We aimed to use deep learning (DL) techniques to accurately differentiate Parkinson’s disease (PD) from multiple system atrophy (MSA), which share similar clinical presentations. In this retrospective analysis, 206 patients who underwent PET/MR imaging at the Chinese PLA General Hospital were included, having been clinically diagnosed with either PD or MSA; an additional 38 healthy volunteers served as normal controls (NC). All subjects were randomly assigned to the training and test sets at a ratio of 7:3. The input to the model consists of 10 two-dimensional (2D) slices in axial, coronal, and sagittal planes from multi-modal images. A modified Residual Block Network with 18 layers (ResNet18) was trained with different modal images, to classify PD, MSA, and NC. A four-fold cross-validation method was applied in the training set. Performance evaluations included accuracy, precision, recall, F1 score, Receiver operating characteristic (ROC), and area under the ROC curve (AUC). Six single-modal models and seven multi-modal models were trained and tested. The PET models outperformed MRI models. The 11C-methyl-N-2β-carbomethoxy-3β-(4-fluorophenyl)-tropanel (11C-CFT) -Apparent Diffusion Coefficient (ADC) model showed the best classification, which resulted in 0.97 accuracy, 0.93 precision, 0.95 recall, 0.92 F1, and 0.96 AUC. In the test set, the accuracy, precision, recall, and F1 score of the CFT-ADC model were 0.70, 0.73, 0.93, and 0.82, respectively. The proposed DL method shows potential as a high-performance assisting tool for the accurate diagnosis of PD and MSA. A multi-modal and multi-sequence model could further enhance the ability to classify PD.

‌Purpose‌ This phase I, first-in-human study evaluated the safety, dosimetry, and preliminary efficacy of ¹⁷⁷Lu-FAPI-XT, a novel fibroblast activation protein (FAP)-targeted radioligand, in patients with advanced solid tumors refractory to standard therapies. ‌Methods‌ This open-label, non-randomized, dose-escalation, investigator-initiated trial had a “accelerated titration” and “3 + 3” design involved a 6-week ¹⁷⁷Lu-FAPI-XT treatment cycle in patients with advanced-stage solid tumors at 3.7GBq initially, with subsequent cohorts receiving an incremental 1.85GBq dose increase until dose limiting toxicity (DLT) was observed. ‌Results‌ ¹⁷⁷Lu-FAPI-XT was well tolerated and no DLT or grade ≥ 3 treatment-related adverse events (TRAE) were observed. The whole-body effective dose was 0.039 ± 0.013 Sv/GBq. The mean effective half-lives for the whole-body and tumor lesions were 52.13 ± 12.33 h and 31.78 ± 15.03 h, respectively. According to RECIST stable disease (SD) were observed in 5 (5/14, 35.7%) patients, with the highest benefit observed in fibroblastic sarcoma subtypes with SD presented in 4 (4/8, 50%) patients. The median progression-free survival (PFS) was 4.63 m (95%CI: 1.25, NE). Metabolic reductions on PET/CT were observed in 2 sarcoma cases. (URL: hclinicaltrials.gov. Trial registration: NCT06211647, NCT06197139. Registered 4 January 2024.) ‌conclusion‌ ¹⁷⁷Lu-FAPI-XT demonstrated a favorable safety profile, with preliminary signs of tumor response observed. Complete/partial response deficiency in ¹⁷⁷Lu-FAPI-XT therapy necessitate optimized strategies for efficacy improvement.

Purpose Although neuroinflammation may play a key role in the pathology of migraine and its progression to chronic migraine (CM), its specific involvement-particularly the role of microglia- remains unclear. We investigated whether neuroinflammation is involved in the pathophysiology of CM and whether pro-inflammatory signals are associated with its clinical features. Methods Nineteen individuals with CM and 10 healthy controls (HCs) underwent integrated brain positron emission tomography (PET)/magnetic resonance (MR) using the translocator protein (TSPO) radioligand ([¹¹C] PBR28, a marker of glial activation, together with the quantification of blood plasma inflammatory cytokine/chemokine. Volumes in regions of interest (ROI) were calculated based on MRI data and the standardized uptake value ratio (SUVR) for [¹¹C] PBR28 was extracted for each ROI. The Spearman’s rank correlation coefficient between [¹¹C] PBR28 SUVR and changes in plasma factors was calculated. Results CM patients had a significantly higher Hamilton Depression Rating Scale (HAMD) and Hamilton Anxiety Rating Scale (HAMA) scores than that in HCs (p < 0.05). Participants with CM also exhibited reduced volume in the thalamus (p = 0.012), compared with HCs. Moreover [11C] PBR28 binding was increased in the midbrain, occipital lobe and vermis, along with increased interictal plasma interleukin-8 (IL-8) and CX3CL1 levels, in individuals with CM compared with HCs. Notably, the midbrain levels of TSPO were negatively correlated with the headache frequency (r=-0.462, p = 0.046). Conclusions These findings demonstrate increased central inflammation in CM participants compared to HCs, providing imaging evidence for the potential involvement of neuroinflammation in CM pathophysiology. Additionally, the observed reduction in thalamic volume may contribute to the chronification of migraine. Clinical trial number Not applicable.

Objective: To evaluate amyloid-β deposition with ¹⁸F-florbetaben (FBB) PET imaging against ¹¹C-PIB PET in cognitive normal controls (NC), mild cognitive impairment (MCI), and Alzheimer’s disease (AD) patients. Methods: We recruited 45 subjects (15 in each group of NC, MCI, and mild/moderate AD) who had undergone dynamic ¹⁸F-FBB amyloid PET imaging. For comparison study, 17 participants, including six NC, five MCI, and six AD patients, also underwent ¹¹C-PIB PET imaging on separate days. Standardized uptake value ratios (SUVR) were calculated using the cerebellar cortex as the reference region with regions of interest (ROI) manually defined on co-registered CT. Quantitative analysis of mean cortical uptake was calculated using global SUVR. Spearman correlation analysis between MMSE scores and SUVR of 18F-FBB and 11C-PIB images were calculated. Results: One (7%) of the 15 NC participants, nine (60%) of 15 MCI patients, and 12 (80%) of 15 AD patients had amyloid-positive lesions on ¹⁸F-FBB PET images. In AD patients, global SUVR was significantly higher than those of MCI patients (1.73 ± 0.62 vs. 1.55 ± 0.11, P < 0.001) and NC subjects (1.73 ± 0.62 vs. 1.13 ± 0.43, P < 0.001). In the comparison study, one NC participant, five MCI patients, and five AD patients had amyloid-positive lesions on ¹¹C-PIB PET images. There was a significant linear correlation (r² = 0.81, P < 0.001) between ¹⁸F-FBB and PIB global SUVR values. MMSE scores had negative correlations with SUVR on ¹¹C-PIB PET (r₁ = –0.650, P = 0.005) or SUVR on ¹⁸F-FBB PET (r₂ = –0.754, P < 0.0001). Conclusion: Our study suggests that ¹⁸F-FBB is a useful tracer for the evaluation of amyloid-β deposition in vivo and that global SUVR of ¹⁸F-FBB PET might be a reliable tool in the diagnosis of AD.

Autosomal dominant Alzheimer’s disease (ADAD) offers a distinct framework to study the preclinical phase of Alzheimer’s disease (AD), due to its predictable symptom onset and high penetrance of causative mutations. The study aims to examine the spatial distribution and temporal progression of amyloid-beta (Aβ) and tau pathologies, along with mapping the pathology-functional connectivity network, in asymptomatic ADAD mutation carriers using hybrid positron emission tomography/magnetic resonance imaging (PET/MRI). Participants were recruited from the Chinese Familial Alzheimer’s Disease Network, comprising 14 asymptomatic ADAD mutation carriers and 20 cognitively normal healthy controls (CN). Aβ deposition was evaluated using 11C-PIB PET, while tau aggregation was assessed via 18F-MK6240 PET imaging. Resting-state functional connectivity (rsFC) was analyzed to investigate relationships between pathological burden and neural network changes. Through qualitative analysis, ADAD carriers with marked 18F-MK6240 uptake in intracranial regions were categorized into Group 2, while others were designated as Group 1. Asymptomatic ADAD carriers demonstrated a significantly greater Aβ burden across the cortex and striatum compared to CN, although tau PET binding did not differ significantly between the groups. Group 2 participants exhibited elevated 11C-PIB uptake in the neocortex and striatum, and increased 18F-MK6240-PET uptake in the medial temporal and other cortical regions. Compared with Group 1, network mapping of rsFC in Group 2 indicated increased connectivity associated with tau deposition in limbic, posterior cortical, and bilateral temporal regions, overlapping with the default mode network, suggesting potential compensatory mechanisms. Additionally, reduced connectivity in the left medial inferior temporal cortex and fusiform gyrus aligned with findings in sporadic AD cases. This study shows the spatiotemporal progression of Aβ and tau pathologies in preclinical ADAD, supporting the hypothesis that Aβ deposition precedes tau pathology. The rsFC alterations observed associate with tau deposition in asymptomatic carriers indicate early network disruptions. Tau network mapping presents a valuable approach for assessing individualized brain connectivity changes in preclinical AD, mitigating single-subject variability and advancing precision assessment in early-stage AD diagnosis.