Pediatric posterior fossa tumors (PFTs) are successfully treated in approximately 70% of patients. However, most survivors experience long-term working memory impairment. The present study examined whether the parameters of diffusion MRI tractography could serve as working memory impairment biomarkers in 60 pediatric PFT survivors. Participants were at least 5 years post-treatment and had received treatment appropriate for their age and type of tumor. Groups included irradiated PFT had undergone radiotherapy, nonirradiated PFT had not, and age, sex, and handedness-matched healthy controls. All participants underwent a cognitive assessment and multimodal MRI including a diffusion MRI sequence. We combined fMRI data collected from the Human Connectome Project database with the acquired diffusion MRI data, to extract the working memory tract and determine tractography parameters for quantitative insights. Participants in the irradiated PFT group exhibited reduced tract volume, fiber density, fiber connectivity, mean length of streamlines, and number of streamlines, compared with both nonirradiated PFT and control groups. Participants in the nonirradiated PFT group also exhibited reduced fiber density, number of streamlines, and mean curvature of streamlines, compared with controls. Poorer working memory scores for the irradiated PFT group correlated with lower tract volume, fiber density, and number of streamlines for verbal working memory. Additionally, these lower scores correlated with reduced fiber density, mean length of streamlines, and number of streamlines for visual working memory. These tractography parameters could serve as biomarkers of working memory deficits and shed light on the detrimental impact of radiotherapy on the working memory tract.

Convolutional neural networks are gradually being recognized in the neuroimaging community as a powerful tool for image analysis. In the present study, we tested the ability of 3D convolutional neural networks to discriminate between whole-brain parametric maps obtained from diffusion-weighted magnetic resonance imaging. Original parametric maps were subjected to intensity-based region-specific alterations, to create altered maps. To analyze how position, size and intensity of altered regions affected the networks’ learning process, we generated monoregion and biregion maps by systematically modifying the size and intensity of one or two brain regions in each image. We assessed network performance over a range of intensity increases and combinations of maps, carrying out 10-fold cross-validation and using a hold-out set for testing. We then tested the networks trained with monoregion images on the corresponding biregion images and vice versa. Results showed an inversely proportional link between size and intensity for the monoregion networks, in that the larger the region, the smaller the increase in intensity needed to achieve good performances. Accuracy was better for biregion networks than for their monoregion counterparts, showing that altering more than one region in the brain can improve discrimination. Monoregion networks correctly detected their target region in biregion maps, whereas biregion networks could only detect one of the two target regions at most. Biregion networks therefore learned a more complex pattern that was absent from the monoregion images. This deep learning approach could be tailored to explore the behavior of other convolutional neural networks for other regions of interest.