Preliminary study of longitudinal changes in the pituitary and brain of children on growth hormone therapy

Altered individual-level morphological similarity network in children with growth hormone deficiency

BACKGROUND

Accumulating evidences indicate regional grey matter (GM) morphology alterations in pediatric growth hormone deficiency (GHD); however, large-scale morphological brain networks (MBNs) undergo these patients remains unclear.

OBJECTIVE

To investigate the topological organization of individual-level MBNs in pediatric GHD.

METHODS

Sixty-one GHD and 42 typically developing controls (TDs) were enrolled. Inter-regional morphological similarity of GM was taken to construct individual-level MBNs. Between-group differences of topological parameters and network-based statistics analysis were compared. Finally, association relationship between network properties and clinical variables was analyzed.

RESULTS

Compared to TDs, GHD indicated a disturbance in the normal small-world organization, reflected by increased Lp, γ, λ, σ and decreased Cp, Eglob (all PFDR < 0.017). Regarding nodal properties, GHD exhibited increased nodal profiles at cerebellum 4-5, central executive network-related left inferior frontal gyrus, limbic regions-related right posterior cingulate gyrus, left hippocampus, and bilateral pallidum, thalamus (all PFDR < 0.05). Meanwhile, GHD exhibited decreased nodal profiles at sensorimotor network -related bilateral paracentral lobule, default-mode network-related left superior frontal gyrus, visual network -related right lingual gyrus, auditory network-related right superior temporal gyrus and bilateral amygdala, right cerebellum 3, bilateral cerebellum 10, vermis 1-2, 3, 4-5, 6 (all PFDR < 0.05). Furthermore, serum markers and behavior scores in GHD group were correlated with altered nodal profiles (P ≤ 0.046, uncorrected).

CONCLUSION

GHD undergo an extensive reorganization in large-scale individual-level MBNs, probably due to abnormal cortico-striatal-thalamo-cerebellum loops, cortico-limbic-cerebellum, dorsal visual-sensorimotor-striatal, and auditory-cerebellum circuitry. This study highlights the crucial role of abnormal morphological connectivity underlying GHD, which might result in their relatively slower development in motor, cognitive, and linguistic functional within behavior problem performance.

deepPGSegNet: MRI-based pituitary gland segmentation using deep learning

Introduction

In clinical research on pituitary disorders, pituitary gland (PG) segmentation plays a pivotal role, which impacts the diagnosis and treatment of conditions such as endocrine dysfunctions and visual impairments. Manual segmentation, which is the traditional method, is tedious and susceptible to inter-observer differences. Thus, this study introduces an automated solution, utilizing deep learning, for PG segmentation from magnetic resonance imaging (MRI).

Methods

A total of 153 university students were enrolled, and their MRI images were used to build a training dataset and ground truth data through manual segmentation of the PGs. A model was trained employing data augmentation and a three-dimensional U-Net architecture with a five-fold cross-validation. A predefined field of view was applied to highlight the PG region to optimize memory usage. The model's performance was tested on an independent dataset. The model's performance was tested on an independent dataset for evaluating accuracy, precision, recall, and an F1 score.

Results and discussion

The model achieved a training accuracy, precision, recall, and an F1 score of 92.7%, 0.87, 0.91, and 0.89, respectively. Moreover, the study explored the relationship between PG morphology and age using the model. The results indicated a significant association between PG volume and midsagittal area with age. These findings suggest that a precise volumetric PG analysis through an automated segmentation can greatly enhance diagnostic accuracy and surveillance of pituitary disorders.