IntroductionState of the art artificial intelligence (AI) models have the potential to become a “one-stop shop” to improve diagnosis and prognosis in several oncological settings. The external validation of AI models on independent cohorts is essential to evaluate their generalization ability, hence their potential utility in clinical practice. In this study we tested on a large, separate cohort a recently proposed state-of-the-art convolutional neural network for the automatic segmentation of intraprostatic cancer lesions on PSMA PET images.MethodsEighty-five biopsy proven prostate cancer patients who underwent 68Ga PSMA PET for staging purposes were enrolled in this study. Images were acquired with either fully hybrid PET/MRI (N = 46) or PET/CT (N = 39); all participants showed at least one intraprostatic pathological finding on PET images that was independently segmented by two Nuclear Medicine physicians. The trained model was available at https://gitlab.com/dejankostyszyn/prostate-gtv-segmentation and data processing has been done in agreement with the reference work.ResultsWhen compared to the manual contouring, the AI model yielded a median dice score = 0.74, therefore showing a moderately good performance. Results were robust to the modality used to acquire images (PET/CT or PET/MRI) and to the ground truth labels (no significant difference between the model’s performance when compared to reader 1 or reader 2 manual contouring).DiscussionIn conclusion, this AI model could be used to automatically segment intraprostatic cancer lesions for research purposes, as instance to define the volume of interest for radiomics or deep learning analysis. However, more robust performance is needed for the generation of AI-based decision support technologies to be proposed in clinical practice.
Recent European guidelines recommend using brain FDG-PET to differentiate between Alzheimer’s disease (AD) and depressive pseudodementia (DP), with specific hypometabolism patterns across the former group, and typically normal or frontal hypometabolism in the latter. We report the case of a 74 years-old man with DP (MMSE 16/30), whose FDG-PET visual rating and semiquantitative analysis closely mimicked the typical AD pattern, showing severe hypometabolism in bilateral precuneus, parietal and temporal lobes, and sparing frontal areas, suggesting the diagnosis of moderate AD. Shortly after starting antidepressant polytherapy, he underwent formal NPS testing, which revealed moderate impairment of episodic memory and mild impairment on executive and visuospatial tests, judged consistent with neurodegenerative dementia and concomitant depression. Over the following two years, he improved dramatically: repeated NPS assessment did not show significant deficits, and FDG-PET showed restoration of cerebral metabolism. The confirmation of PET findings via semiquantitative analysis, and their reversion to normality with antidepressant treatment, proved the non-neurodegenerative origin of the initial AD-like FDG-PET abnormalities. We review similar cases and provide a comprehensive analysis of their implications, concluding that reversible FDG-PET widespread hypometabolism might represent a biomarker of pseudodementia. Therefore, we suggest caution when interpreting FDG-PET scans of depressed patients with cognitive impairment.
Background: Early- and late-onset dementia with Lewy bodies (EO-DLB and LO-DLB) are similar in terms of core symptoms. However, LO-DLB presents with more amnestic deficits, while EO-DLB shows a rapid cognitive decline and more severe neuropsychiatric symptoms at onset. A contribution of neurotransmitter dysfunction was suggested but never explored, as a possible factor contributing to the reported clinical differences. By using FDG-PET brain metabolism imaging, we aimed to assess the differences between EO-DLB and LO-DLB regarding brain hypometabolism, related neurotransmitter functional topography, and metabolic connectivity. Methods: We included a total of 62 patients, 21 EO-DLB and 41 LO-DLB patients. Statistical parametric mapping (SPM) voxel-wise comparison with a validated dataset of healthy controls (N=112) provided brain hypometabolism patterns. A metabolic connectivity analysis assessed whole-brain and resting-state network (RSN) alterations. Furthermore, we used the JuSpace toolbox to evaluate the correlations between neurotransmitter pathways topography and brain hypometabolism. Results: Both EO- and LO-DLB groups showed typical bilateral occipito-parieto-frontal hypometabolism. Direct between-group comparison revealed a more severe hypometabolism in posterior cingulate cortex (PCC), precuneus, and occipital cortex for EO-DLB and a more severe hypometabolism in fronto-insular cortices for LO-DLB. Metabolic connectivity analysis showed significant reductions in posterior brain regions in both clinical groups compared to controls, as well as connectivity increases in the EO-DLB only. There were differences in the involvement of temporo-parietal and occipital pathological nodes. Specific RSN vulnerabilities were observed in the executive, default mode and limbic networks for EO-DLB and in the attentional network for LO-DLB. The spatial association analysis based on the metabolic differences in neurotransmission showed significant correlations with acetylcholine, gamma-aminobutyric acid (GABA), serotonin, dopamine maps, and hypometabolism in both EO and LO-DLB groups. Of note, the between-group comparison showed a higher correlation for the EO-DLB in the presynaptic serotonergic system. Overall, this indicates the biochemical involvement of metabolic impairment. Conclusions: This metabolic imaging study indicates similarities and differences between EO- and LO-DLB, both in terms of brain hypometabolism, across different neurotransmission networks, and altered connectivity, adding novel biological evidence to the DLB syndromes.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2025 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.