Clinical Implication of Individually Tailored Segmentation Method for Distorted Hypothalamus in Craniopharyngioma

Obesity mechanism after hypothalamic damage: Cohort analysis of neuroimaging, psychological, cognitive, and clinical phenotyping data

OBJECTIVE

The hypothalamus regulates energy homeostasis, and its damage results in severe obesity. We aimed to investigate the multifaceted characteristics of hypothalamic obesity.

METHODS

We performed multidimensional analyses of brain structure/function and psychological and behavioral phenotypes in 29 patients with hypothalamic damage (HD) (craniopharyngioma) and 31 controls (non-functional pituitary adenoma). Patients underwent structural and functional magnetic resonance imaging and completed self-reports and cognitive tasks.

RESULTS

Patients with HD showed significantly higher postoperative weight gain than controls. The HD group also showed significant hypothalamic damage and lower neural activation in the left caudate nucleus in response to food images. The HD group had significantly higher food inattention, lower satiety, and higher restrained eating behavior. Within the HD group, higher restrained eating behavior was significantly associated with lower activation in the bilateral fusiform gyrus.

CONCLUSION

These results suggest that hypothalamic damage contributes to weight gain by altering the brain response, attention, satiety, and eating behaviors. The present study proposes novel neuro-psycho-behavioral mechanisms targeted for patients with hypothalamic obesity.

Deep-learning-based automatic segmentation and classification for craniopharyngiomas

Objective

Neuronavigation and classification of craniopharyngiomas can guide surgical approaches and prognostic information. The QST classification has been developed according to the origin of craniopharyngiomas; however, accurate preoperative automatic segmentation and the QST classification remain challenging. This study aimed to establish a method to automatically segment multiple structures in MRIs, detect craniopharyngiomas, and design a deep learning model and a diagnostic scale for automatic QST preoperative classification.

Methods

We trained a deep learning network based on sagittal MRI to automatically segment six tissues, including tumors, pituitary gland, sphenoid sinus, brain, superior saddle cistern, and lateral ventricle. A deep learning model with multiple inputs was designed to perform preoperative QST classification. A scale was constructed by screening the images.

Results

The results were calculated based on the fivefold cross-validation method. A total of 133 patients with craniopharyngioma were included, of whom 29 (21.8%) were diagnosed with type Q, 22 (16.5%) with type S and 82 (61.7%) with type T. The automatic segmentation model achieved a tumor segmentation Dice coefficient of 0.951 and a mean tissue segmentation Dice coefficient of 0.8668 for all classes. The automatic classification model and clinical scale achieved accuracies of 0.9098 and 0.8647, respectively, in predicting the QST classification.

Conclusions

The automatic segmentation model can perform accurate multi-structure segmentation based on MRI, which is conducive to clearing tumor location and initiating intraoperative neuronavigation. The proposed automatic classification model and clinical scale based on automatic segmentation results achieve high accuracy in the QST classification, which is conducive to developing surgical plans and predicting patient prognosis.

Resting energy expenditure in children at risk of hypothalamic dysfunction

Objective

Children with suprasellar brain damage are at risk of hypothalamic dysfunction (HD). HD may lead to decreased resting energy expenditure (REE). Decreased REE, however, is not present in all children with HD. Our aim was to assess which children suspect for HD have low REE, and its association with clinical severity of HD or radiological hypothalamic damage.

Patients and methods

A retrospective cohort study was performed. Measured REE (mREE) of children at risk of HD was compared to predicted REE (pREE). Low REE was defined as mREE <90% of predicted. The mREE/pREE quotient was associated to a clinical score for HD symptoms and to radiological hypothalamic damage.

Results

In total, 67 children at risk of HD (96% brain tumor diagnosis) with a mean BMI SDS of +2.3 ± 1.0 were included. Of these, 45 (67.2%) had low mREE. Children with severe HD had a significant lower mean mREE/pREE quotient compared to children with no, mild, or moderate HD. Mean mREE/pREE quotient of children with posterior hypothalamic damage was significantly lower compared to children with no or anterior damage. Tumor progression or tumor recurrence, severe clinical HD, and panhypopituitarism with diabetes insipidus (DI) were significant risk factors for reduced REE.

Conclusion

REE may be lowered in children with hypothalamic damage and is associated to the degree of clinical HD. REE is, however, not lowered in all children suspect for HD. For children with mild or moderate clinical HD symptoms, REE measurements may be useful to distinguish between those who may benefit from obesity treatment that increases REE from those who would be better helped using other obesity interventions.