{"id":5201,"date":"2024-11-28T14:40:45","date_gmt":"2024-11-28T06:40:45","guid":{"rendered":"https:\/\/nullthought.net\/?p=5201"},"modified":"2024-11-28T15:51:01","modified_gmt":"2024-11-28T07:51:01","slug":"%e5%a4%9a%e7%a7%8d%e5%a4%a7%e8%af%ad%e8%a8%80%e6%a8%a1%e5%9e%8b%ef%bc%88llms%ef%bc%89%e5%9c%a8%e7%a3%81%e5%85%b1%e6%8c%af%e6%88%90%e5%83%8f%ef%bc%88mri%ef%bc%89%e6%8a%80%e6%9c%af%e9%97%ae%e9%a2%98","status":"publish","type":"post","link":"https:\/\/nullthought.net\/?p=5201","title":{"rendered":"\u591a\u79cd\u5927\u8bed\u8a00\u6a21\u578b\uff08LLMs\uff09\u5728\u78c1\u5171\u632f\u6210\u50cf\uff08MRI\uff09\u6280\u672f\u95ee\u9898\u56de\u7b54\u4e2d\u7684\u8868\u73b0"},"content":{"rendered":"\n<p>\u8bba\u6587<strong><a href=\"https:\/\/arxiv.org\/abs\/2411.12238\" target=\"_blank\" rel=\"noreferrer noopener\">Performance of Large Language Models in Technical MRI Question Answering: A Comparative Study<\/a><\/strong>\u8be6\u7ec6\u8bc4\u4f30\u4e86\u591a\u79cd\u5927\u8bed\u8a00\u6a21\u578b\uff08LLMs\uff09\u5728\u78c1\u5171\u632f\u6210\u50cf\uff08MRI\uff09\u6280\u672f\u95ee\u9898\u56de\u7b54\u4e2d\u7684\u8868\u73b0\uff0c\u4ece\u6570\u636e\u6765\u6e90\u3001\u6a21\u578b\u9009\u62e9\u3001\u5b9e\u9a8c\u65b9\u6cd5\u3001\u7ed3\u679c\u5206\u6790\u5230\u8ba8\u8bba\u548c\u672a\u6765\u5c55\u671b\uff0c\u90fd\u8fdb\u884c\u4e86\u7cfb\u7edf\u6027\u5206\u6790\u3002<\/p>\n\n\n\n<p>\u8bba\u6587\u4f5c\u8005\u4e3aAlan B McMillan\uff0c\u6765\u81eaUniversity of Wisconsin\u3002<\/p>\n\n\n\n<h4 class=\"wp-block-heading\">\u4e00\u3001<strong>\u80cc\u666f<\/strong><\/h4>\n\n\n\n<p>\u78c1\u5171\u632f\u6210\u50cf\uff08MRI\uff09\u4f5c\u4e3a\u73b0\u4ee3\u533b\u5b66\u5f71\u50cf\u5b66\u7684\u91cd\u8981\u7ec4\u6210\u90e8\u5206\uff0c\u5176\u56fe\u50cf\u8d28\u91cf\u76f4\u63a5\u4f9d\u8d56\u64cd\u4f5c\u4eba\u5458\u7684\u6280\u672f\u6c34\u5e73\u3002\u7136\u800c\uff0c\u5728\u5730\u7406\u504f\u8fdc\u5730\u533a\u6216\u8d44\u6e90\u53d7\u9650\u7684\u533b\u7597\u73af\u5883\u4e2d\uff0c\u7531\u4e8e\u7f3a\u4e4f\u7ecf\u9a8c\u4e30\u5bcc\u7684MRI\u6280\u5e08\u6216\u653e\u5c04\u79d1\u533b\u751f\uff0c\u8fd9\u79cd\u6280\u672f\u6c34\u5e73\u5e38\u5e38\u5b58\u5728\u663e\u8457\u5dee\u5f02\u3002\u8fd9\u79cd\u5dee\u5f02\u4e0d\u4ec5\u5f71\u54cd\u56fe\u50cf\u8d28\u91cf\uff0c\u8fd8\u53ef\u80fd\u5bfc\u81f4\u4ee5\u4e0b\u95ee\u9898\uff1a<\/p>\n\n\n\n<ol class=\"wp-block-list\">\n<li><strong>\u8bca\u65ad\u9519\u8bef<\/strong>\uff1a\u4f8b\u5982\uff0c\u7531\u4e8e\u5206\u8fa8\u7387\u4f4e\u5bfc\u81f4\u5c0f\u80bf\u7624\u6216\u8840\u7ba1\u5f02\u5e38\u672a\u88ab\u53d1\u73b0\u3002<\/li>\n\n\n\n<li><strong>\u5ef6\u8bef\u6cbb\u7597<\/strong>\uff1a\u6280\u672f\u4e0d\u5f53\u53ef\u80fd\u5bfc\u81f4\u91cd\u590d\u68c0\u67e5\uff0c\u5ef6\u8bef\u60a3\u8005\u7684\u8bca\u6cbb\u3002<\/li>\n\n\n\n<li><strong>\u7814\u7a76\u504f\u5dee<\/strong>\uff1a\u5728\u7eb5\u5411\u7814\u7a76\u6216\u7597\u6548\u8bc4\u4f30\u4e2d\uff0c\u56fe\u50cf\u4e0d\u4e00\u81f4\u4f1a\u5bf9\u7ed3\u679c\u4ea7\u751f\u5e72\u6270\u3002<\/li>\n<\/ol>\n\n\n\n<p>\u4e3a\u4e86\u51cf\u5c11\u8fd9\u4e9b\u5dee\u5f02\uff0c\u4f20\u7edf\u65b9\u6cd5\u4f9d\u8d56\u4e8e\u5e7f\u6cdb\u7684\u6280\u5e08\u57f9\u8bad\u548c\u6807\u51c6\u5316\u7684\u6210\u50cf\u534f\u8bae\u3002\u7136\u800c\uff0c\u8fd9\u4e9b\u65b9\u6cd5\u5728\u8d44\u6e90\u6709\u9650\u7684\u73af\u5883\u4e2d\u96be\u4ee5\u5168\u9762\u63a8\u5e7f\u3002\u56e0\u6b64\uff0c\u7814\u7a76\u4eba\u5de5\u667a\u80fd\uff0c\u7279\u522b\u662f\u5927\u8bed\u8a00\u6a21\u578b\uff08LLMs\uff09\u7684\u5e94\u7528\uff0c\u6210\u4e3a\u4e00\u79cd\u6f5c\u5728\u7684\u89e3\u51b3\u65b9\u6848\u3002\u8fd9\u4e9b\u6a21\u578b\u56e0\u5176\u5f3a\u5927\u7684\u81ea\u7136\u8bed\u8a00\u5904\u7406\u80fd\u529b\uff0c\u6709\u671b\u4e3aMRI\u64cd\u4f5c\u63d0\u4f9b\u5b9e\u65f6\u6280\u672f\u6307\u5bfc\uff0c\u5f25\u5408\u6280\u672f\u6c34\u5e73\u5dee\u8ddd\uff0c\u6539\u5584\u4e34\u5e8a\u5b9e\u8df5\u7684\u4e00\u81f4\u6027\u3002<\/p>\n\n\n\n<p>\u8bba\u6587\u660e\u786e\u63d0\u51fa\u4e86\u4e24\u5927\u6838\u5fc3\u95ee\u9898\uff1a<\/p>\n\n\n\n<ol class=\"wp-block-list\">\n<li><strong>LLMs\u662f\u5426\u80fd\u591f\u51c6\u786e\u56de\u7b54MRI\u6280\u672f\u76f8\u5173\u95ee\u9898\uff1f<\/strong><\/li>\n\n\n\n<li><strong>\u66f4\u5927\u3001\u66f4\u5148\u8fdb\u7684\u6a21\u578b\u662f\u5426\u5728\u56de\u7b54\u6df1\u5ea6\u548c\u51c6\u786e\u6027\u4e0a\u4f18\u4e8e\u5c0f\u578b\u6a21\u578b\uff1f<\/strong><\/li>\n<\/ol>\n\n\n\n<h4 class=\"wp-block-heading\">\u4e8c\u3001<strong>\u7814\u7a76\u76ee\u6807<\/strong><\/h4>\n\n\n\n<ol class=\"wp-block-list\">\n<li><strong>\u5168\u9762\u8bc4\u4f30<\/strong>\uff1a\u7cfb\u7edf\u6bd4\u8f83\u591a\u79cd\u5927\u8bed\u8a00\u6a21\u578b\u5728MRI\u6280\u672f\u95ee\u9898\u56de\u7b54\u4e2d\u7684\u8868\u73b0\uff0c\u91cf\u5316\u5176\u51c6\u786e\u6027\u3002<\/li>\n\n\n\n<li><strong>\u63a2\u7d22\u5dee\u5f02<\/strong>\uff1a\u5206\u6790\u6a21\u578b\u5728\u4e0d\u540cMRI\u4e3b\u9898\uff08\u5982\u57fa\u7840\u539f\u7406\u3001\u4f2a\u5f71\u4fee\u6b63\u3001\u56fe\u50cf\u52a0\u6743\u7b49\uff09\u4e0a\u7684\u8868\u73b0\u5dee\u5f02\u3002<\/li>\n\n\n\n<li><strong>\u8bc4\u4f30\u6f5c\u529b<\/strong>\uff1a\u63a2\u8ba8\u8fd9\u4e9b\u6a21\u578b\u5728\u6807\u51c6\u5316MRI\u5b9e\u8df5\u548c\u4e34\u5e8a\u5b9e\u65f6\u652f\u6301\u4e2d\u7684\u6f5c\u5728\u5e94\u7528\u4ef7\u503c\u3002<\/li>\n\n\n\n<li><strong>\u660e\u786e\u4e0d\u8db3<\/strong>\uff1a\u8bc6\u522b\u5f53\u524d\u6a21\u578b\u5728\u533b\u5b66\u6210\u50cf\u9886\u57df\u7684\u6280\u672f\u77ed\u677f\uff0c\u4e3a\u672a\u6765\u6539\u8fdb\u63d0\u4f9b\u53c2\u8003\u3002<\/li>\n<\/ol>\n\n\n\n<h4 class=\"wp-block-heading\">\u4e09\u3001<strong>\u7814\u7a76\u65b9\u6cd5<\/strong><\/h4>\n\n\n\n<h5 class=\"wp-block-heading\">1. <strong>\u6570\u636e\u51c6\u5907<\/strong><\/h5>\n\n\n\n<p>\u6570\u636e\u6765\u6e90\u4e8e\u300aThe MRI Study Guide for Technologists\u300b\uff0c\u8fd9\u662f\u4e00\u672c\u7528\u4e8eMRI\u6280\u672f\u5458\u8ba4\u8bc1\u8003\u8bd5\u7684\u6743\u5a01\u6559\u6750\uff0c\u95ee\u9898\u5185\u5bb9\u5168\u9762\u8986\u76d6MRI\u64cd\u4f5c\u7684\u6838\u5fc3\u9886\u57df\u3002\u7ecf\u8fc7\u7b5b\u9009\u548c\u5206\u7c7b\u540e\uff0c\u6700\u7ec8\u4fdd\u7559\u4e86570\u9053\u95ee\u9898\uff0c\u5e76\u5206\u4e3a\u4ee5\u4e0b\u4e5d\u5927\u4e3b\u9898\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u5386\u53f2\uff08History\uff0c24\u9898\uff09<\/strong>\uff1a\u6db5\u76d6MRI\u6280\u672f\u53d1\u5c55\u7684\u91cc\u7a0b\u7891\u548c\u5386\u53f2\u4e8b\u4ef6\u3002<\/li>\n\n\n\n<li><strong>\u57fa\u7840\u539f\u7406\uff08Basic Principles\uff0c64\u9898\uff09<\/strong>\uff1a\u6d89\u53caMRI\u7684\u7269\u7406\u57fa\u7840\uff0c\u5305\u62ec\u78c1\u573a\u3001\u5c04\u9891\u8109\u51b2\u53ca\u539f\u5b50\u6838\u6210\u50cf\u539f\u7406\u3002<\/li>\n\n\n\n<li><strong>\u56fe\u50cf\u52a0\u6743\u4e0e\u5bf9\u6bd4\uff08Image Weighting and Contrast\uff0c59\u9898\uff09<\/strong>\uff1a\u8ba8\u8bba\u56fe\u50cf\u5bf9\u6bd4\u5ea6\u7684\u5f62\u6210\u673a\u5236\u3002<\/li>\n\n\n\n<li><strong>\u56fe\u50cf\u751f\u4ea7\uff08Image Production\uff0c115\u9898\uff09<\/strong>\uff1a\u8be6\u7ec6\u63cf\u8ff0\u4e86\u751f\u6210MRI\u56fe\u50cf\u7684\u8fc7\u7a0b\u3002<\/li>\n\n\n\n<li><strong>\u8109\u51b2\u5e8f\u5217\uff08Pulse Sequences\uff0c41\u9898\uff09<\/strong>\uff1a\u5206\u6790\u4e0d\u540c\u8109\u51b2\u5e8f\u5217\u7684\u7279\u70b9\u53ca\u5e94\u7528\u3002<\/li>\n\n\n\n<li><strong>\u4f2a\u5f71\u4e0e\u4fee\u6b63\uff08Artifacts and Corrections\uff0c55\u9898\uff09<\/strong>\uff1a\u7814\u7a76\u5e38\u89c1\u6210\u50cf\u4f2a\u5f71\u53ca\u5176\u89e3\u51b3\u65b9\u6848\u3002<\/li>\n\n\n\n<li><strong>\u8840\u6d41\/\u5fc3\u810f\u6210\u50cf\uff08Flow\/Cardiac Imaging\uff0c82\u9898\uff09<\/strong>\uff1a\u4e13\u6ce8\u4e8e\u5fc3\u8840\u7ba1\u6210\u50cf\u7684\u7279\u6b8a\u6280\u672f\u3002<\/li>\n\n\n\n<li><strong>\u4eea\u5668\u64cd\u4f5c\uff08Instrumentation\uff0c56\u9898\uff09<\/strong>\uff1a\u63cf\u8ff0MRI\u8bbe\u5907\u7684\u6280\u672f\u6784\u6210\u53ca\u5176\u64cd\u4f5c\u3002<\/li>\n\n\n\n<li><strong>\u5b89\u5168\uff08Safety\uff0c74\u9898\uff09<\/strong>\uff1a\u5f3a\u8c03\u64cd\u4f5c\u89c4\u8303\u53ca\u60a3\u8005\/\u6280\u5e08\u5b89\u5168\u3002<\/li>\n<\/ul>\n\n\n\n<p>\u8fd9\u4e9b\u95ee\u9898\u88ab\u8bbe\u8ba1\u4e3a\u7eaf\u6587\u672c\u5f62\u5f0f\uff0c\u4ee5\u907f\u514d\u56e0\u9700\u8981\u89c6\u89c9\u53c2\u8003\u800c\u5f15\u5165\u7684\u504f\u5dee\u3002<\/p>\n\n\n\n<h5 class=\"wp-block-heading\">2. <strong>\u6a21\u578b\u9009\u62e9<\/strong><\/h5>\n\n\n\n<p>\u8bba\u6587\u6d4b\u8bd5\u4e86\u591a\u79cd\u5c01\u95ed\u6e90\u548c\u5f00\u6e90\u6a21\u578b\uff0c\u8fd9\u4e9b\u6a21\u578b\u6db5\u76d6\u4ece\u5c0f\u578b\u5230\u8d85\u5927\u578b\u7684\u53c2\u6570\u8303\u56f4\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u5c01\u95ed\u6e90\u6a21\u578b<\/strong>\uff1a\n<ul class=\"wp-block-list\">\n<li>OpenAI \u7684 GPT-4 Turbo\u3001GPT-4o \u548c o1 \u7cfb\u5217\uff08\u5305\u62ec o1 Mini \u548c o1 Preview\uff09\u3002<\/li>\n\n\n\n<li>Anthropic \u7684 Claude \u7cfb\u5217\uff08\u5982 Claude 3.5 Haiku\uff09\u3002<\/li>\n\n\n\n<li>Google \u7684 Gemini \u7cfb\u5217\u3002<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>\u5f00\u6e90\u6a21\u578b<\/strong>\uff1a\n<ul class=\"wp-block-list\">\n<li>Microsoft \u7684 Phi 3.5 Mini\u3002<\/li>\n\n\n\n<li>Meta \u7684 Llama 3.1\u3002<\/li>\n\n\n\n<li>Hugging Face \u7684 smolLM2\u3002<\/li>\n\n\n\n<li>\u5176\u4ed6\u5982 Mistral \u7cfb\u5217\u548c Falcon 2\u3002<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<p>\u8fd9\u4e9b\u6a21\u578b\u4ee3\u8868\u4e86\u4e0d\u540c\u7684\u8bad\u7ec3\u65b9\u6cd5\u3001\u53c2\u6570\u89c4\u6a21\u548c\u6027\u80fd\u76ee\u6807\uff0c\u4e3a\u5168\u9762\u5206\u6790\u63d0\u4f9b\u4e86\u591a\u6837\u6027\u3002<\/p>\n\n\n\n<h5 class=\"wp-block-heading\">3. <strong>\u5b9e\u9a8c\u6d41\u7a0b<\/strong><\/h5>\n\n\n\n<ol class=\"wp-block-list\">\n<li><strong>\u95ee\u9898\u63d0\u4ea4<\/strong>\uff1a\u91c7\u7528 LangChain \u6846\u67b6\u7edf\u4e00\u683c\u5f0f\u5316\u95ee\u9898\u5e76\u5411\u6a21\u578b\u63d0\u4ea4\uff0c\u786e\u4fdd\u4e0d\u540c\u6a21\u578b\u7684\u5b9e\u9a8c\u6761\u4ef6\u4e00\u81f4\u3002<\/li>\n\n\n\n<li><strong>\u56de\u7b54\u8bc4\u4f30<\/strong>\uff1a\u5229\u7528\u81ea\u52a8\u8bc4\u5206\u534f\u8bae\u5bf9\u56de\u7b54\u8fdb\u884c\u8bc4\u4f30\u3002\u8bc4\u5206\u6d41\u7a0b\u5305\u62ec\uff1a\n<ul class=\"wp-block-list\">\n<li>\u5982\u679c\u6a21\u578b\u7684\u56de\u7b54\u4e3a\u7b80\u5355\u7684\u5b57\u6bcd\u6216\u77ed\u8bed\uff0c\u76f4\u63a5\u5339\u914d\u6b63\u786e\u7b54\u6848\u3002<\/li>\n\n\n\n<li>\u5bf9\u8f83\u957f\u56de\u7b54\uff0c\u4f7f\u7528 Levenshtein \u8ddd\u79bb\u8ba1\u7b97\u6587\u672c\u76f8\u4f3c\u5ea6\uff0c\u4ee5\u5224\u65ad\u662f\u5426\u4e0e\u6b63\u786e\u7b54\u6848\u8bed\u4e49\u4e00\u81f4\u3002<\/li>\n\n\n\n<li>\u5728\u8bed\u4e49\u5339\u914d\u5931\u8d25\u65f6\uff0c\u91c7\u7528\u6a21\u7cca\u5339\u914d\u6280\u672f\u63d0\u9ad8\u8bc4\u5206\u7cbe\u5ea6\u3002<\/li>\n<\/ul>\n<\/li>\n<\/ol>\n\n\n\n<p>\u8bc4\u5206\u7ed3\u679c\u4ee5\u6a21\u578b\u7684\u51c6\u786e\u7387\u8868\u793a\uff0c\u5373\u6b63\u786e\u56de\u7b54\u7684\u6bd4\u4f8b\u3002<\/p>\n\n\n\n<ol start=\"3\" class=\"wp-block-list\">\n<li><strong>\u57fa\u51c6\u8bbe\u7f6e<\/strong>\uff1a\u968f\u673a\u731c\u6d4b\u7684\u51c6\u786e\u7387\u57fa\u51c6\u503c\u4e3a26.5%\uff0c\u7528\u4e8e\u5bf9\u6bd4\u6a21\u578b\u8868\u73b0\u3002<\/li>\n<\/ol>\n\n\n\n<h4 class=\"wp-block-heading\">\u56db\u3001<strong>\u7814\u7a76\u7ed3\u679c<\/strong><\/h4>\n\n\n\n<h5 class=\"wp-block-heading\">1. <strong>\u6574\u4f53\u8868\u73b0<\/strong><\/h5>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u6700\u9ad8\u51c6\u786e\u7387<\/strong>\uff1aOpenAI \u7684 o1 Preview \u8fbe\u523094%\uff0c\u663e\u8457\u4f18\u4e8e\u5176\u4ed6\u6a21\u578b\u3002<\/li>\n\n\n\n<li><strong>\u5176\u4ed6\u8868\u73b0\u4f18\u79c0\u7684\u5c01\u95ed\u6e90\u6a21\u578b<\/strong>\uff1aGPT-4o \u548c o1 Mini\uff0888%\uff09\uff0cClaude 3.5 Haiku \u548c GPT-4 Turbo\uff0884%\uff09\u3002<\/li>\n\n\n\n<li><strong>\u5f00\u6e90\u6a21\u578b\u4e2d\u8868\u73b0\u6700\u4f73<\/strong>\uff1aMicrosoft \u7684 Phi 3.5 Mini \u8fbe\u523078%\uff0c\u63a5\u8fd1\u5c01\u95ed\u6e90\u6a21\u578b\u7684\u6027\u80fd\u3002<\/li>\n\n\n\n<li><strong>\u5c0f\u578b\u6a21\u578b\u7684\u6f5c\u529b<\/strong>\uff1asmolLM2\uff0869%\uff09\u8868\u73b0\u4f18\u4e8e\u9884\u671f\uff0c\u4e0e\u90e8\u5206\u5927\u578b\u5c01\u95ed\u6e90\u6a21\u578b\u76f8\u8fd1\u3002<\/li>\n<\/ul>\n\n\n\n<h5 class=\"wp-block-heading\">2. <strong>\u4e13\u9898\u8868\u73b0<\/strong><\/h5>\n\n\n\n<p>\u4e0d\u540c\u6a21\u578b\u5728\u4e5d\u5927MRI\u4e3b\u9898\u4e2d\u7684\u8868\u73b0\u5dee\u5f02\u663e\u8457\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u8868\u73b0\u6700\u4f73\u7684\u4e3b\u9898<\/strong>\uff1a\n<ul class=\"wp-block-list\">\n<li>\u57fa\u7840\u539f\u7406\uff0897%\uff09\u548c\u4eea\u5668\u64cd\u4f5c\uff0896%\uff09\uff0c\u8bf4\u660e\u6a21\u578b\u5bf9\u7269\u7406\u57fa\u7840\u548c\u786c\u4ef6\u77e5\u8bc6\u638c\u63e1\u8f83\u597d\u3002<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>\u8868\u73b0\u6700\u5f31\u7684\u4e3b\u9898<\/strong>\uff1a\n<ul class=\"wp-block-list\">\n<li>\u56fe\u50cf\u52a0\u6743\u4e0e\u5bf9\u6bd4\uff0881%\uff09\u548c\u4f2a\u5f71\u4e0e\u4fee\u6b63\uff0878%\uff09\uff0c\u53cd\u6620\u51fa\u6a21\u578b\u5728\u590d\u6742\u6210\u50cf\u539f\u7406\u548c\u4f2a\u5f71\u5904\u7406\u4e0a\u7684\u7406\u89e3\u4ecd\u663e\u4e0d\u8db3\u3002<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h5 class=\"wp-block-heading\">3. <strong>\u5f00\u6e90\u4e0e\u5c01\u95ed\u6e90\u5bf9\u6bd4<\/strong><\/h5>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u5c01\u95ed\u6e90\u6a21\u578b\u56e0\u5176\u66f4\u5927\u7684\u53c2\u6570\u89c4\u6a21\u548c\u66f4\u5e7f\u6cdb\u7684\u8bad\u7ec3\u6570\u636e\uff0c\u5728\u6574\u4f53\u51c6\u786e\u7387\u4e0a\u666e\u904d\u4f18\u4e8e\u5f00\u6e90\u6a21\u578b\u3002<\/li>\n\n\n\n<li>Phi 3.5 Mini \u662f\u5f00\u6e90\u6a21\u578b\u4e2d\u7684\u4eae\u70b9\uff0c\u8868\u73b0\u63a5\u8fd1\u90e8\u5206\u5c01\u95ed\u6e90\u6a21\u578b\u3002<\/li>\n<\/ul>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h4 class=\"wp-block-heading\">\u4e94\u3001<strong>\u8ba8\u8bba\u4e0e\u610f\u4e49<\/strong><\/h4>\n\n\n\n<h5 class=\"wp-block-heading\">\u4e00\uff09\u3001<strong>LLMs\u7684\u4f18\u52bf<\/strong><\/h5>\n\n\n\n<ol class=\"wp-block-list\">\n<li><strong>\u6280\u672f\u652f\u6301\u5de5\u5177<\/strong>\uff1a\u9ad8\u6027\u80fd\u6a21\u578b\u53ef\u4f5c\u4e3aMRI\u6280\u672f\u5458\u7684\u5b9e\u65f6\u53c2\u8003\uff0c\u89e3\u51b3\u64cd\u4f5c\u96be\u9898\u3002<\/li>\n\n\n\n<li><strong>\u6807\u51c6\u5316\u5b9e\u8df5<\/strong>\uff1a\u901a\u8fc7\u63d0\u4f9b\u4e00\u81f4\u7684\u6280\u672f\u6307\u5bfc\uff0c\u964d\u4f4e\u5730\u57df\u5dee\u5f02\uff0c\u6539\u5584\u6210\u50cf\u8d28\u91cf\u3002<\/li>\n\n\n\n<li><strong>\u6559\u80b2\u4ef7\u503c<\/strong>\uff1a\u5e2e\u52a9\u65b0\u624b\u6280\u5e08\u5feb\u901f\u638c\u63e1\u6280\u672f\u8981\u70b9\uff0c\u540c\u65f6\u4e3a\u8d44\u6df1\u6280\u5e08\u63d0\u4f9b\u6301\u7eed\u5b66\u4e60\u7684\u652f\u6301\u3002<\/li>\n<\/ol>\n\n\n\n<h5 class=\"wp-block-heading\">\u4e8c\uff09\u3001<strong>\u6a21\u578b\u7684\u5c40\u9650\u6027<\/strong><\/h5>\n\n\n\n<ol class=\"wp-block-list\">\n<li><strong>\u6570\u636e\u96c6\u5355\u4e00<\/strong>\uff1a\u53ea\u6d4b\u8bd5\u4e86\u6559\u6750\u4e2d\u7684\u95ee\u9898\uff0c\u53ef\u80fd\u65e0\u6cd5\u6db5\u76d6MRI\u64cd\u4f5c\u7684\u590d\u6742\u6027\u3002<\/li>\n\n\n\n<li><strong>\u7f3a\u4e4f\u900f\u660e\u6027<\/strong>\uff1a\u5c01\u95ed\u6e90\u6a21\u578b\u7684\u8bad\u7ec3\u6570\u636e\u548c\u65b9\u6cd5\u4e0d\u516c\u5f00\uff0c\u9650\u5236\u4e86\u7ed3\u679c\u7684\u53ef\u89e3\u91ca\u6027\u3002<\/li>\n\n\n\n<li><strong>\u4efb\u52a1\u7c7b\u578b\u6709\u9650<\/strong>\uff1a\u4ec5\u6d4b\u8bd5\u4e86\u9009\u62e9\u9898\uff0c\u672a\u5305\u542b\u5f00\u653e\u6027\u4efb\u52a1\uff0c\u53ef\u80fd\u4f4e\u4f30\u4e86\u6a21\u578b\u7684\u80fd\u529b\u3002<\/li>\n<\/ol>\n\n\n\n<h5 class=\"wp-block-heading\">\u4e09\uff09\u3001<strong>\u6539\u8fdb\u65b9\u5411<\/strong><\/h5>\n\n\n\n<ol class=\"wp-block-list\">\n<li><strong>\u9886\u57df\u5fae\u8c03<\/strong>\uff1a\u5bf9\u5f00\u6e90\u6a21\u578b\u8fdb\u884cMRI\u7279\u5b9a\u6570\u636e\u7684\u5fae\u8c03\uff0c\u53ef\u663e\u8457\u63d0\u9ad8\u5176\u8868\u73b0\u3002<\/li>\n\n\n\n<li><strong>\u591a\u6837\u5316\u95ee\u9898\u6d4b\u8bd5<\/strong>\uff1a\u7eb3\u5165\u66f4\u590d\u6742\u548c\u5f00\u653e\u6027\u7684\u95ee\u9898\u7c7b\u578b\uff0c\u5168\u9762\u8bc4\u4f30\u6a21\u578b\u7684\u80fd\u529b\u3002<\/li>\n\n\n\n<li><strong>\u4e34\u5e8a\u6574\u5408\u7814\u7a76<\/strong>\uff1a\u63a2\u7d22\u6a21\u578b\u5728\u5b9e\u9645\u5de5\u4f5c\u6d41\u7a0b\u4e2d\u7684\u5e94\u7528\uff0c\u8bc4\u4f30\u5176\u5bf9\u60a3\u8005\u62a4\u7406\u548c\u6280\u5e08\u5de5\u4f5c\u7684\u5b9e\u9645\u5f71\u54cd\u3002<\/li>\n<\/ol>\n","protected":false},"excerpt":{"rendered":"<p>\u8bba\u6587Performance of Large Language Models in Technical MRI [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"set","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[8],"tags":[39,45,86],"class_list":["post-5201","post","type-post","status-publish","format-standard","hentry","category-tech","tag-ai","tag-medicine","tag-llm"],"rttpg_featured_image_url":null,"rttpg_author":{"display_name":"NullThought","author_link":"https:\/\/nullthought.net\/?author=1"},"rttpg_comment":0,"rttpg_category":"<a href=\"https:\/\/nullthought.net\/?cat=8\" rel=\"category\">Tech<\/a>","rttpg_excerpt":"\u8bba\u6587Performance of Large Language Models in Technical MRI&hellip;","_links":{"self":[{"href":"https:\/\/nullthought.net\/index.php?rest_route=\/wp\/v2\/posts\/5201","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/nullthought.net\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/nullthought.net\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/nullthought.net\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/nullthought.net\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=5201"}],"version-history":[{"count":1,"href":"https:\/\/nullthought.net\/index.php?rest_route=\/wp\/v2\/posts\/5201\/revisions"}],"predecessor-version":[{"id":5202,"href":"https:\/\/nullthought.net\/index.php?rest_route=\/wp\/v2\/posts\/5201\/revisions\/5202"}],"wp:attachment":[{"href":"https:\/\/nullthought.net\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5201"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/nullthought.net\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5201"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/nullthought.net\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5201"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}