@article {995, title = {Validation of deep learning techniques for quality augmentation in diffusion MRI for clinical studies}, journal = {NeuroImage: Clinical}, volume = {39}, year = {2023}, pages = {103483}, abstract = {

The objective of this study is to evaluate the efficacy of deep learning (DL) techniques in improving the quality of diffusion MRI (dMRI) data in clinical applications. The study aims to determine whether the use of artificial intelligence (AI) methods in medical images may result in the loss of critical clinical information and/or the appearance of false information. To assess this, the focus was on the angular resolution of dMRI and a clinical trial was conducted on migraine, specifically between episodic and chronic migraine patients. The number of gradient directions had an impact on white matter analysis results, with statistically significant differences between groups being drastically reduced when using 21 gradient directions instead of the original 61. Fourteen teams from different institutions were tasked to use DL to enhance three diffusion metrics (FA, AD and MD) calculated from data acquired with 21 gradient directions and a b-value of 1000 s/mm2. The goal was to produce results that were comparable to those calculated from 61 gradient directions. The results were evaluated using both standard image quality metrics and Tract-Based Spatial Statistics (TBSS) to compare episodic and chronic migraine patients. The study results suggest that while most DL techniques improved the ability to detect statistical differences between groups, they also led to an increase in false positive. The results showed that there was a constant growth rate of false positives linearly proportional to the new true positives, which highlights the risk of generalization of AI-based tasks when assessing diverse clinical cohorts and training using data from a single group. The methods also showed divergent performance when replicating the original distribution of the data and some exhibited significant bias. In conclusion, extreme caution should be exercised when using AI methods for harmonization or synthesis in clinical studies when processing heterogeneous data in clinical studies, as important information may be altered, even when global metrics such as structural similarity or peak signal-to-noise ratio appear to suggest otherwise.

}, keywords = {Angular resolution, Artificial Intelligence, Deep learning, Diffusion tensor, diffusion MRI, machine learning}, issn = {2213-1582}, doi = {https://doi.org/10.1016/j.nicl.2023.103483}, url = {https://www.sciencedirect.com/science/article/pii/S2213158223001742}, author = {Santiago Aja-Fern{\'a}ndez and Carmen Mart{\'\i}n-Mart{\'\i}n and {\'A}lvaro Planchuelo-G{\'o}mez and Abrar Faiyaz and Md Nasir Uddin and Giovanni Schifitto and Abhishek Tiwari and Saurabh J. Shigwan and Rajeev Kumar Singh and Tianshu Zheng and Zuozhen Cao and Dan Wu and Stefano B. Blumberg and Snigdha Sen and Tobias Goodwin-Allcock and Paddy J. Slator and Mehmet Yigit Avci and Zihan Li and Berkin Bilgic and Qiyuan Tian and Xinyi Wang and Zihao Tang and Mariano Cabezas and Amelie Rauland and Dorit Merhof and Renata Manzano Maria and Vin{\'\i}cius Paran{\'\i}ba Campos and Tales Santini and Marcelo Andrade da Costa Vieira and SeyyedKazem HashemizadehKolowri and Edward DiBella and Chenxu Peng and Zhimin Shen and Zan Chen and Irfan Ullah and Merry Mani and Hesam Abdolmotalleby and Samuel Eckstrom and Steven H. Baete and Patryk Filipiak and Tanxin Dong and Qiuyun Fan and Rodrigo de Luis-Garc{\'\i}a and Antonio Trist{\'a}n-Vega and Tomasz Pieciak} } @conference {655, title = {Variance Stabilization of Noncentral-Chi Data: Application to Noise Estimation in MRI}, booktitle = {2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI), Prague, 2016}, year = {2016}, month = {2016}, address = {Prague, Czech Republic}, author = {Tomasz Pieciak and Gonzalo Vegas-S{\'a}nchez-Ferrero and Santiago Aja-Fern{\'a}ndez} }