4zhvml8 发表于 2024-8-22 15:03:18

用ChatGPT直接来构建预测模型,效果怎么样?有作者用它写了一篇SCI论文


    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q3.itc.cn/q_70/images01/20240624/78bbcaf3d0be4746a1fdd6204ec6144f.png" style="width: 50%; margin-bottom: 20px;"></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">详情请点击右方:</strong><a style="color: black;"><strong style="color: blue;">欢迎报名!2024年临床预测模型培训班,零代码构建预测模型</strong></a></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q4.itc.cn/q_70/images01/20240613/e77fc27e9f074079a4ac1308d6c9ce69.png" style="width: 50%; margin-bottom: 20px;"></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">此刻</span>应该没人不<span style="color: black;">晓得</span><strong style="color: blue;">ChatGPT</strong>吧?就算<span style="color: black;">无</span>用过,<span style="color: black;">亦</span>应该听过。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q1.itc.cn/images01/20240625/3c4902796a854090bae894d44593cacb.png" style="width: 50%; margin-bottom: 20px;"><span style="color: black;">图源百度</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">做为</span>一个大型的自然语言模型,ChatGPT在语言生成和响应方面有着天然的<span style="color: black;">优良</span>。简单<span style="color: black;">来讲</span>,<span style="color: black;">咱们</span><span style="color: black;">能够</span>利用它来润色<span style="color: black;">文案</span>,阅读文献。<strong style="color: blue;">但最新<span style="color: black;">发布</span>的一篇<span style="color: black;">文案</span>却利用ChatGPT构建了一个新型临床预测模型,<span style="color: black;">发帖</span>一区(IF=4.1)!</strong></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">接下来本文将对这篇<span style="color: black;">文案</span>进行简单介绍。</p>
    <h1 style="color: black; text-align: left; margin-bottom: 10px;"><strong style="color: blue;"><span style="color: black;">文案</span>介绍</strong></h1>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">2024年5月31日,外国<span style="color: black;">科研</span>团队在期刊 <strong style="color: blue;">《American Journal of Ophthalmology》(医学一区top,IF=4.1)</strong><span style="color: black;">发布</span>了一篇题为:“Predicting glaucoma before onset using a large language model chatbot Xylitol is prothrombotic and associated with cardiovascular risk ”的<span style="color: black;">科研</span>论文,旨在<strong style="color: blue;">探索大型语言模型ChatGPT在预测从高眼压症(OHT)发展为青光眼方面的能力。</strong></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q2.itc.cn/images01/20240625/6ce4544a3ee74c6495089e9375ac6921.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">(<strong style="color: blue;">关注“医学论文与统计分析”公z号</strong>回复<strong style="color: blue;">“ 原文”</strong><span style="color: black;">就可</span><span style="color: black;">得到</span>文献PDF等资料)</p>
    <h1 style="color: black; text-align: left; margin-bottom: 10px;"><strong style="color: blue;"> 怎么用ChatGPT发一区</strong></h1>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">什么是<strong style="color: blue;">青光眼?</strong><span style="color: black;">做为</span>一种以病理性眼压<span style="color: black;">上升</span>,<span style="color: black;">导致</span>视神经萎缩和视野<span style="color: black;">缺失</span>为<span style="color: black;">一起</span>特征的眼睛<span style="color: black;">疾患</span>,青光眼有5%~20%<span style="color: black;">上下</span>的概率会<span style="color: black;">引起</span><span style="color: black;">病人</span><span style="color: black;">眼瞎</span>。预防青光眼的关键<span style="color: black;">便是</span>早<span style="color: black;">发掘</span>早治疗,但<span style="color: black;">日前</span>的诊断和预测手段仍存在局限性,<span style="color: black;">不可</span>做出更准确地判断。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">在探索青光眼发展预测模型的过程中,不少<span style="color: black;">科研</span><span style="color: black;">已然</span>利用人工智能(AI)和深度学习(DL)取得了可喜的成果。但这些模型的运行需要特殊的<span style="color: black;">要求</span>,并不适宜大范围开展。而在过去几个月,更新后更稳定、速度更快的ChatGPT4.0<span style="color: black;">始终</span><span style="color: black;">处在</span><span style="color: black;">专题</span>的中心。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">既然人工智能(AI)<span style="color: black;">能够</span>纳入预测模型,<span style="color: black;">那样</span>将性能更稳定的ChatGPT4.0纳入预测模型,会不会有更好的结果?</strong></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">基于先前的结果,<span style="color: black;">科研</span>团队开展了这项<span style="color: black;">科研</span>。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">该项<span style="color: black;">科研</span><span style="color: black;">经过</span><strong style="color: blue;">回顾性病例对照<span style="color: black;">科研</span>设计</strong>,纳入了<strong style="color: blue;">OHTS(高眼压症治疗<span style="color: black;">科研</span>)数据集</strong>中<strong style="color: blue;">1504名受试者的3008只眼睛</strong>,并收集了受试者在青光眼发病前一年内的<span style="color: black;">关联</span>参数。</p><strong style="color: blue;"><span style="color: black;">关联</span>参数</strong><span style="color: black;">包含</span>:青光眼<span style="color: black;">出现</span>前一年的人口统计学、临床、眼部、视神经头和视野(VF)参数。<span style="color: black;">按照</span><strong style="color: blue;">青光眼视神经病变(GON)或VF<span style="color: black;">反常</span></strong>(校正模式标准差&lt;0.05或青光眼半视野测试超出正常范围)的任何体征标记为<strong style="color: blue;">原发性开角型青光眼(POAG)</strong>,以先满足的标准为准。所有的眼睛在青光眼发病前被标记为“POAG未转化”或“POAG转化”。最后,<span style="color: black;">倘若</span>受试者<span style="color: black;">最少</span>有一只眼睛被标记为POAG,则被标记为“转换POAG”,否则被标记为“未转换POAG”。随后,基于所有参与者的双眼将表格参数转换为<strong style="color: blue;">文本格式</strong>来<span style="color: black;">研发</span><span style="color: black;">查找</span>,并<strong style="color: blue;"><span style="color: black;">经过</span>ChatGPT接口自动进行预测。</strong>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">重点</span><span style="color: black;">评估</span>指标:</strong>准确性、ROC曲线下面积(AUC)、<span style="color: black;">敏锐</span>性、特异性和加权F1得分。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">(<strong style="color: blue;">关注“医学论文与统计分析”公z号</strong>回复<strong style="color: blue;">“ 原文”</strong><span style="color: black;">就可</span><span style="color: black;">得到</span>文献PDF等资料)</p>
    <h1 style="color: black; text-align: left; margin-bottom: 10px;"><strong style="color: blue;"> ChatGPT有<span style="color: black;">必定</span>的预测效果 </strong></h1>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">在后续开展的<span style="color: black;">实验</span>中,<span style="color: black;">科研</span>团队<span style="color: black;">发掘</span><strong style="color: blue;">ChatGPT在预测青光眼发病方面展现出较好的性能,其中ChatGPT4.0的预测性能优于ChatGPT3.5。</strong></p><span style="color: black;">详细</span><span style="color: black;">来讲</span>,<strong style="color: blue;">ChatGPT4.0在预测青光眼发病前一年的准确性为75%,AUC为0.67,<span style="color: black;">敏锐</span>性为56%,特异性为78%,加权F1得分为0.77;</strong>相比之下,<strong style="color: blue;">ChatGPT3.5的表现较差,其准确性为61%,AUC为0.62,<span style="color: black;">敏锐</span>性为64%,特异性为59%,加权F1得分为0.63。</strong>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q4.itc.cn/images01/20240625/4caa3a1f46284aa4937d3e54ec8d6575.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">(<strong style="color: blue;">关注“医学论文与统计分析”公z号</strong>回复<strong style="color: blue;">“ 原文”</strong><span style="color: black;">就可</span><span style="color: black;">得到</span>文献PDF等资料)</p>
    <h1 style="color: black; text-align: left; margin-bottom: 10px;"><strong style="color: blue;"> ChatGPT能取代<span style="color: black;">大夫</span>吗?</strong></h1>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">本文的创新点显而易见,首次将ChatGPT应用于<span style="color: black;">疾患</span>的预测,展现其在医学<span style="color: black;">行业</span>强大的<span style="color: black;">潜能</span>。<span style="color: black;">同期</span>,<span style="color: black;">因为</span>ChatGPT<span style="color: black;">拥有</span>强大的学习能力,<span style="color: black;">能够</span><span style="color: black;">经过</span>整合现有青光眼数据集的多种模式,对ChatGPT进行训练,<span style="color: black;">研发</span>青光眼<span style="color: black;">行业</span>特定的聊天<span style="color: black;">设备</span>人,以期得到更准确地预测结果。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">然则</span>这篇<span style="color: black;">文案</span>的结果<span style="color: black;">吗</span>,就马马虎虎了,AUC面积<span style="color: black;">亦</span>不大,<span style="color: black;">无</span><span style="color: black;">通常</span>应用临床的0.70以上,GPT 4.0还凑合。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">但随着<span style="color: black;">更加多</span><span style="color: black;">繁杂</span>大模型的介入,<span style="color: black;">例如</span>GPT4o <span style="color: black;">或</span>后面GPT5的<span style="color: black;">面世</span>,<span style="color: black;">亦</span>许效果会更好、</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">不外</span>问题来了,它能应用吗?</strong></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">从预测模型<span style="color: black;">方向</span><span style="color: black;">来讲</span>,它很难做到可视化,<span style="color: black;">亦</span>很难变成一个工具<span style="color: black;">或</span>一个临床模型。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">不外</span>从<span style="color: black;">病人</span><span style="color: black;">方向</span>,自我诊断与预判的<span style="color: black;">方向</span>,<span style="color: black;">拥有</span><span style="color: black;">必定</span>的用途,至于从医护<span style="color: black;">方向</span>,在临床应用上,还远着呢。当然,今后就难说了</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">ChatGPT刚更新的时候,各行各业的人都在担心自己的工作被取代。但老郑我想说的是,能取代你的从来不是ChatGPT,而是会用ChatGPT的人!</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">(<strong style="color: blue;">关注“医学论文与统计分析”公z号</strong>回复<strong style="color: blue;">“ 原文”</strong><span style="color: black;">就可</span><span style="color: black;">得到</span>文献PDF等资料)</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q4.itc.cn/q_70/images01/20240613/e77fc27e9f074079a4ac1308d6c9ce69.png" style="width: 50%; margin-bottom: 20px;"></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q7.itc.cn/q_70/images01/20240621/77920ea8cb2140779d8586a484de4342.png" style="width: 50%; margin-bottom: 20px;"></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">详情请点击右方:</strong><a style="color: black;"><strong style="color: blue;">Nhanes一对一高级学习班,不做实验<span style="color: black;">发布</span>SCI论文</strong></a><a style="color: black;"><span style="color: black;">返回<span style="color: black;">外链论坛:www.fok120.com</span>,查看<span style="color: black;">更加多</span></span></a></p>

    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">责任编辑:网友投稿</span></p>




4zhvml8 发表于 2024-10-13 11:22:27

感谢你的精彩评论,为我的思绪打开了新的窗口。
页: [1]
查看完整版本: 用ChatGPT直接来构建预测模型,效果怎么样?有作者用它写了一篇SCI论文