A5: Workflows for Annotation-Efficient Machine Learning in Biomedical Imaging Research

@article{SCHINTKE2024,

title = {Validity constraints for data analysis workflows},

author = {Florian Schintke and Khalid Belhajjame and Ninon De Mecquenem and David Frantz and Vanessa Emanuela Guarino and Marcus Hilbrich and Fabian Lehmann and Paolo Missier and Rebecca Sattler and Jan Arne Sparka and Daniel T. Speckhard and Hermann Stolte and Anh Duc Vu and Ulf Leser},

url = {https://www.sciencedirect.com/science/article/pii/S0167739X24001079},

doi = {https://doi.org/10.1016/j.future.2024.03.037},

issn = {0167-739X},

year  = {2024},

date = {2024-01-01},

urldate = {2024-01-01},

journal = {Future Generation Computer Systems},

volume = {157},

pages = {82--97},

abstract = {Porting a scientific data analysis workflow (DAW) to a cluster infrastructure, a new software stack, or even only a new dataset with some notably different properties is often challenging. Despite the structured definition of the steps (tasks) and their interdependencies during a complex data analysis in the DAW specification, relevant assumptions may remain unspecified and implicit. Such hidden assumptions often lead to crashing tasks without a reasonable error message, poor performance in general, non-terminating executions, or silent wrong results of the DAW, to name only a few possible consequences. Searching for the causes of such errors and drawbacks in a distributed compute cluster managed by a complex infrastructure stack, where DAWs for large datasets typically are executed, can be tedious and time-consuming. We propose validity constraints (VCs) as a new concept for DAW languages to alleviate this situation. A VC is a constraint specifying logical conditions that must be fulfilled at certain times for DAW executions to be valid. When defined together with a DAW, VCs help to improve the portability, adaptability, and reusability of DAWs by making implicit assumptions explicit. Once specified, VCs can be controlled automatically by the DAW infrastructure, and violations can lead to meaningful error messages and graceful behaviour (e.g., termination or invocation of repair mechanisms). We provide a broad list of possible VCs, classify them along multiple dimensions, and compare them to similar concepts one can find in related fields. We also provide a proof-of-concept implementation for the workflow system Nextflow.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{oliveira7benchmarking,

title = {Benchmarking the Influence of Pre-training on Explanation Performance in MR Image Classification},

author = {Marta Oliveira and Rick Wilming and Benedict Clark and Céline Budding and Fabian Eitel and Kerstin Ritter and Stefan Haufe},

year  = {2024},

date = {2024-00-00},

journal = {Frontiers in Artificial Intelligence},

volume = {7},

pages = {1330919},

publisher = {Frontiers},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {Similar neural pathways link psychological stress and structural brain health in health and multiple sclerosis},

author = {Marc-Andre Schulz and Fabian Eitel and Susanna Asseyer and Lil Meyer-Arndt and Tanja Schmitz-Hübsch and Judith Bellmann-Strobl and James Cole and Stefan Gold and Friedemann Paul and Kerstin Ritter and Martin Weygandt},

doi = {https://doi.org/10.1101/2022.12.19.521098},

year  = {2023},

date = {2023-08-18},

urldate = {2022-00-00},

journal = {iScience},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Schulz2022.02.23.481601,

title = {Performance reserves in brain-imaging-based phenotype prediction},

author = {Marc-Andre Schulz and Danilo Bzdok and Stefan Haufe and John-Dylan Haynes and Kerstin Ritter},

url = {https://www.biorxiv.org/content/early/2022/02/25/2022.02.23.481601},

doi = {10.1101/2022.02.23.481601},

year  = {2022},

date = {2022-01-01},

urldate = {2022-01-01},

journal = {bioRxiv},

publisher = {Cold Spring Harbor Laboratory},

abstract = {Machine learning studies have shown that various phenotypes can be predicted from structural and functional brain images. However, in most such studies, prediction performance ranged from moderate to disappointing. It is unclear whether prediction performance will substantially improve with larger sample sizes or whether insufficient predictive information in brain images impedes further progress. Here, we systematically assess the effect of sample size on prediction performance using sample sizes far beyond what is possible in common neuroimaging studies. We project 3-9 fold improvements in prediction performance for behavioral and mental health phenotypes when moving from one thousand to one million samples. Moreover, we find that moving from single imaging modalities to multimodal input data can lead to further improvements in prediction performance, often on par with doubling the sample size. Our analyses reveal considerable performance reserves for neuroimaging-based phenotype prediction. Machine learning models may benefit much more from extremely large neuroimaging datasets than currently believed.Competing Interest StatementThe authors have declared no competing interest.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {Promises and pitfalls of deep neural networks in neuroimaging-based psychiatric research},

author = {Fabian Eitel and Marc-Andre Schulz and Moritz Seiler and Henrik Walter and Kerstin Ritter},

year  = {2021},

date = {2021-02-04},

urldate = {2021-02-04},

journal = {Experimental Neurology},

volume = {339},

keywords = {},

pubstate = {published},

tppubtype = {article}

}