前列腺癌检测数据集|深度学习数据集

该数据集包含130名PSA水平升高的无症状男性的外周血多参数流式细胞术免疫表型数据，用于基于深度学习的前列腺癌检测与风险分层。数据规模为130例样本（42例良性，88例前列腺癌），涵盖6个关键免疫表型特征及年龄、PSA等变量。特点在于结合biLSTM模型实现高精度分类：癌症检测模型准确率86.79%，AUC达89.31%；风险预测模型准确率94.90%，AUC 94.06%。适用于开发非侵入性AI辅助诊断工具，支持前列腺癌早期筛查与临床风险评估。

This dataset contains multiparametric flow cytometry immunophenotypic data of peripheral blood from 130 asymptomatic men with elevated prostate-specific antigen (PSA) levels, for deep learning-driven prostate cancer detection and risk stratification. It consists of 130 samples, including 42 benign cases and 88 prostate cancer cases, covering 6 key immunophenotypic features plus clinical variables such as age and PSA. A prominent feature of this dataset is its compatibility with biLSTM models to deliver high-precision classification performance: the cancer detection model achieves an accuracy of 86.79% and an AUC of 89.31%, while the risk prediction model attains an accuracy of 94.90% and an AUC of 94.06%. This dataset is suitable for developing non-invasive AI-assisted diagnostic tools to support early screening and clinical risk assessment of prostate cancer.