TabPFN在小数据下预测阿尔茨海默病转化效果优于传统模型
Evaluating TabPFN for Mild Cognitive Impairment to Alzheimer's Disease Conversion in Data Limited Settings
- 用预训练表格式基础模型TabPFN处理小样本医疗数据
- 在50个样本时仍保持AUC=0.892，优于其他模型
- 适合数据稀缺的神经退行性疾病预测研究
准确预测轻度认知障碍（MCI）向阿尔茨海默病（AD）转化对早期干预至关重要，但受限于纵向数据不足，构建可靠预测模型困难。本文使用来自ADNI的TADPOLE数据集，基于人口学、APOE4、MRI体积、脑脊液标志物和PET影像等多模态生物标志物，评估了TabPFN相较于传统机器学习方法在预测3年MCI转AD上的表现。实验在不同训练集规模（N=50至1000）下进行，对比了XGBoost、随机森林、LightGBM和逻辑回归模型。结果表明，TabPFN取得最高性能（AUC=0.892），优于LightGBM（AUC=0.860），尤其在低数据场景下优势明显：当训练样本仅50例时，其性能仍稳定，而传统模型显著下降。这表明基础模型在数据有限的疾病预测中具有巨大潜力。
原文摘要 · Abstract (English)
Accurate prediction of conversion from Mild Cognitive Impairment (MCI) to Alzheimers Diseases (AD) is essential for early intervention, however, developing reliable conversion predictive models is difficult to develop due to limited longitudinal data availability We evaluate TabPFN (Tabular Pre-Trained Foundation Network) against traditional machine learning methods for predicting 3 year MCI to AD conversion using the TADPOLE dataset derived from ADNI. Using multimodal biomarker features extracted from demographics, APOE4, MRI volumes, CSF markers, and PET imaging, we conducted an experimental comparison across varying training set sizes (N=50 to 1000) and models including XGBoost, Random Forest, LightGBM, and Logistic Regression. TabPFN achieved one the highest performance (AUC=0.892), outperforming LightGBM (AUC=0.860) and demonstrating advantages in low data settings. At N=50 training samples, TabPFN maintained strong AUC while the traditional machine learning models struggles at small training samples. These findings demonstrate that foundation models are promising for disease prediction in data limited scenarios, such as Alzheimers diseases.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。