Editorial: Spatiotemporal & AI trends in neuroscience, neuroimaging, and neurooncology

Editorial on the Research Topic
Spatiotemporal & AI trends in neuroscience, neuroimaging, and neurooncology

Longitudinal brain changes over years, or even days, can reflect changes and insights that can be captured and studied with data science. In this context, this special issue focuses on neuroscience, neurooncology, and neuroimaging. Powerful paradigms, such as advanced artificial intelligence (AI) (Zhang et al., 2019) and spatio-temporal modeling, either biophysical or biomechanical (Mang et al., 2020), have been contributing in transformational ways. This is changing our understanding of brain activity, disease diagnosis, and response to treatment.

It is indisputable that brain function is spatio-temporal. Indeed, we use different methods to quantify different structural and functional aspects of the brain: Diffusion tensor imaging (DTI) maps the structural pathways that support communication, electroencephalography (EEG) magnetoencephalography (MEG) record rapid electrical and magnetic fluctuations, and functional magnetic resonance imaging (fMRI) records slow hemodynamic correlating to neural activity throughout the brain. We need all those modalities to grasp a minimum of understanding of the complex responses of the neurological system. As evidenced by the work in this issue, novel techniques are emerging to model brain activity as a continuous, high-dimensional spatio-temporal process. Dynamic Functional Connectivity (dFC) analyses reveal how networks reconfigure according to timescales (Preti et al., 2017), while graph neural networks (GNNs) are being used to model the propagation of pathological activity in epilepsy or the spread of neurodegeneration (Li et al., 2022).

The rise of AI, especially deep learning, is the driving force behind this revolution. Without the need for strong a priori hypotheses, these models are uniquely suited to reveal complex, non-linear patterns hidden within massive neuroimaging datasets. This enables sophisticated AI architectures to handle multi-modal data integration/fusion, combining MRI, metabolic, histology, genomics, and clinical records, to create a more comprehensive view of the patient.

Nowhere is the impact of this spatiotemporal-AI synergy more impressive than in neurooncology. Management of brain tumors requires precise characterization (e.g., imaging genotyping), delineation of invasive margins, differentiation of treatment effects from true progressive disease (i.e., tumor recurrence), and prediction of patient overall survival and treatment response. All of these are complex spatio-temporal problems. For example, to predict the grade of glioma and molecular markers, such as the status of IDH mutation (Pytlarz et al., 2024), advanced AI models have been applied to multiparametric MRI, allowing non-invasive characterization and personalized treatment planning (Akbari et al., 2020). Spatiotemporal models can track tumor growth trajectories (Mang et al., 2020) and simulate the effects of radiotherapy or resection recovery (Falcó-Roget et al., 2024), moving toward truly personalized treatment planning.

To comprehend trajectories of degeneration or recovery, longitudinal studies with many time-points spread over years have been increasingly useful. Comprehensive multiyear studies such as the Alzheimer's Disease Neuroimaging Initiative (ADNI) (Mueller et al., 2005), which provide multimodal longitudinal data sets that track individuals from healthy aging through mild cognitive impairment to Alzheimer's dementia, have been extremely beneficial to this effort. Instead of relying on only single-time point cross-sectional images, AI models can now use these longitudinal data to assess the temporal progression of the disease. Recurrent neural networks (RNNs) and latent variable models are being trained to predict future rates of cortical thinning, hippocampal atrophy, and cognitive decline based on an individual's initial scan and clinical profile (Li and Fan, 2019). These techniques bring us closer to developing personalized disease development timelines, identifying pivotal moments in pathology, and evaluating the efficacy of treatments by detecting subtle year-over-year changes that are imperceptible to the naked eye.

The research presented in this Research Topic offers an engaging look at a rapidly evolving topic. We are transitioning from static snapshots to dynamic, predictive models of brain health and disease through the combination of spatio-temporal analytics and AI.

In order to speed up the process from the bench to the bedside, we anticipate that this will encourage further innovation and cooperation between the fields of neuroscientific, clinical, and computational science in the future.

Author contributions

AC: Writing – original draft. SB: Writing – review & editing.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The author(s) declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.

Generative AI statement

The author(s) declare that no Gen AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Akbari, H., Rathore, S., Bakas, S., Nasrallah, M. P., Shukla, G., Mamourian, E., et al. (2020). Imaging surrogates of infiltration obtained via multiparametric imaging pattern analysis predict subsequent location of recurrence of glioblastoma. Neuro-Oncology 32, 352–362. doi: 10.1227/NEU.0000000000001202

PubMed Abstract | Crossref Full Text | Google Scholar

Falcó-Roget, J., Cacciola, A., Sambataro, F., and Crimi, A. (2024). Functional and structural reorganization in brain tumors: a machine learning approach using desynchronized functional oscillations. Commun. Biol. 7:419. doi: 10.1038/s42003-024-06119-3

PubMed Abstract | Crossref Full Text | Google Scholar

Li, H., and Fan, Y. (2019). “Early prediction of Alzheimer's disease dementia based on baseline hippocampal MRI and 1-year follow-up cognitive measures using deep recurrent neural networks,” in 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019) (IEEE), 368–371. doi: 10.1109/isbi.2019.8759397

Crossref Full Text | Google Scholar

Li, Z., Hwang, K., Li, K., Wu, J., and Ji, T. (2022). Graph-generative neural network for EEG-based epileptic seizure detection via discovery of dynamic brain functional connectivity. Sci. Rep. 12:18998. doi: 10.1038/s41598-022-23656-1

PubMed Abstract | Crossref Full Text | Google Scholar

Mang, A., Bakas, S., Subramanian, S., Davatzikos, C., and Biros, G. (2020). Integrated biophysical modeling and image analysis: application to neuro-oncology. Ann. Rev. Biomed. Eng. 22, 309–341. doi: 10.1146/annurev-bioeng-062117-121105

PubMed Abstract | Crossref Full Text | Google Scholar

Mueller, S. G., Weiner, M. W., Thal, L. J., Petersen, R. C., Jack, C. R., Jagust, W., et al. (2005). Ways toward an early diagnosis in Alzheimer's disease: the Alzheimer's Disease Neuroimaging Initiative (ADNI). Alzheimers Dement. 1, 55–66.

Google Scholar

Preti, M. G., Bolton, T. A. W., and Van De Ville, D. (2017). The dynamic functional connectome: State-of-the-art and perspectives. NeuroImage 160, 41–54. doi: 10.1016/j.neuroimage.2016.12.061

PubMed Abstract | Crossref Full Text | Google Scholar

Pytlarz, M., Wojnicki, K., Pilanc, P., Kaminska, B., and Crimi, A. (2024). Deep learning glioma grading with the tumor microenvironment analysis protocol for comprehensive learning, discovering, and quantifying microenvironmental features. J. Imaging Inform. Med. 37, 1711–1727. doi: 10.1007/s10278-024-01008-x

PubMed Abstract | Crossref Full Text | Google Scholar

Zhang, L., Lu, L., Wang, X., Zhu, R. M., Bagheri, M., Summers, R. M., et al. (2019). Spatio-temporal convolutional lstms for tumor growth prediction by learning 4d longitudinal patient data. IEEE Trans. Med. Imaging 39, 1114–1126. doi: 10.1109/TMI.2019.2943841

PubMed Abstract | Crossref Full Text | Google Scholar