Clinical data warehouses (CDW) are centralised repositories that store and organise a wide range of clinical data within a healthcare organisation. CDWs offer fantastic opportunities for research due to their unique characteristics: i) they constitute very large scale datasets gathering up to millions of patients; ii) they integrate very diverse data types, including electronic health records, laboratory tests and medical imaging; iii) they allow longitudinal studies since they store historical patient data over time; iv) they provide a unique source of real-world data. However, harnessing CDWs for research raises major challenges, such as that related to data quality and heterogeneity. Indeed, the heterogeneity of data quality can be a source of bias in statistical analysis that can lead to erroneous conclusions, in particular when the quality of the data is correlated to the outcome of interest.
Main publications
Quality control
- Bottani, S., Burgos, N., Maire, A., Wild, A., Ströer, S., Dormont, D., and Colliot, O.: Automatic Quality Control of Brain T1-Weighted Magnetic Resonance Images for a Clinical Data Warehouse. Medical Image Analysis, 75:102219, 2022. doi:10.1016/j.media.2021.102219 Available on HAL
- Loizillon, S., Colliot, O., Chougar, L., Stroer, S., Jacob, Y., Maire, A., Dormont, D., and Burgos, N.: Semi-Supervised Domain Adaptation for Automatic Quality Control of FLAIR MRIs in a Clinical Data Warehouse. In Domain Adaptation and Representation Transfer, LNCS, 14293: 84–93, 2024. doi:10.1007/978-3-031-45857-6_9
- Loizillon, S., Bottani, S., Maire, A., Ströer, S., Dormont, D., Colliot, O., and Burgos, N.: Automatic Motion Artefact Detection in Brain T1-Weighted Magnetic Resonance Images from a Clinical Data Warehouse Using Synthetic Data. Medical Image Analysis, 93: 103073, 2024. doi:10.1016/j.media.2023.103073 Available on HAL
- Loizillon, S., Bottani, S., Mabille, S., Jacob, Y., Maire, A., Ströer, S., Dormont, D., Colliot, O., and Burgos, N.: Automated MRI Quality Assessment of Brain T1-Weighted MRI in Clinical Data Warehouses: A Transfer Learning Approach Relying on Artefact Simulation. Machine Learning for Biomedical Imaging, 2(June 2024 issue): 888–915, 2024. doi:10.59275/j.melba.2024-7fgd Available on HAL
Dataset homogenisation
- Bottani, S., Thibeau-Sutre, E., Maire, A., Ströer, S., Dormont, D., Colliot, O., and Burgos, N.: Contrast-Enhanced to Non-Contrast-Enhanced Image Translation to Exploit a Clinical Data Warehouse of T1-Weighted Brain MRI. BMC Medical Imaging, 24(1): 67, 2024. doi:10.1186/s12880-024-01242-3 Available on HAL
Computer-aided diagnosis of dementia
- Bottani, S., Burgos, N., Maire, A., Saracino, D., Ströer, S., Dormont, D., and Colliot, O.: Evaluation of MRI-Based Machine Learning Approaches for Computer-Aided Diagnosis of Dementia in a Clinical Data Warehouse. Medical Image Analysis, 89: 102903, 2023. doi:10.1016/j.media.2023.102903 Available on HAL
- Loizillon, S., Jacob, Y., Aurélien, M., Dormont, D., Colliot, O., and Burgos, N.: Detecting Brain Anomalies in Clinical Routine with the β-VAE: Feasibility Study on Age-Related White Matter Hyperintensities. In Medical Imaging with Deep Learning, 2024. https://openreview.net/forum?id=YFfOvLf2T1 Available on HAL