{"id":1056,"date":"2026-05-14T19:16:07","date_gmt":"2026-05-14T11:16:07","guid":{"rendered":"https:\/\/www.eutaboo.com\/index.php\/2026\/05\/14\/%e3%80%8annu-net-0%e5%9f%ba%e7%a1%80%e5%85%a5%e9%97%a8%ef%bc%888%ef%bc%89%ef%bc%9a%e7%90%86%e8%a7%a3-nnu-net-v2-%e5%86%85%e9%83%a8%e6%a1%86%e6%9e%b6%ef%bc%8cfingerprint%e3%80%81plans%e3%80%81preproces\/"},"modified":"2026-05-14T19:16:07","modified_gmt":"2026-05-14T11:16:07","slug":"%e3%80%8annu-net-0%e5%9f%ba%e7%a1%80%e5%85%a5%e9%97%a8%ef%bc%888%ef%bc%89%ef%bc%9a%e7%90%86%e8%a7%a3-nnu-net-v2-%e5%86%85%e9%83%a8%e6%a1%86%e6%9e%b6%ef%bc%8cfingerprint%e3%80%81plans%e3%80%81preproces","status":"publish","type":"post","link":"https:\/\/www.eutaboo.com\/index.php\/2026\/05\/14\/%e3%80%8annu-net-0%e5%9f%ba%e7%a1%80%e5%85%a5%e9%97%a8%ef%bc%888%ef%bc%89%ef%bc%9a%e7%90%86%e8%a7%a3-nnu-net-v2-%e5%86%85%e9%83%a8%e6%a1%86%e6%9e%b6%ef%bc%8cfingerprint%e3%80%81plans%e3%80%81preproces\/","title":{"rendered":"\u300annU-Net 0\u57fa\u7840\u5165\u95e8\uff088\uff09\uff1a\u7406\u89e3 nnU-Net v2 \u5185\u90e8\u6846\u67b6\uff0cfingerprint\u3001plans\u3001preprocessing\u3001trainer\u300b"},"content":{"rendered":"<h2>\u672c\u7bc7\u5b66\u4e60\u76ee\u6807<\/h2>\n<p>\u8fd9\u662f\u300annU-Net 0\u57fa\u7840\u5165\u95e8\u300b\u7cfb\u5217\u7684\u7b2c 8 \u7bc7\u3002\u4ece\u8fd9\u4e00\u7bc7\u5f00\u59cb\uff0c\u6211\u4eec\u4e0d\u518d\u53ea\u505c\u7559\u5728\u201c\u4f1a\u7528\u547d\u4ee4\u201d\uff0c\u800c\u662f\u8fdb\u5165 nnU-Net v2 \u7684\u5185\u90e8\u6846\u67b6\u3002\u7406\u89e3\u8fd9\u4e9b\u6a21\u5757\uff0c\u662f\u540e\u9762\u4fee\u6539 Trainer\u3001loss\u3001augmentation \u548c network architecture \u7684\u524d\u63d0\u3002<\/p>\n<p>\u8bfb\u5b8c\u672c\u6587\uff0c\u4f60\u5e94\u8be5\u80fd\u591f\uff1a<\/p>\n<ol>\n<li>\u8bf4\u6e05\u695a fingerprint\u3001plans\u3001preprocessing\u3001Trainer\u3001Predictor \u5206\u522b\u8d1f\u8d23\u4ec0\u4e48\u3002<\/li>\n<li>\u7406\u89e3 <code>nnUNetPlans.json<\/code> \u4e3a\u4ec0\u4e48\u662f\u8fde\u63a5\u6570\u636e\u5206\u6790\u548c\u8bad\u7ec3\u914d\u7f6e\u7684\u5173\u952e\u6587\u4ef6\u3002<\/li>\n<li>\u77e5\u9053\u54ea\u4e9b\u4fee\u6539\u901a\u5e38\u9700\u8981\u91cd\u8dd1 preprocessing\uff0c\u54ea\u4e9b\u901a\u5e38\u4e0d\u9700\u8981\u3002<\/li>\n<li>\u5efa\u7acb\u4e00\u5f20\u6e90\u7801\u7ea7\u7684 nnU-Net v2 \u6846\u67b6\u5730\u56fe\u3002<\/li>\n<\/ol>\n<h2>1. \u5148\u770b\u603b\u56fe\uff1annU-Net v2 \u5185\u90e8\u6570\u636e\u6d41<\/h2>\n<p>nnU-Net v2 \u7684\u5185\u90e8\u6d41\u7a0b\u53ef\u4ee5\u7406\u89e3\u4e3a\u201c\u6570\u636e\u7279\u5f81 \u2192 \u5b9e\u9a8c\u8ba1\u5212 \u2192 \u9884\u5904\u7406\u6570\u636e \u2192 \u8bad\u7ec3\u5668 \u2192 \u63a8\u7406\u5668\u201d\u3002<\/p>\n<pre><code class=\"language-mermaid\">flowchart TD\n    A[\u539f\u59cb\u6570\u636e DatasetXXX] --> B[Dataset Fingerprint Extractor]\n    B --> C[dataset_fingerprint.json]\n    C --> D[Experiment Planner]\n    D --> E[nnUNetPlans.json]\n    E --> F[Preprocessor]\n    F --> G[\u9884\u5904\u7406\u540e\u7684\u8bad\u7ec3\u6570\u636e]\n    E --> H[nnUNetTrainer]\n    G --> H\n    H --> I[checkpoints \/ validation \/ logs]\n    E --> J[nnUNetPredictor]\n    I --> J\n    J --> K[\u9884\u6d4b\u7ed3\u679c]\n<\/code><\/pre>\n<p>\u8fd9\u5f20\u56fe\u8bf4\u660e\u4e00\u4e2a\u6838\u5fc3\u4e8b\u5b9e\uff1annU-Net v2 \u4e0d\u662f\u76f4\u63a5\u628a\u539f\u56fe\u4e22\u7ed9 Trainer\u3002\u5b83\u4f1a\u5148\u5206\u6790\u6570\u636e\uff0c\u518d\u751f\u6210 plans\uff0c\u7136\u540e\u7528 plans \u6307\u5bfc\u9884\u5904\u7406\u3001\u8bad\u7ec3\u548c\u63a8\u7406\u3002<\/p>\n<h2>2. fingerprint\uff1a\u6570\u636e\u96c6\u7684\u201c\u4f53\u68c0\u62a5\u544a\u201d<\/h2>\n<p><strong>dataset fingerprint<\/strong> \u53ef\u4ee5\u7406\u89e3\u4e3a\u6570\u636e\u96c6\u7684\u4f53\u68c0\u62a5\u544a\u3002\u5b83\u4e0d\u662f\u8bad\u7ec3\u7ed3\u679c\uff0c\u800c\u662f\u5bf9\u8bad\u7ec3\u6570\u636e\u672c\u8eab\u7684\u7edf\u8ba1\u548c\u63cf\u8ff0\uff0c\u4f8b\u5982\uff1a<\/p>\n<ul>\n<li>\u56fe\u50cf\u5c3a\u5bf8\u5206\u5e03\u3002<\/li>\n<li>\u4f53\u7d20 spacing \u5206\u5e03\u3002<\/li>\n<li>\u5f3a\u5ea6\u7edf\u8ba1\u4fe1\u606f\u3002<\/li>\n<li>\u901a\u9053\u548c\u6807\u7b7e\u76f8\u5173\u4fe1\u606f\u3002<\/li>\n<\/ul>\n<p>\u8fd9\u4e9b\u4fe1\u606f\u4f1a\u4fdd\u5b58\u5230 <code>dataset_fingerprint.json<\/code>\u3002\u540e\u7eed planner \u4f1a\u8bfb\u53d6\u5b83\uff0c\u5e76\u6839\u636e\u6570\u636e\u7279\u5f81\u51b3\u5b9a\u9884\u5904\u7406\u548c\u8bad\u7ec3\u914d\u7f6e\u3002<\/p>\n<p>\u4f60\u53ef\u4ee5\u628a fingerprint \u60f3\u6210\u533b\u751f\u68c0\u67e5\u75c5\u4eba\u7684\u57fa\u7840\u6307\u6807\uff1a\u4e0d\u5148\u91cf\u8eab\u9ad8\u4f53\u91cd\u3001\u8840\u538b\u3001\u8840\u5e38\u89c4\uff0c\u5c31\u76f4\u63a5\u5f00\u8bad\u7ec3\u65b9\u6848\uff0c\u98ce\u9669\u5f88\u5927\u3002<\/p>\n<h2>3. plans\uff1annU-Net \u7684\u5b9e\u9a8c\u8ba1\u5212\u4e66<\/h2>\n<p><code>nnUNetPlans.json<\/code> \u662f\u7b2c 8 \u7bc7\u6700\u91cd\u8981\u7684\u6587\u4ef6\u3002\u5b98\u65b9 reference \u8bf4\u660e\uff0cplans \u6587\u4ef6\u5b9a\u4e49\u4e86 nnU-Net \u5982\u4f55\u9884\u5904\u7406\u6570\u636e\uff0c\u4ee5\u53ca\u6bcf\u4e2a configuration \u5982\u4f55\u8bad\u7ec3\u3002<\/p>\n<p>\u5b83\u5305\u542b\u4e24\u7c7b\u4fe1\u606f\uff1a<\/p>\n<ul>\n<li><strong>global dataset-level settings<\/strong>\uff1a\u6570\u636e\u96c6\u7ea7\u522b\u8bbe\u7f6e\uff0c\u4f8b\u5982 reader\/writer\u3001label manager\u3001\u8f6c\u7f6e\u89c4\u5219\u3001\u6570\u636e\u96c6\u540d\u79f0\u3002<\/li>\n<li><strong>configuration-level settings<\/strong>\uff1a\u6bcf\u4e2a\u914d\u7f6e\u81ea\u5df1\u7684\u8bbe\u7f6e\uff0c\u4f8b\u5982 spacing\u3001patch size\u3001batch size\u3001preprocessor\u3001normalization\u3001network architecture\u3002<\/li>\n<\/ul>\n<table>\n<thead>\n<tr>\n<th>plans \u5b57\u6bb5<\/th>\n<th>\u542b\u4e49<\/th>\n<th>\u5f71\u54cd<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><code>spacing<\/code><\/td>\n<td>\u76ee\u6807 spacing<\/td>\n<td>\u5f71\u54cd\u91cd\u91c7\u6837\u548c\u56fe\u50cf\u5c3a\u5ea6<\/td>\n<\/tr>\n<tr>\n<td><code>patch_size<\/code><\/td>\n<td>\u8bad\u7ec3 patch \u5927\u5c0f<\/td>\n<td>\u5f71\u54cd\u4e0a\u4e0b\u6587\u8303\u56f4\u548c\u663e\u5b58<\/td>\n<\/tr>\n<tr>\n<td><code>batch_size<\/code><\/td>\n<td>batch \u5927\u5c0f<\/td>\n<td>\u5f71\u54cd\u8bad\u7ec3\u7a33\u5b9a\u6027\u548c\u663e\u5b58<\/td>\n<\/tr>\n<tr>\n<td><code>preprocessor_name<\/code><\/td>\n<td>\u9884\u5904\u7406\u5668\u540d\u79f0<\/td>\n<td>\u51b3\u5b9a\u9884\u5904\u7406\u903b\u8f91<\/td>\n<\/tr>\n<tr>\n<td><code>normalization_schemes<\/code><\/td>\n<td>\u5404\u901a\u9053\u5f52\u4e00\u5316\u7b56\u7565<\/td>\n<td>\u5f71\u54cd\u8f93\u5165\u5f3a\u5ea6\u5206\u5e03<\/td>\n<\/tr>\n<tr>\n<td><code>network_arch_class_name<\/code><\/td>\n<td>\u7f51\u7edc\u67b6\u6784\u7c7b<\/td>\n<td>\u5f71\u54cd\u6a21\u578b\u7ed3\u6784<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u6240\u4ee5 plans \u4e0d\u662f\u666e\u901a\u65e5\u5fd7\uff0c\u800c\u662f\u540e\u7eed\u8bad\u7ec3\u548c\u63a8\u7406\u5171\u4eab\u7684\u914d\u7f6e\u6765\u6e90\u3002\u968f\u610f\u624b\u6539 plans \u53ef\u80fd\u5bfc\u81f4\u8bad\u7ec3\u3001\u63a8\u7406\u3001\u9884\u5904\u7406\u4e0d\u4e00\u81f4\u3002<\/p>\n<h2>4. preprocessing\uff1a\u628a\u539f\u59cb\u6570\u636e\u53d8\u6210\u8bad\u7ec3\u6570\u636e<\/h2>\n<p><strong>preprocessing<\/strong> \u662f\u9884\u5904\u7406\u3002\u5b83\u8d1f\u8d23\u628a\u539f\u59cb\u533b\u5b66\u56fe\u50cf\u8f6c\u6362\u6210 nnU-Net \u8bad\u7ec3\u65f6\u771f\u6b63\u8bfb\u53d6\u7684\u6570\u636e\u3002\u5e38\u89c1\u5de5\u4f5c\u5305\u62ec\uff1a<\/p>\n<ul>\n<li>\u6309 plans \u4e2d\u7684 spacing \u91cd\u91c7\u6837\u3002<\/li>\n<li>\u6309\u901a\u9053\u5f52\u4e00\u5316\u5f3a\u5ea6\u3002<\/li>\n<li>\u88c1\u526a\u65e0\u6548\u80cc\u666f\u533a\u57df\u3002<\/li>\n<li>\u4fdd\u5b58\u53ef\u5feb\u901f\u8bfb\u53d6\u7684\u4e2d\u95f4\u6570\u636e\u3002<\/li>\n<\/ul>\n<p>\u9884\u5904\u7406\u7ed3\u679c\u901a\u5e38\u653e\u5728 <code>nnUNet_preprocessed\/DatasetXXX_Name<\/code> \u4e0b\u3002\u8bad\u7ec3\u65f6 Trainer \u8bfb\u53d6\u7684\u4e0d\u662f <code>nnUNet_raw<\/code> \u91cc\u7684\u539f\u56fe\uff0c\u800c\u662f\u8fd9\u91cc\u7684\u9884\u5904\u7406\u6570\u636e\u3002<\/p>\n<p>\u8fd9\u4e5f\u89e3\u91ca\u4e86\u4e00\u4e2a\u5e38\u89c1\u95ee\u9898\uff1a\u5982\u679c\u4f60\u6539\u4e86 spacing\u3001normalization \u6216 preprocessor\uff0c\u4f46\u6ca1\u6709\u91cd\u8dd1 preprocessing\uff0c\u8bad\u7ec3\u53ef\u80fd\u4ecd\u7136\u8bfb\u65e7\u6570\u636e\u3002<\/p>\n<h2>5. Trainer\uff1a\u8bad\u7ec3\u884c\u4e3a\u7684\u6838\u5fc3\u5165\u53e3<\/h2>\n<p><strong>Trainer<\/strong> \u662f\u63a7\u5236\u8bad\u7ec3\u6d41\u7a0b\u7684\u7c7b\u3002nnU-Net v2 \u4e2d\u6700\u6838\u5fc3\u7684\u662f <code>nnUNetTrainer<\/code>\u3002\u5b83\u8d1f\u8d23\u628a\u6570\u636e\u52a0\u8f7d\u3001\u7f51\u7edc\u6784\u5efa\u3001loss\u3001optimizer\u3001\u8bad\u7ec3\u5faa\u73af\u3001\u9a8c\u8bc1\u3001checkpoint \u4fdd\u5b58\u7b49\u6b65\u9aa4\u7ec4\u7ec7\u8d77\u6765\u3002<\/p>\n<p>\u4ece\u5b66\u4e60\u8def\u5f84\u770b\uff0cTrainer \u662f\u6700\u91cd\u8981\u7684\u6269\u5c55\u5165\u53e3\u3002\u540e\u9762\u6211\u4eec\u8981\u4fee\u6539 loss\u3001augmentation\u3001optimizer \u6216\u8bad\u7ec3\u6d41\u7a0b\u65f6\uff0c\u901a\u5e38\u4e0d\u662f\u76f4\u63a5\u6539 nnU-Net \u6838\u5fc3\u6587\u4ef6\uff0c\u800c\u662f\u7ee7\u627f <code>nnUNetTrainer<\/code> \u5199\u4e00\u4e2a\u81ea\u5df1\u7684 Trainer \u53d8\u4f53\u3002<\/p>\n<table>\n<thead>\n<tr>\n<th>\u60f3\u4fee\u6539\u4ec0\u4e48<\/th>\n<th>\u901a\u5e38\u4ece\u54ea\u91cc\u5165\u624b<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>loss<\/td>\n<td>\u81ea\u5b9a\u4e49 Trainer \u4e2d\u8986\u76d6 loss \u6784\u5efa\u903b\u8f91<\/td>\n<\/tr>\n<tr>\n<td>optimizer \/ scheduler<\/td>\n<td>\u81ea\u5b9a\u4e49 Trainer \u4e2d\u8986\u76d6\u4f18\u5316\u5668\u76f8\u5173\u65b9\u6cd5<\/td>\n<\/tr>\n<tr>\n<td>augmentation<\/td>\n<td>\u81ea\u5b9a\u4e49 Trainer \u4e2d\u4fee\u6539 transform \u6784\u5efa<\/td>\n<\/tr>\n<tr>\n<td>network architecture<\/td>\n<td>\u8986\u76d6\u6216\u8c03\u6574\u7f51\u7edc\u6784\u5efa\u903b\u8f91\uff0c\u5e76\u4fdd\u8bc1 plans \u517c\u5bb9<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2>6. Predictor\uff1a\u628a\u8bad\u7ec3\u7ed3\u679c\u7528\u4e8e\u65b0\u6570\u636e<\/h2>\n<p><strong>Predictor<\/strong> \u8d1f\u8d23\u63a8\u7406\u6d41\u7a0b\u3002\u5b98\u65b9\u6e90\u7801\u4e2d\uff0c\u63a8\u7406\u76f8\u5173\u903b\u8f91\u96c6\u4e2d\u5728 <code>nnunetv2\/inference\/predict_from_raw_data.py<\/code>\u3002\u5b83\u4f1a\u8bfb\u53d6\u8bad\u7ec3\u4ea7\u751f\u7684\u6a21\u578b\u3001plans \u548c dataset \u4fe1\u606f\uff0c\u5bf9\u8f93\u5165\u56fe\u50cf\u6267\u884c\u4e0e\u8bad\u7ec3\u5339\u914d\u7684\u9884\u5904\u7406\u548c\u9884\u6d4b\u3002<\/p>\n<p>\u8fd9\u5c31\u662f\u4e3a\u4ec0\u4e48\u63a8\u7406\u65f6\u4e0d\u80fd\u53ea\u62ff\u4e00\u4e2a checkpoint \u968f\u4fbf\u5199 PyTorch \u4ee3\u7801\u52a0\u8f7d\u3002nnU-Net \u63a8\u7406\u4e0d\u4ec5\u9700\u8981\u6743\u91cd\uff0c\u8fd8\u9700\u8981\u77e5\u9053\u8bad\u7ec3\u65f6\u7684 plans\u3001normalization\u3001spacing\u3001patch size\u3001\u6ed1\u7a97\u9884\u6d4b\u7b56\u7565\u7b49\u4fe1\u606f\u3002<\/p>\n<h2>7. \u54ea\u4e9b\u4fee\u6539\u9700\u8981\u91cd\u8dd1 preprocessing<\/h2>\n<p>\u5b98\u65b9 plans reference \u7ed9\u51fa\u4e00\u4e2a\u91cd\u8981\u5224\u65ad\uff1a\u5982\u679c\u4fee\u6539\u5f71\u54cd\u5df2\u7ecf\u51c6\u5907\u597d\u7684\u6570\u636e\uff0c\u5c31\u901a\u5e38\u9700\u8981\u91cd\u8dd1 preprocessing\u3002<\/p>\n<table>\n<thead>\n<tr>\n<th>\u4fee\u6539\u5185\u5bb9<\/th>\n<th>\u901a\u5e38\u662f\u5426\u9700\u8981\u91cd\u8dd1 preprocessing<\/th>\n<th>\u539f\u56e0<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><code>spacing<\/code><\/td>\n<td>\u9700\u8981<\/td>\n<td>\u91cd\u91c7\u6837\u7ed3\u679c\u4f1a\u53d8<\/td>\n<\/tr>\n<tr>\n<td><code>preprocessor_name<\/code><\/td>\n<td>\u9700\u8981<\/td>\n<td>\u9884\u5904\u7406\u903b\u8f91\u4f1a\u53d8<\/td>\n<\/tr>\n<tr>\n<td><code>normalization_schemes<\/code><\/td>\n<td>\u9700\u8981<\/td>\n<td>\u8f93\u5165\u5f3a\u5ea6\u5206\u5e03\u4f1a\u53d8<\/td>\n<\/tr>\n<tr>\n<td>resampling functions<\/td>\n<td>\u9700\u8981<\/td>\n<td>\u91cd\u91c7\u6837\u7b97\u6cd5\u4f1a\u53d8<\/td>\n<\/tr>\n<tr>\n<td><code>batch_size<\/code><\/td>\n<td>\u901a\u5e38\u4e0d\u9700\u8981<\/td>\n<td>\u53ea\u5f71\u54cd\u8bad\u7ec3\u52a0\u8f7d\u6279\u91cf\uff0c\u4e0d\u6539\u53d8\u9884\u5904\u7406\u6570\u636e<\/td>\n<\/tr>\n<tr>\n<td>\u90e8\u5206 architecture-only \u8bbe\u7f6e<\/td>\n<td>\u901a\u5e38\u4e0d\u9700\u8981<\/td>\n<td>\u82e5\u4ecd\u590d\u7528\u540c\u4e00\u4efd\u9884\u5904\u7406\u6570\u636e\uff0c\u5219\u4e0d\u5fc5\u91cd\u8dd1<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u5b9e\u8df5\u4e2d\uff0c\u5982\u679c\u4f60\u4e0d\u786e\u5b9a\u67d0\u4e2a\u6539\u52a8\u662f\u5426\u6539\u53d8\u9884\u5904\u7406\u6570\u636e\uff0c\u4fdd\u5b88\u505a\u6cd5\u662f\u65b0\u5efa\u4e00\u4e2a configuration \u548c\u65b0\u7684 <code>data_identifier<\/code>\uff0c\u907f\u514d\u65e7\u6570\u636e\u548c\u65b0\u5b9e\u9a8c\u6df7\u5728\u4e00\u8d77\u3002<\/p>\n<h2>8. \u6846\u67b6\u5730\u56fe\uff1a\u4ece\u547d\u4ee4\u5230\u6a21\u5757<\/h2>\n<p>\u628a\u524d\u9762\u7528\u8fc7\u7684\u547d\u4ee4\u548c\u5185\u90e8\u6a21\u5757\u5bf9\u5e94\u8d77\u6765\uff0c\u53ef\u4ee5\u5f97\u5230\u8fd9\u5f20\u8868\uff1a<\/p>\n<table>\n<thead>\n<tr>\n<th>\u547d\u4ee4<\/th>\n<th>\u4e3b\u8981\u6a21\u5757<\/th>\n<th>\u5173\u952e\u4ea7\u7269<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><code>nnUNetv2_plan_and_preprocess<\/code><\/td>\n<td>fingerprint extractor\u3001planner\u3001preprocessor<\/td>\n<td><code>dataset_fingerprint.json<\/code>\u3001<code>nnUNetPlans.json<\/code>\u3001\u9884\u5904\u7406\u6570\u636e<\/td>\n<\/tr>\n<tr>\n<td><code>nnUNetv2_train<\/code><\/td>\n<td><code>nnUNetTrainer<\/code><\/td>\n<td>checkpoint\u3001\u65e5\u5fd7\u3001validation \u7ed3\u679c<\/td>\n<\/tr>\n<tr>\n<td><code>nnUNetv2_find_best_configuration<\/code><\/td>\n<td>\u6a21\u578b\u9009\u62e9\u3001ensemble\u3001postprocessing \u8bc4\u4f30<\/td>\n<td><code>inference_instructions.txt<\/code><\/td>\n<\/tr>\n<tr>\n<td><code>nnUNetv2_predict<\/code><\/td>\n<td><code>nnUNetPredictor<\/code><\/td>\n<td>\u9884\u6d4b\u5206\u5272\u6587\u4ef6\u3001\u53ef\u9009 probability files<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2>9. \u5b98\u65b9\u8d44\u6599\u5165\u53e3<\/h2>\n<p>\u672c\u6587\u4e3b\u8981\u53c2\u8003\uff1a<\/p>\n<ul>\n<li><a href=\"https:\/\/github.com\/MIC-DKFZ\/nnUNet\/blob\/master\/documentation\/explanation\/how-nnunet-works.md\" target=\"_blank\"  rel=\"nofollow\" >How nnU-Net Works<\/a><\/li>\n<li><a href=\"https:\/\/github.com\/MIC-DKFZ\/nnUNet\/blob\/master\/documentation\/reference\/plans-and-configuration.md\" target=\"_blank\"  rel=\"nofollow\" >Plans and Configuration Reference<\/a><\/li>\n<li><a href=\"https:\/\/github.com\/MIC-DKFZ\/nnUNet\/blob\/master\/documentation\/extending_nnunet.md\" target=\"_blank\"  rel=\"nofollow\" >Extending nnU-Net<\/a><\/li>\n<li><a href=\"https:\/\/github.com\/MIC-DKFZ\/nnUNet\/blob\/master\/nnunetv2\/training\/nnUNetTrainer\/nnUNetTrainer.py\" target=\"_blank\"  rel=\"nofollow\" >nnUNetTrainer.py \u5b98\u65b9\u6e90\u7801<\/a><\/li>\n<li><a href=\"https:\/\/github.com\/MIC-DKFZ\/nnUNet\/blob\/master\/nnunetv2\/inference\/predict_from_raw_data.py\" target=\"_blank\"  rel=\"nofollow\" >predict_from_raw_data.py \u5b98\u65b9\u6e90\u7801<\/a><\/li>\n<\/ul>\n<h2>\u672c\u7bc7\u603b\u7ed3<\/h2>\n<p>nnU-Net v2 \u7684\u5185\u90e8\u6846\u67b6\u53ef\u4ee5\u6309 fingerprint\u3001plans\u3001preprocessing\u3001Trainer\u3001Predictor \u6765\u7406\u89e3\u3002fingerprint \u63cf\u8ff0\u6570\u636e\uff0cplans \u8fde\u63a5\u6570\u636e\u5206\u6790\u548c\u8bad\u7ec3\u914d\u7f6e\uff0cpreprocessing \u751f\u6210\u8bad\u7ec3\u53ef\u8bfb\u7684\u6570\u636e\uff0cTrainer \u63a7\u5236\u8bad\u7ec3\u8fc7\u7a0b\uff0cPredictor \u4f7f\u7528\u6a21\u578b\u548c plans \u5bf9\u65b0\u6570\u636e\u63a8\u7406\u3002\u7406\u89e3\u8fd9\u6761\u94fe\u8def\u540e\uff0c\u540e\u9762\u4fee\u6539 loss\u3001augmentation \u548c network architecture \u624d\u4e0d\u4f1a\u53d8\u6210\u76f2\u6539\u3002<\/p>\n<h2>\u4e0b\u4e00\u7bc7\u9884\u544a<\/h2>\n<p>\u4e0b\u4e00\u7bc7\u6211\u4eec\u6b63\u5f0f\u5f00\u59cb\u4fee\u6539 nnU-Net v2\uff1a\u4ece\u7ee7\u627f <code>nnUNetTrainer<\/code> \u5199\u4e00\u4e2a\u81ea\u5b9a\u4e49 Trainer \u5f00\u59cb\u3002\u6211\u4eec\u4f1a\u8bb2\u4e3a\u4ec0\u4e48\u4e0d\u8981\u76f4\u63a5\u6539\u6838\u5fc3\u6e90\u7801\uff0c\u4ee5\u53ca\u5982\u4f55\u8ba9 <code>nnUNetv2_train<\/code> \u8c03\u7528\u4f60\u7684 Trainer\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u672c\u7bc7\u5efa\u7acb nnU-Net v2 \u5185\u90e8\u6846\u67b6\u5730\u56fe\uff0c\u89e3\u91ca fingerprint\u3001plans\u3001preprocessing\u3001Trainer \u548c Predictor \u5982\u4f55\u4e32\u8d77\u6765\u3002<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"emotion":"","emotion_color":"","title_style":"","license":"","footnotes":""},"categories":[83],"tags":[],"class_list":["post-1056","post","type-post","status-publish","format-standard","hentry","category-83"],"views":2,"_links":{"self":[{"href":"https:\/\/www.eutaboo.com\/index.php\/wp-json\/wp\/v2\/posts\/1056","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.eutaboo.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.eutaboo.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.eutaboo.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.eutaboo.com\/index.php\/wp-json\/wp\/v2\/comments?post=1056"}],"version-history":[{"count":0,"href":"https:\/\/www.eutaboo.com\/index.php\/wp-json\/wp\/v2\/posts\/1056\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.eutaboo.com\/index.php\/wp-json\/wp\/v2\/media?parent=1056"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.eutaboo.com\/index.php\/wp-json\/wp\/v2\/categories?post=1056"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.eutaboo.com\/index.php\/wp-json\/wp\/v2\/tags?post=1056"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}