio=\"0.8296296296296296\" data-src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-2.webp\" data-type=\"jpeg\" data-w=\"1080\" style=\"display:block;max-width:90%;width:90%;height:auto;aspect-ratio:1 \u002F 0.82963;object-fit:contain;margin:24px auto;\" src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-2.webp\">\u003C\u002Ffigure>\n\n\n\n\u003Cspan style=\"font-size: 17px;font-family: mp-quote, "PingFang SC", system-ui, -apple-system, BlinkMacSystemFont, "Helvetica Neue", "Hiragino Sans GB", "Microsoft YaHei UI", "Microsoft YaHei", Arial, sans-serif;line-height: 1.6;letter-spacing: 0.034em;font-style: normal;font-weight: normal\">\u003Cspan style=\"font-size: 14px\">AACS结构：促排卵决策模型和四种处方模型\u003C\u002Fspan>\u003C\u002Fspan>\n\n\n\n\u003Ch2>研究结果\u003C\u002Fh2>\n\n\n\n\u003Cspan style=\"font-size: 17px;font-family: mp-quote, "PingFang SC", system-ui, -apple-system, BlinkMacSystemFont, "Helvetica Neue", "Hiragino Sans GB", "Microsoft YaHei UI", "Microsoft YaHei", Arial, sans-serif;line-height: 1.6;letter-spacing: 0.034em;font-style: normal;font-weight: normal\">研究结果显示，促排卵决策模型在测试集中的AUC达到0.964，准确率高达96.9%。通过特征重要性分析发现，右侧卵巢第二大卵泡大小（重要性评分0.53）和左侧卵巢最大卵泡直径（评分0.14）是影响决策的核心因素。与此同时，处方推断模型的整体AUC为0.948，准确率为86.9%，其中hMG预测模型（AUC=0.914）的关键特征为AMH和LH水平，而Cetrorelix预测模型（AUC=0.966）则主要依赖刺激天数（stimdays）。SHAP分析进一步揭示了各临床变量对处方决策的具体影响，为模型的透明性和可解释性提供了有力支持。\u003C\u002Fspan>\n\n\n\n\u003Cfigure class=\"wp-block-image size-full\">\u003Cimg decoding=\"async\" alt=\"图片\" class=\"rich_pages wxw-img fertsy-content-image fertsy-content-image--wide\" data-backh=\"182\" data-backw=\"553\" data-imgfileid=\"100000397\" data-ratio=\"0.3287037037037037\" data-src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-3.webp\" data-type=\"jpeg\" data-w=\"1080\" style=\"display:block;max-width:90%;width:90%;height:auto;aspect-ratio:1 \u002F 0.328704;object-fit:contain;margin:24px auto;\" src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-3.webp\">\u003C\u002Ffigure>\n\n\n\n\u003Cfigure class=\"wp-block-image size-full\">\u003Cimg decoding=\"async\" alt=\"图片\" class=\"rich_pages wxw-img fertsy-content-image fertsy-content-image--wide\" data-backh=\"206\" data-backw=\"553\" data-imgfileid=\"100000398\" data-ratio=\"0.3712962962962963\" data-src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-4.webp\" data-type=\"jpeg\" data-w=\"1080\" style=\"display:block;max-width:90%;width:90%;height:auto;aspect-ratio:1 \u002F 0.371296;object-fit:contain;margin:24px auto;\" src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-4.webp\">\u003C\u002Ffigure>\n\n\n\n\u003Cfigure class=\"wp-block-image size-full\">\u003Cimg decoding=\"async\" alt=\"图片\" class=\"rich_pages wxw-img fertsy-content-image fertsy-content-image--wide\" data-backh=\"193\" data-backw=\"553\" data-imgfileid=\"100000399\" data-ratio=\"0.3490740740740741\" data-src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-5.webp\" data-type=\"jpeg\" data-w=\"1080\" style=\"display:block;max-width:90%;width:90%;height:auto;aspect-ratio:1 \u002F 0.349074;object-fit:contain;margin:24px auto;\" src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-5.webp\">\u003C\u002Ffigure>\n\n\n\n\u003Cfigure class=\"wp-block-image size-full\">\u003Cimg decoding=\"async\" alt=\"图片\" class=\"rich_pages wxw-img fertsy-content-image fertsy-content-image--wide\" data-backh=\"183\" data-backw=\"553\" data-imgfileid=\"100000396\" data-ratio=\"0.3314814814814815\" data-src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-6.webp\" data-type=\"jpeg\" data-w=\"1080\" style=\"display:block;max-width:90%;width:90%;height:auto;aspect-ratio:1 \u002F 0.331481;object-fit:contain;margin:24px auto;\" src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-6.webp\">\u003C\u002Ffigure>\n\n\n\n\u003Cfigure class=\"wp-block-image size-full\">\u003Cimg decoding=\"async\" alt=\"图片\" class=\"rich_pages wxw-img fertsy-content-image fertsy-content-image--wide\" data-backh=\"186\" data-backw=\"553\" data-imgfileid=\"100000401\" data-ratio=\"0.33611111111111114\" data-src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-7.webp\" data-type=\"jpeg\" data-w=\"1080\" style=\"display:block;max-width:90%;width:90%;height:auto;aspect-ratio:1 \u002F 0.336111;object-fit:contain;margin:24px auto;\" src=\"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-image-7.webp\">\u003C\u002Ffigure>\n\n\n\n\u003Cspan style=\"font-size: 17px;font-family: mp-quote, "PingFang SC", system-ui, -apple-system, BlinkMacSystemFont, "Helvetica Neue", "Hiragino Sans GB", "Microsoft YaHei UI", "Microsoft YaHei", Arial, sans-serif;line-height: 1.6;letter-spacing: 0.034em;font-style: normal;font-weight: normal\">\u003Cspan style=\"font-size: 14px\">所有模型的性能\u003C\u002Fspan>\u003C\u002Fspan>\n\n\n\n\u003Ch2>研究创新点\u003C\u002Fh2>\n\n\n\n\u003Cspan style=\"font-size: 17px;font-family: mp-quote, "PingFang SC", system-ui, -apple-system, BlinkMacSystemFont, "Helvetica Neue", "Hiragino Sans GB", "Microsoft YaHei UI", "Microsoft YaHei", Arial, sans-serif;line-height: 1.6;letter-spacing: 0.034em;font-style: normal;font-weight: normal\">本研究在辅助生殖领域实现了多项重要创新：AACS系统作为首个覆盖促排卵全流程的AI决策支持工具，通过整合取卵时机判断与个性化药物推荐功能，填补了临床决策支持系统的空白；该系统创新性地将高预测性能（AUC>0.9）与可解释性相结合，借助特征重要性分析和SHAP可视化技术，使复杂的AI决策过程变得透明可信；尤为重要的是，系统已成功应用于实际临床工作，不仅为医师提供实时决策支持，更通过标准化建议显著提升了年轻医师的培养效率，实现了从理论研究到临床实践的重要跨越。\u003C\u002Fspan>\n\n\n\n\u003Ch2>研究局限性\u003C\u002Fh2>\n\n\n\n\u003Cspan style=\"font-size: 17px;font-family: mp-quote, "PingFang SC", system-ui, -apple-system, BlinkMacSystemFont, "Helvetica Neue", "Hiragino Sans GB", "Microsoft YaHei UI", "Microsoft YaHei", Arial, sans-serif;line-height: 1.6;letter-spacing: 0.034em;font-style: normal;font-weight: normal\">尽管取得了显著成果，本研究仍存在一定局限性。由于数据来源于单一医疗中心，模型的泛化能力可能受到限制；此外，当前系统仅适配拮抗剂方案，尚未涵盖激动剂或微刺激方案；同时，本研究基于回顾性数据集，训练结果以医生的判断为准，没有考虑促排卵成功与否，因而其判断逻辑更多体现的是专家经验，而非结局导向。\u003C\u002Fspan>\n\n\n\n\u003Ch2>临床意义与展望\u003C\u002Fh2>\n\n\n\n\u003Cspan style=\"font-size: 17px;font-family: mp-quote, "PingFang SC", system-ui, -apple-system, BlinkMacSystemFont, "Helvetica Neue", "Hiragino Sans GB", "Microsoft YaHei UI", "Microsoft YaHei", Arial, sans-serif;line-height: 1.6;letter-spacing: 0.034em;font-style: normal;font-weight: normal\">本研究开发的AACS系统具有重要的临床应用价值，其通过人工智能技术实现了促排卵决策流程的标准化和智能化，显著降低了不同经验水平医师间的诊疗差异，使治疗效果更加稳定可靠。系统提供的客观量化指标和可视化决策依据，不仅优化了临床决策过程，也为医患沟通提供了科学参考。未来，研究团队计划将该系统扩展至更多方案类型，并引入更强的语言模型技术，进一步提升医患沟通效率。总体而言，该系统为辅助生殖医疗向标准化、数据化方向发展提供了新的实践路径，也为提高临床一致性与治疗成功率带来了新的可能。\u003C\u002Fspan>",false,"https:\u002F\u002Fshengshengyi.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fwechat-ar2vqt359oykmfl5vi50pqcsoyc7xvwza6dcc4bw-p6-tfpurpp7zvvppbrgrpqd-0-cover.jpg",[17],11,[],0,"AI赋能促排卵","AI赋能促排卵：一种基于人工智能的促排卵决策支持系统 | Fertsy","AI赋能促排卵：本文梳理相关背景、核心进展与行业意义，并介绍生生易在人工智能辅助生殖和IVF技术领域的实践。",1786874476559]