HICH-IT高血压性脑内出血数据集
包含头部CT图像与电子病历文本的高血压性脑内出血数据集，标注关键解剖结构，用于AI辅助诊断与治疗。
基本信息
资源简介
HICH-IT高血压性脑内出血数据集
【数据集背景】
本数据集源站标题为《CYBUS123456/HICH-IT-Datasets》。源站首次发布：2023-12-20。
【数据内容】
源站原始描述：HICH Image/Text (HICH-IT): Comprehensive Text and Image Datasets for Hypertensive Intracerebral Hemorrhage Research # HICH Image/Text (HICH-IT) we introduce a new multimodal dataset in the medical field of hypertensive intracerebral hemorrhage(HICH), called as HICH-IT, which includes both textual information and head CT images. # Related notes The dataset only displays a part of the dataset. Due to patient privacy concerns, parts of this dataset are fictionalized. Some data are presented in the…
【适用方向】
影像类数据可用于医学图像分析模型（检测、分割、分类）的训练与评测，也可用于标注质量与成像参数相关性的研究。
【获取与许可】
源站页面：https://github.com/CYBUS123456/HICH-IT-Datasets
【使用注意】
· 源站最后修改于 2024-02-06，引用时请注意版本时效。
下载信息
注册下载
需要注册 Kaggle 账号并登录后下载，适合需要跟踪下载记录和使用 API 的用户。
暂未开放公开下载
无需注册即可直接获取公开样本或文档，适合快速预览和评估数据集质量。
免登录有偿下载
公开数据集受托下载与技术交付服务。
提供高速下载与技术交付服务（收技术服务费，非数据销售）
暂未开放当前数据集主要来源为 Kaggle 公开托管，完整影像包建议通过原始链接或 Kaggle API 获取。
使用方式
数据集获取
git clone https://github.com/CYBUS123456/HICH-IT-Datasets.git
curl -L -o repo.zip https://github.com/CYBUS123456/HICH-IT-Datasets/archive/refs/heads/main.zip
unzip repo.zip
源站 README 摘录（使用方式）
HICH Image/Text (HICH-IT)
we introduce a new multimodal dataset in the medical field of hypertensive intracerebral hemorrhage(HICH), called as HICH-IT, which includes both textual information and head CT images.
Related notes
The dataset only displays a part of the dataset. Due to patient privacy concerns, parts of this dataset are fictionalized. Some data are presented in the form of tensors.If you are interested in this data set and want to obtain the complete data set, please contact email: Li Jiang, drjiangli2019@163.com
Experiment
we use our image data using the U-net segmentation model. Due to the unique structure of the U-net model, which aids in capturing the contextual information of images, the left side of U-net is a deep convolutional network responsible for extracting image features. The right side involves an upsampling process that gradually restores the image resolution. Furthermore, U-net excels in handling detailed aspects of images, enabling precise segmentation of small structures within images, which is crucial for medical imaging.We utilized Python scripting to convert NIfTI format images into PNG format. Subsequently, the images were randomly divided into training and testing datasets in a 7:3 ratio to validate the efficacy of image data when utilized with the U-net model
We use our text data on a Named Entity Recognition(NER) model, Named Entity Recognition (NER) models represent a key task in natural language processing. The objective of NER is to automatically identify named entities in text and categorize them into predefined classes.We divided the text data into training, testing and validation sets. After training,the weight files were imported into the test and validation sets to evaluate the effectiveness of the text data
Related Literature
HICH Image/Text (HICH-IT): Comprehensive Text and Image Datasets for Hypertensive Intracerebral Hemorrhage Research
https://arxiv.org/abs/2401.15934
数据加载示例（图像类）
from PIL import Image
import glob, os
files = (glob.glob(os.path.join(path, "**", "*.png"), recursive=True)
+ glob.glob(os.path.join(path, "**", "*.jpg"), recursive=True)
+ glob.glob(os.path.join(path, "**", "*.tif"), recursive=True))
print("图像文件数:", len(files))
img = Image.open(files[0]); print("尺寸/模式:", img.size, img.mode)
# torchvision Dataset 方式:
# from torchvision import datasets
# ds = datasets.ImageFolder(path) # 要求 子目录=类别
目录组织与标注格式以源站说明和下载后实际文件为准。
数据缺失？
依托陕西、四川两大基地，我们与超过十家三甲医院建立直接合作关系，覆盖合规授权、采集治理、专业标注、数据交付的全流程，为AI医疗团队提供即拿即用的高质量临床数据。
⚡️ 需要数据支持或标注服务？立即联系我们获取专业方案。
