language: "en" tags:

• bert
• 医疗
• 临床
• 死亡率 thumbnail: "https://core.app.datexis.com/static/paper.png"

CORe模型 - 临床死亡风险预测

模型描述

CORe（临床结局表征）模型在论文《利用自监督知识整合从入院记录预测临床结局》中提出。该模型基于BioBERT架构，并针对临床记录、疾病描述和医学文献进行了专门的"临床结局预训练"目标进一步预训练。

此模型检查点专为死亡风险预测任务微调。模型接收患者入院记录作为输入，输出院内死亡风险的预测值。

如何使用CORe死亡风险预测模型

可通过transformers库加载模型：

from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("bvanaken/CORe-clinical-mortality-prediction")
model = AutoModelForSequenceClassification.from_pretrained("bvanaken/CORe-clinical-mortality-prediction")

以下代码展示推理示例：

input = "主诉：头痛\n\n现病史：58岁男性，有高血压史，房颤患者正在服用华法林，因有生以来最剧烈头痛急诊入院。"

tokenized_input = tokenizer(input, return_tensors="pt")
output = model(**tokenized_input)

import torch
predictions = torch.softmax(output.logits.detach(), dim=1)
mortality_risk_prediction = predictions[0][1].item()

更多信息

有关CORe的详细信息和联系方式，请访问CORe.app.datexis.com。

引用

@inproceedings{vanaken21,
  author    = {Betty van Aken and
               Jens-Michalis Papaioannou and
               Manuel Mayrdorfer and
               Klemens Budde and
               Felix A. Gers and
               Alexander Löser},
  title     = {Clinical Outcome Prediction from Admission Notes using Self-Supervised
               Knowledge Integration},
  booktitle = {Proceedings of the 16th Conference of the European Chapter of the
               Association for Computational Linguistics: Main Volume, {EACL} 2021,
               Online, April 19 - 23, 2021},
  publisher = {Association for Computational Linguistics},
  year      = {2021},
}