Datawhale AI夏令营第三期 - 脑PET图像分析和疾病预测挑战赛——环境搭建及 PyTorch Baseline跑通

原教程链接AI夏令营第三期 - 脑PET图像分析和疾病预测挑战赛教程 - 飞书云文档 (feishu.cn)

一、环境搭建

本次采用Anaconda+Pycharm的方式进行，首先在Anaconda中创建环境，使用以下命令创建python版本为3.10名称为PETCompete的虚拟环境

conda create -n PETCompete python=3.10

 激活conda环境

activate PETCompete

使用conda安装cuda11.8及PyTorch

conda install pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia

安装juypter

pip install jupyter

在pycharm中创建项目并使用刚刚创建conda环境

1.在新建项目时添加本地环境

 选择刚刚创建的conda环境

二、Baseline运行

1.下载数据集并新建jupyter notebook

数据集下载链接

https://ai-contest-static.xfyun.cn/2023/data/%E8%84%91PET%E5%9B%BE%E5%83%8F%E5%88%86%E6%9E%90%E5%92%8C%E7%96%BE%E7%97%85%E9%A2%84%E6%B5%8B%E6%8C%91%E6%88%98%E8%B5%9B%E5%85%AC%E5%BC%80%E6%95%B0%E6%8D%AE.zip

2.安装对应包

pip install pandas tqdm scikit-learn nibabel albumentations 

3.读取数据并进行数据增强

import os, sys, glob, argparse
import pandas as pd
import numpy as np
from tqdm import tqdm

import cv2
from PIL import Image
from sklearn.model_selection import train_test_split, StratifiedKFold, KFold

import torch
torch.manual_seed(0)
torch.backends.cudnn.deterministic = False
torch.backends.cudnn.benchmark = True

import torchvision.models as models
import torchvision.transforms as transforms
import torchvision.datasets as datasets
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torch.autograd import Variable
from torch.utils.data.dataset import Dataset

import nibabel as nib
from nibabel.viewers import OrthoSlicer3D
import albumentations as A

#读取数据集
train_path = glob.glob('./脑PET图像分析和疾病预测挑战赛公开数据/Train/*/*')
test_path = glob.glob('./脑PET图像分析和疾病预测挑战赛公开数据/Test/*')
#将训练集和测试集随机打乱
np.random.shuffle(train_path)
np.random.shuffle(test_path)

DATA_CACHE = {}
#定义数据集读取方法
class XunFeiDataset(Dataset):
    def __init__(self, img_path, transform=None):
        self.img_path = img_path
        if transform is not None:
            self.transform = transform
        else:
            self.transform = None
    
    def __getitem__(self, index):
        if self.img_path[index] in DATA_CACHE:
            img = DATA_CACHE[self.img_path[index]]
        else:
            img = nib.load(self.img_path[index]) 
            img = img.dataobj[:,:,:, 0]
            DATA_CACHE[self.img_path[index]] = img
        
        # 随机选择一些通道            
        idx = np.random.choice(range(img.shape[-1]), 50)
        img = img[:, :, idx]
        img = img.astype(np.float32)

        if self.transform is not None:
            img = self.transform(image = img)['image']
        
        img = img.transpose([2,0,1])
        return img,torch.from_numpy(np.array(int('NC' in self.img_path[index])))
    
    def __len__(self):
        return len(self.img_path)
        
#训练集载入与并对图像镜像增强（随机旋转90度，随机裁剪，随机水平翻转，随机对比度，随机亮度对比度）
train_loader = torch.utils.data.DataLoader(
    XunFeiDataset(train_path[:-10],
            A.Compose([
            A.RandomRotate90(),    
            A.RandomCrop(120, 120),    
            A.HorizontalFlip(p=0.5),    
            A.RandomContrast(p=0.5),
            A.RandomBrightnessContrast(p=0.5),
        ])
    ), batch_size=2, shuffle=True, num_workers=1, pin_memory=False
)
#验证集载入及数据增强（随机裁剪）
val_loader = torch.utils.data.DataLoader(
    XunFeiDataset(train_path[-10:],
            A.Compose([
            A.RandomCrop(120, 120),
        ])
    ), batch_size=2, shuffle=False, num_workers=1, pin_memory=False
)
#测试集载入及数据增强（随机裁剪，随机水平翻转，随机对比度）
test_loader = torch.utils.data.DataLoader(
    XunFeiDataset(test_path,
            A.Compose([
            A.RandomCrop(128, 128),
            A.HorizontalFlip(p=0.5),
            A.RandomContrast(p=0.5),
        ])
    ), batch_size=2, shuffle=False, num_workers=1, pin_memory=False
)

2.定义模型，Baseline采用PyTorch自带的Resnet18网络，也可使用Resnet50，Resnet101等，增加了一层卷积层对图像特征镜像进一步提取，增加了512*2的全连接层，用于将输出转为二分类问题

class XunFeiNet(nn.Module):
    def __init__(self):
        super(XunFeiNet, self).__init__()
                
        model = models.resnet18(True)
        model.conv1 = torch.nn.Conv2d(50, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)
        model.avgpool = nn.AdaptiveAvgPool2d(1)
        model.fc = nn.Linear(512, 2)
        self.resnet = model
        
    def forward(self, img):        
        out = self.resnet(img)
        return out
        
model = XunFeiNet()
model = model.to('cuda')
criterion = nn.CrossEntropyLoss().cuda()
optimizer = torch.optim.AdamW(model.parameters(), 0.001)

3.模型训练及验证（WIndows用户需要将原Baseline这一段9和33行的target修改为target.long()）

def train(train_loader, model, criterion, optimizer):
    model.train()
    train_loss = 0.0
    for i, (input, target) in enumerate(train_loader):
        input = input.cuda(non_blocking=True)
        target = target.cuda(non_blocking=True)

        output = model(input)
        loss = criterion(output, target.long())

        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        if i % 20 == 0:
            print(loss.item())
            
        train_loss += loss.item()
    
    return train_loss/len(train_loader)
            
def validate(val_loader, model, criterion):
    model.eval()
    val_acc = 0.0
    
    with torch.no_grad():
        for i, (input, target) in enumerate(val_loader):
            input = input.cuda()
            target = target.cuda()

            # compute output
            output = model(input)
            loss = criterion(output, target.long())
            
            val_acc += (output.argmax(1) == target).sum().item()
            
    return val_acc / len(val_loader.dataset)
    
for _  in range(3):
    train_loss = train(train_loader, model, criterion, optimizer)
    val_acc  = validate(val_loader, model, criterion)
    train_acc = validate(train_loader, model, criterion)
    
    print(train_loss, train_acc, val_acc)

4.模型预测与提交，WIndows用户需将Baseline的25行中的/替换为\\

def predict(test_loader, model, criterion):
    model.eval()
    val_acc = 0.0
    
    test_pred = []
    with torch.no_grad():
        for i, (input, target) in enumerate(test_loader):
            input = input.cuda()
            target = target.cuda()

            output = model(input)
            test_pred.append(output.data.cpu().numpy())
            
    return np.vstack(test_pred)
    
pred = None
for _ in range(10):
    if pred is None:
        pred = predict(test_loader, model, criterion)
    else:
        pred += predict(test_loader, model, criterion)
        
submit = pd.DataFrame(
    {
        'uuid': [int(x.split('\\')[-1][:-4]) for x in test_path],
        'label': pred.argmax(1)
})
submit['label'] = submit['label'].map({1:'NC', 0: 'MCI'})
submit = submit.sort_values(by='uuid')
submit.to_csv('submit2.csv', index=None)

5.提交生成的submit2.csv文件，获得评分0.73418