Quantifying Confounding Bias in Neuroimaging Datasets with Causal Inference

被引:8
|
作者
Wachinger, Christian [1 ]
Becker, Benjamin Gutierrez [1 ]
Rieckmann, Anna [2 ]
Poelsterl, Sebastian [1 ]
机构
[1] LMU Munchen, KJP, Artificial Intelligence Med Imaging AI Med, Munich, Germany
[2] Umea Univ, Dept Radiat Sci, Umea, Sweden
关键词
SEGMENTATION;
D O I
10.1007/978-3-030-32251-9_53
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Neuroimaging datasets keep growing in size to address increasingly complex medical questions. However, even the largest datasets today alone are too small for training complex machine learning models. A potential solution is to increase sample size by pooling scans from several datasets. In this work, we combine 12,207 MRI scans from 15 studies and show that simple pooling is often ill-advised due to introducing various types of biases in the training data. First, we systematically define these biases. Second, we detect bias by experimentally showing that scans can be correctly assigned to their respective dataset with 73.3% accuracy. Finally, we propose to tell causal from confounding factors by quantifying the extent of confounding and causality in a single dataset using causal inference. We achieve this by finding the simplest graphical model in terms of Kolmogorov complexity. As Kolmogorov complexity is not directly computable, we employ the minimum description length to approximate it. We empirically show that our approach is able to estimate plausible causal relationships from real neuroimaging data.
引用
收藏
页码:484 / 492
页数:9
相关论文
共 50 条
  • [1] Confounding and collapsibility in causal inference
    Greenland, S
    Robins, JM
    Pearl, J
    [J]. STATISTICAL SCIENCE, 1999, 14 (01) : 29 - 46
  • [2] Confounding Equivalence in Causal Inference
    Pearl, Judea
    Paz, Azaria
    [J]. JOURNAL OF CAUSAL INFERENCE, 2014, 2 (01) : 75 - 93
  • [3] Quantifying biases in causal models:: Classical confounding vs collider-stratification bias
    Greenland, S
    [J]. EPIDEMIOLOGY, 2003, 14 (03) : 300 - 306
  • [4] Disentangled Latent Representation Learning for Tackling the Confounding M-Bias Problem in Causal Inference
    Cheng, Debo
    Xie, Yang
    Xu, Ziqi
    Li, Jiuyong
    Liu, Lin
    Liu, Jixue
    Zhang, Yinghao
    Feng, Zaiwen
    [J]. 23RD IEEE INTERNATIONAL CONFERENCE ON DATA MINING, ICDM 2023, 2023, : 51 - 60
  • [5] Causal inference in epidemiological studies with strong confounding
    Moore, Kelly L.
    Neugebauer, Romain
    van der Laan, Mark J.
    Tager, Ira B.
    [J]. STATISTICS IN MEDICINE, 2012, 31 (13) : 1380 - 1404
  • [6] Discovery and Inference of a Causal Network with Hidden Confounding
    Chen, Li
    Li, Chunlin
    Shen, Xiaotong
    Pan, Wei
    [J]. JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 2023,
  • [7] On the causal structure of information bias and confounding bias in randomized trials
    Shahar, Eyal
    Shahar, Doron J.
    [J]. JOURNAL OF EVALUATION IN CLINICAL PRACTICE, 2009, 15 (06) : 1214 - 1216
  • [8] Sensitivity to Confounding in Causal Inference: From Childhood to Adulthood
    Schofield, E. Christina
    Cheng, Patricia W.
    [J]. PROCEEDINGS OF THE TWENTY-SIXTH ANNUAL CONFERENCE OF THE COGNITIVE SCIENCE SOCIETY, 2004, : 1215 - 1218
  • [9] Causal inference on neuroimaging data with Mendelian randomisation
    Taschler, Bernd
    Smith, Stephen M.
    Nichols, Thomas E.
    [J]. NEUROIMAGE, 2022, 258
  • [10] Sensitivity to Confounding in Causal Inference: From Childhood to Adulthood
    Ford, E. Christina
    Cheng, Patricia W.
    [J]. PROCEEDINGS OF THE TWENTY-SIXTH ANNUAL CONFERENCE OF THE COGNITIVE SCIENCE SOCIETY, 2004, : 398 - 403