上海市HER2型乳腺癌辅助诊断模型训练数据

通过对样本的数据处理和数据加工，提供给辅助诊断人工智能模型进行训练，帮助人工智能模型更好地理解上海市样本场景下HER2型乳腺癌的情况，提取特征，发现规律，最终提高诊断人工智能模型的准确性、鲁棒性和泛化能力。1数据采集：通过正式合作协议，从医疗机构取得匿名化的样本临床数据，包括是否有术后病理结果，术后Her2情况，术后fish情况；2数据处理：对数据进行检查核对，确保所有数据去标识化，处于完全匿名化状态且不可还原的状态，将没有病理结果的数据去除，对异常数据进行清洗去除，对部分缺失数据进行生成式补充；3数据加工：基于原始数据以及算法规则HER2型乳腺癌的术后状态，生成阴性阳性分型标记，具体规则为：如果Her2满足3+或Her2满足2+且同时术后Fish为扩增则标记为阳性，其余标记为阴性。

This dataset undergoes processing and curation to support the training of auxiliary diagnostic artificial intelligence (AI) models, with the goal of enabling the AI models to better understand the scenario of HER2-positive breast cancer among patient samples in Shanghai, extract meaningful features, identify underlying patterns, and ultimately improve the accuracy, robustness and generalization performance of the diagnostic AI models.
1. Data Collection: Anonymized clinical sample data is acquired from medical institutions via formal cooperation agreements, covering postoperative pathological results, postoperative Her2 status, and postoperative FISH status.
2. Data Preprocessing: Data inspection and validation are carried out to ensure that all data is de-identified, fully anonymized and non-reidentifiable. Data without pathological results is removed, abnormal data is cleaned and filtered out, and generative imputation is performed for partially missing entries.
3. Data Curation: Based on the original data and algorithmic rules, negative and positive classification labels are generated for the postoperative status of HER2-positive breast cancer. The specific labeling rules are as follows: samples are labeled as positive if Her2 is 3+ or Her2 is 2+ with concurrent postoperative FISH amplification; otherwise, they are labeled as negative.

搜集汇总

数据集介绍

特点

该数据集包含150条上海市HER2型乳腺癌的临床数据，用于训练辅助诊断人工智能模型，以提高诊断的准确性和泛化能力。数据经过匿名化处理和清洗，每年更新一次，适用于医疗健康领域的AI模型开发。