Search

Introduction: Over the past few decades, numerous researchers have explored the application of machine learning for assessing children's neurological development. Developmental changes in the brain could be utilized to gauge the alignment of its maturation status with the child's chronological age. AI is trained to analyze changes in different modalities and estimate the brain age of subjects. Disparities between the predicted and chronological age can be viewed as a biomarker for a pathological condition. This literature review aims to illuminate research studies that have employed AI to predict children's brain age. Methods: The inclusion criteria for this study were predicting brain age via AI in healthy children up to 12 years. The search term was centered around the keywords "pediatric," "artificial intelligence," and "brain age" and was utilized in PubMed and IEEEXplore. The selected literature was then examined for information on data acquisition methods, the age range of the study population, pre-processing, methods and AI techniques utilized, the quality of the respective techniques, model explanation, and clinical applications. Results: Fifty one publications from 2012 to 2024 were included in the analysis. The primary modality of data acquisition was MRI, followed by EEG. Structural and functional MRI-based studies commonly used publicly available datasets, while EEG-based studies typically relied on self-recruitment. Many studies utilized pre-processing pipelines provided by toolkit suites, particularly in MRI-based research. The most frequently used model type was kernel-based learning algorithms, followed by convolutional neural networks. Overall, prediction accuracy may improve when multiple acquisition modalities are used, but comparing studies is challenging. In EEG, the prediction error decreases as the number of electrodes increases. Approximately one-third of the studies used explainable artificial intelligence methods to explain the model and chosen parameters. However, there is a significant clinical translation gap as no study has tested their model in a clinical routine setting. Discussion: Further research should test on external datasets and include low-quality routine images for MRI. T2-weighted MRI was underrepresented. Furthermore, different kernel types should be compared on the same dataset. Implementing modern model architectures, such as convolutional neural networks, should be the next step in EEG-based research studies. [ABSTRACT FROM AUTHOR]

Objective To examine age-related changes in T 2 relaxation times during infancy and childhood in order to assess T 2 values obtained from routine MRI as a biomarker. Methods From our pool of clinical pediatric MRI examinations at 3 T all patients with normal conventional MRI scans were retrospectively selected. Depending on their clinical findings the identified 99 patients (0–199 months) were divided into 43 healthy controls and 56 diseased children with various clinical abnormalities (developmental delay, epilepsy, prematurity, and deafness). T 2 maps based on routinely performed triple echo turbo spin echo sequences were created. T 2 values were measured in 22 brain regions to determine age-related changes. We also investigated whether such changes differ between healthy and diseased children. Results Age significantly reduced T 2 relaxation times across all regions (p < 0.05), but health status had no impact. With increasing age, T 2 values decreased continuously, with declines faster over the first 10 months and slower thereafter. Early rapid and later slow decline was similar in healthy and diseased groups. Conclusions Using T 2 maps based on clinical MRI data we could determine age-related T 2 relaxation times in 22 brain regions during infancy and childhood. Our data have relevance for future investigator independent T 2 relaxation time measurements in determining whether T 2 values are within the normal range or should be considered as potentially pathologic. [ABSTRACT FROM AUTHOR]

Purpose: The purpose of this study was to examine age-related, infratentorial changes in T relaxation times during infancy and childhood using routine MRI data at 3 Tesla. Methods: One hundred patients (0-199 months) without signal abnormalities on conventional MRI were retrospectively selected from our pool of pediatric MRI examinations. T maps based on our routinely acquired triple-echo turbo spin-echo (TSE) sequence were created. Based on their clinical symptoms, the children were divided into 43 controls and 57 diseased children with different clinical diseases. T relaxation times were measured in 15 infratentorial brain regions (medullary pyramid, ventral and dorsal pons, middle cerebellar peduncle, dentate nucleus, medial and lateral cerebellar hemisphere each on both sides, and in the cerebellar vermis) investigating age-related changes. Secondly, this study examined whether those changes in T values differed between healthy and diseased children. Results: Age significantly reduced T relaxation time in all infratentorial brain regions ( p < 0.05). With increasing age, the T relaxation times decreased continuously, faster in the first 9 months and slower thereafter. Overall, controls did not differ significantly from diseased children ( p > 0.05) apart from the dentate nucleus and cerebellar hemispheres in terms of rapid decline (larger in controls) and the right dorsal pons and left pyramid in terms of slow decline (larger in diseased children). In both groups, the later slow decline was almost negligible. Conclusions: Using T maps, it was possible to determine age-related T relaxation times in the different infratentorial brain regions in this preliminary study. Between neurologically healthy controls and diseased children, no significant differences in T relaxation times could be found overall in the studied regions. [ABSTRACT FROM AUTHOR]