Understanding the complex cellular and spatial organization of glioblastoma (GBM) and its microenvironment is crucial for improving diagnosis and treatment. Here we integrated 121 spatial transcriptomics, single-cell RNA sequencing, single-cell assay for transposase-accessible chromatin using sequencing and patch sequencing profiles from 100 patients to characterize primary GBM tissue. We identified four malignant cellular communities that exhibited consistent patterns of cell-type compositions, gene expression and intercellular interactions across patients. We identified two subpopulations of mesenchymal-like (MES-like) tumor cells: MES-Hyp, colocalized with monocyte-derived brain macrophages in hypoxic regions; and MES-Ast, associated with endothelial cells, pericytes and vascular smooth muscle cells. We also predicted and experimentally verified cell subtypes and ligand-receptor pairs involved in intercellular communications in each cellular community. Furthermore, patch sequencing analysis revealed that synaptic connections with glioma cells were predominantly formed between neurons and oligodendrocyte-progenitor-like tumor cells. Overall, our study provides insights into the spatial organization and intercellular communication in GBM, offering potential therapeutic targets.

Deep brain stimulation (DBS) is an established therapy for Parkinson’s disease, yet conventional onsite programming mandates frequent travel to specialized centres, imposing substantial burdens on patients. Here we present a large real-world analysis of remote programming (RP) for DBS in China, drawing on 20,383 patients with Parkinson’s disease and 42,163 RP sessions (2012–2024). RP achieves comparable satisfaction and effectiveness to onsite programming while reducing the healthcare access inequality index by 30%. These access gains translate into disproportionate economic benefits for the most vulnerable groups, with cost savings two to ten times greater among low-income, remote and advanced-disease populations. Integrated clinical–labour–economy modelling projects annual direct economic benefits of ¥1.09 billion from reduced domestic medical tourism and ¥8.15 billion from labour-cost savings, with cumulative benefits of ¥115–270 billion by 2050 (3.9–9.2% of China’s 2024 basic medical insurance fund). These findings suggest that RP could be a clinically equivalent, more equitable and economically advantageous approach for postoperative DBS management worldwide. After more than 10 years of real-world deployment involving 20,383 patients, this analysis reports on the cost-effectiveness of remote programming of deep brain stimulation, outlining the implications for reducing inequalities in healthcare access and projecting scenarios of large-scale implementation in the context of an ageing population.

BACKGROUND:Previous studies show that quantitative R2* mapping can reveal iron deposition and tissue alterations, potentially aiding Parkinson's disease (PD) management. However, R2* maps are not commonly used in clinical practice due to the extra time required and sensitivity to susceptibility artifacts. PURPOSE:In this work, we propose to evaluate the feasibility of using generative-adversarial-networks (GANs) for synthesis of R2* maps from T1-weighted (T1W) and T2-weighted (T2W) images. METHODS:A GAN model was developed to synthesize R2* maps from T1W and T2W images. 572 internal participants and 268 external participants were included. The internal-dataset was divided into training (344), validation (114), and test data (114), while the external-dataset was an independent test-set. The performance of the proposed model was compared with a 2D Unet model without adversarial loss. The performance of the two models was evaluated using normalized mean square error, peak signal to noise ratio, the structure similarity index measure (SSIM), feature similarity index (FSIM), root mean square difference, average absolute difference, and relative error. Pearson method was used to assess the correlation between synthetic and real values. The area-under-the-receiver-operating-characteristic-curve (AUC) was calculated to evaluate the diagnostic efficacy of R2* maps in distinguishing PD from healthy controls (HC), with a focus on the substantia nigra pars compacta (SNpc). RESULTS:The proposed model performed better than the 2D Unet model. In internal test-set, high correlations were observed between synthetic and real R2* maps across various brain regions, with coefficients ranging from 0.75 to 0.87. The AUC was 0.79 and 0.80 for synthetic and real maps (p = 0.76), respectively. In external test-set, the AUC was 0.84 for synthetic R2* maps. Longitudinal analysis showed a positive correlation between ∆R2* and ∆UPDRS (Unified-Parkinson's-Disease-Rating-Scale) in SNpc (R = 0.69, p = 0.01) and substantia nigra pars reticulata (SNpr) (R = 0.64, p = 0.02) for PD group. CONCLUSION:The synthetic R2* maps demonstrated good correlation with real maps and performed well in diagnosing and evaluating PD in both internal and external datasets, indicating their potential value for PD diagnosis and assessment.

Parkinson’s disease (PD) has afflicted numerous patients and troubled countless families worldwide, but an effective therapeutic approach has not been discovered so far. Yet, emerging evidence suggested that photobiomodulation (PBM) can fundamentally delay and inhibit neuronal degeneration, and thus promisingly serve as a non-invasive alternative to conventional drug and surgical treatments. Nevertheless, the light wavelength and spectrum are crucial to PBM. To optimize the spectral formula of photomedicine and reveal photobiological effects, an acute PD model by injecting paraquat into mice abdomen was established in this study. Then, these mice were treated with the narrowband light-emitting diode (LED)-chip light peaked at 670 nm, the broadband phosphor-converted LED light peaked at 840 nm in 600−1000 nm region, and their combination. The results indicate that the PBM is safe, and the combined (narrow 670 + broad 840 nm) light exerted the most potent therapeutic effects. It significantly attenuated oxidative stress levels, enhanced axonal regeneration, protected dopaminergic neurons in the substantia nigra pars compacta (SNpc), preserved striatal neuronal integrity, and improved neuronal morphology and marker expression. Thereby, broadband wavelength through multitarget synergistic therapy helps to improving pathological features and facilitating neural function repair in acute PD mice. However, behavioral validation remains necessary in future studies and further richens the spectrum engineering to PBM.

BACKGROUND:Deep brain stimulation (DBS) has been increasingly introduced for patients with Parkinson's disease (PD). However, there has been extensive controversy regarding its surgical timing. This study aimed to evaluate surgical outcomes of DBS across different PD durations and identify key prognostic factors. METHODS AND FINDINGS:In this multicenter cohort study, patients with PD who underwent subthalamic DBS between 1/1/2011 and 12/31/2020 from seven representative Chinese national centers were included. Two-year follow-up data were analyzed, accordingly. These patients were classified into short (<5 years), mid (5-10 years), and long (≥10 years) PD duration groups. Primary assessments included part III of the Movement Disorder Society-sponsored revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS-III) at the off-medicine state, Hamilton Anxiety Rating Scale (HAM-A), Hamilton Depression Rating Scale (HAM-D), and Parkinson Disease Questionnaire-39 (PDQ-39) scales. Relative changes in scores were analyzed for within- and between-group comparisons, and prognostic factors were identified via multivariable linear regression. A total of 1,859 patients were screened, and 1,717 patients (749 females) were included for analysis. Respectively, 141, 978, and 598 patients underwent surgeries after short-, mid-, and long-duration. The scores of the MDS-UPDRS-III (off-medicine), HAM-A, HAM-D, and PDQ-39 significantly improved by 46.7% ± 14.1% (mean difference [MD] 25.1, 95% confidence interval [CI] [24.5, 25.7], P < 0.001), 54.4% ± 22.4% (MD 8.0, 95%CI [7.5, 8.5], P < 0.001), 43.4% ± 22.6% (MD 6.3, 95%CI [5.8, 6.8], P < 0.001), and 47.9% ± 17.8% (MD 28.0, 95%CI [27.0, 29.0], P < 0.001), respectively, and all the study groups achieved significant improvements (all P < 0.001). Notably, patients with mid-PD duration achieved greatest improvements in motor outcomes (versus short: MD 8.0%, 95%CI [4.7%, 11.3%], P = 0.008; versus long: MD 5.6%, 95%CI [2.8%, 9.4%], P = 0.01), neuropsychological evaluations (anxiety, versus long: MD 15.2%, 95%CI [12.3%, 18.1%], P = 0.002; depression, versus long: MD 19.1%, 95%CI [15.6%, 22.6%], P < 0.001), and quality of life (versus long: MD 7.6%, 95%CI [5.2%, 10.0%], P = 0.007). Levodopa response (short: adjusted β 0.42, 95% CI [0.30, 0.54], P < 0.001; mid: adjusted β 0.17, 95% CI [0.12, 0.22], P < 0.001; long: adjusted β 0.20, 95% CI [0.12, 0.28], P < 0.001) was a unified positive factor of motor response for all three groups. Higher MDS-UPDRS-III (off-medicine) scores (mid: adjusted β 0.10, 95% CI [0.05, 0.15], P < 0.001; long: adjusted β 0.30, 95% CI [0.23, 0.38], P < 0.001) were positively correlated with motor outcomes for the mid- and long-duration groups. Nevertheless, it was a negative factor for the short duration group (adjusted β -0.25, 95% CI [-0.36, -0.14], P < 0.001). The main limitation of this study is the nonrandomized observational nature introduced potential selection bias and imbalanced comparisons. CONCLUSIONS:DBS significantly improved motor, neuropsychological, and quality-of-life outcomes across all PD durations, with the most substantial benefits observed in mid-duration (5-10 years) patients. While levodopa response was a consistent positive prognostic factor for motor response, caution is warranted for short-duration patients with rapidly progressive motor symptoms, as they exhibited less favorable outcomes.

Sleep disorder is a common concomitant symptom of Parkinson's disease (PD). We retrospectively analyzed the medical records and questionnaire responses of 468 patients with PD who received DBS of the subthalamic nucleus (STN) between 2017 and 2020. These patients were categorized into two groups based on whether their PD Sleep Disorder Scale (PDSS) scores showed improvement three years post-surgery: the improved group and the non-improved group. To identify factors that influence sleep disorder improvement, we conducted both univariate and multivariate regression analyses. Subsequently, we developed a nomogram to predict the likelihood of sleep disturbance improvement. We assessed the nomogram's accuracy and predictive performance through calibration plots, Receiver Operating Characteristic (ROC) curves, and Decision Curve Analysis (DCA). Patients who experienced improvement in sleep disorders following surgery showed better preoperative responses to medication, higher Mini-Mental State Examination (MMSE) scores, and lower Hamilton Anxiety Scale (HAMA) scores, despite having poorer PDSS scores compared to those without post-surgical sleep disorder improvement. Further analysis using univariate and multivariate regression identified preoperative medication responsiveness, MMSE, HAMA, and PDSS score as independent predictors of postoperative PDSS score improvement in PD patients. Utilizing these findings, we developed a nomogram model, which demonstrated a strong predictive accuracy with an area under the ROC curve of 0.78 (95% CI: 0.69–0.88). Calibration plots and decision curve analysis confirmed the nomogram's excellent alignment between predicted outcomes and actual observations. A nomogram was developed to forecast the likelihood of sleep disorder improvement in PD patients three years following DBS of the STN. This tool may hold significant value in prognosticating sleep quality in PD patients after DBS.

As the aging population in China continues to grow, the country's public health sector faces an urgent need to address the significant social challenges posed by Alzheimer's disease (AD). The available clinical treatments for AD are extremely limited, and the effectiveness of these drugs often diminishes after a period of use. Despite substantial global investment in drug research and development, the progress of clinical trials for AD treatments has been exceedingly slow. Over the past 30 years, only seven AD drugs have been approved by the U. S. Food and Drug Administration (FDA). Traditional drug therapies are expensive and can only slow the progression of AD, without halting the progressive degeneration of neurons. Therefore, exploring and developing emerging treatment methods for AD is imperative. Photobiomodulation (PBM) is a non-invasive therapeutic approach that uses red or near-infrared light to stimulate cellular metabolism and biological responses. PBM has the potential to improve brain metabolism and blood circulation, repair damaged neurons in the brain, and stimulate dendritic and neuronal growth, making it a promising non-invasive neurotherapeutic method that could complement drug treatments. This paper discusses the pathological characteristics and pathogenic mechanisms of AD, as well as the challenges faced by existing treatment strategies. It also reviews the research on PBM treatment in AD cellular and animal models and clinical studies, summarizes the history of phototherapy and the current state of advanced PBM phototherapy device development, and finally offers a perspective on the future development of advanced photonic technologies and therapeutic devices for PBM treatment of AD.

BACKGROUND: The tumor microenvironment (TME) of glioma harbors diverse cell types; however, cell metabolic heterogeneity remains to be explored. This study aims to characterize the metabolic features of different cell types in the TME by integrating multiple datasets, including genomics, bulk and single-cell transcriptomics, and metabolomics. METHODS: Unsupervised machine learning was used to construct an energy metabolic classifier based on the metabolic pathways identified from bulk RNA-seq of gliomas in the TCGA dataset. The classifier was externally validated using multiple datasets, including genomics, bulk RNA-seq, snRNA-seq, and the metabolomics data. Furthermore, metabolic heterogeneity associated with the classifier was further characterized at single-cell resolution. RESULTS: The energy metabolism-based classifier stratified patients into two prognostic clusters: patients in cluster 1 were characterized by high pathway activity of glycolysis, the pentose phosphate pathway (PPP), and fatty acid oxidation (FAO), whereas patients in cluster 2 exhibited higher activity in glutaminolysis. This metabolic classifier revealed both intratumoral and intertumoral metabolic heterogeneity, and the complexity was further validated by the metabolomics profiling and snRNA-seq data from the CPTAC dataset. Notably, OSMR, highly expressed in cluster 1, showed significant co-expression with key glycolytic enzyme genes. The OSM/OSMR/JAK1/STAT3 axis potently drives malignant progression of glioma cells, specially enhancing their invasive and migratory capabilities. Single-cell resolution analyses demonstrated that tumor metabolic heterogeneity is primarily driven by malignant cells rather than non-malignant components, while tumor microenvironment (TME) factors were also found to modulate malignant cell metabolism. Significantly, glycolytic activity in glioma cells increased during the phenotypic transition from PN (proneural) to MES (mesenchymal), with cluster 1 metabolic phenotypes predominating in the tumor core. Compared to cluster 2, cluster 1 patients exhibited higher mRNA expression of immunosuppressive checkpoint genes, which correlated with pronounced immunosuppression in the TME. Furthermore, various immune cells demonstrated distinct metabolic preferences at single-cell resolution. CONCLUSIONS: This study developed an energy metabolic-based classifier for gliomas with prognostic and therapeutic potential. Metabolic reprogramming was linked with the PN-to-MES transition of glioma cells and immunosuppression in the tumor microenvironment. Multi-omics data, especially snRNA-seq, offered insights into metabolism heterogeneity at single-cell resolution, enabling personalized treatment strategies.

OBJECTIVES:Conventional deep brain stimulation (cDBS) is an established treatment for Parkinson's disease (PD), whereas adaptive DBS (aDBS) represents a promising approach with potential advantages in minimizing stimulation-induced side effects and enhancing quality of life. This study evaluated the safety and efficacy of a newly developed aDBS closed-loop neurostimulation (CNS) device for one year across multiple centers, with the primary objective of comparing the outcomes of aDBS and cDBS. MATERIALS AND METHODS:This retrospective study included 62 patients with PD who underwent bilateral subthalamic nucleus (STN) DBS. Outcomes were assessed using the Movement Disorder Society Unified Parkinson's Disease Rating Scale (MDS-UPDRS), Parkinson's Disease Questionnaire-39 (PDQ-39), and Schwab and England Activities Scale, whereas the levodopa-equivalent daily dose (LEDD) and adverse events were monitored. This two-phase trial randomized participants into Stim-on or Stim-off groups for 90-day postoperative comparison followed by nonrandomized evaluation of aDBS vs cDBS at 360 days after surgery. RESULTS:At 90 days postoperatively, the Stim-on group exhibited superior outcomes to those in the Stim-off group except for LEDD and speech in the medication-on state. At the 360-day postoperative assessment, the aDBS group showed significantly greater improvements than did the cDBS group in MDS-UPDRS II (57.29% vs 33.02%, p = 0.022), MDS-UPDRS IV (59.83% vs 36.69%, p = 0.026), PDQ-39 (56.91% vs 27.37%, p = 0.031), and LEDD reduction (53.35% vs 29.16%, p = 0.002). CNS aDBS recorded clear STN-beta signals, which could be adopted as a biomarker. CONCLUSIONS:Both aDBS and cDBS significantly alleviate motor symptoms and enhance quality of life in patients with PD. Although comparable in motor symptom control, aDBS indicated advantages over cDBS across LEDD reduction, MDS-UPDRS II, MDS-UPDRS IV, and PDQ-39 over the long term. Further studies with extended follow-up and larger sample sizes are required to validate these results.

Glioblastoma (GBM) is an aggressive and recurrent malignancy with a poor prognosis. Although temozolomide (TMZ) is a cornerstone of GBM treatment, its efficacy is often compromised by inherent or acquired resistance, underscoring the urgent need to uncover molecular mechanisms, discover new therapeutic targets, and develop innovative treatment strategies. In this study, we found an increased formation of filamentous actin (F-actin) within the nuclei of TMZ-resistant GBM cells. We also showed that overexpression of FSCN1 in TMZ-resistant GBM cells promotes F-actin formation and facilitates the repair of DNA double-strand breaks (DSBs). Further investigation revealed a marked decrease in the expression of YTHDC1 in TMZ-resistant GBM cells, which regulates FSCN1 through m6A modification. Additionally, FSCN1 activates the CDC42/N-WASP/Arp2/3 signaling pathway by recruiting FGD1 to activate CDC42GTP, which drives nuclear F-actin formation. Importantly, combining the FSCN1 inhibitor NP-G2-044, with TMZ therapy resulted in stronger anti-tumor effects both in vitro and in vivo. In conclusion, the study demonstrates that nuclear F-actin formation in GBM promotes DSB repair and reveals that targeting FSCN1 with NP-G2-044 could be a promising strategy for enhancing treatment outcomes and improving the prognosis for GBM patients.

[This corrects the article DOI: 10.1016/j.heliyon.2024.e25912.].

BACKGROUND:Glioblastoma stem cells and their exosomes (exos) are involved in shaping the immune microenvironment, which is important for tumor invasion and recurrence. However, studies involving GSC-derived exosomal circular RNAs (GDE-circRNAs) in regulating tumor microenvironment (TME) remain unknown. Here, we comprehensively evaluated the significance of a novel immune-related GDE-circRNA in the glioma microenvironment. METHODS:GDE-circPRKD3 was screened out through high-throughput sequencing and verified by RT-PCR, sanger sequencing, and RNase R assays. A series of in vitro and in vivo experiments were performed to investigate the function of GDE-circPRKD3. RNA-seq, RNA immunoprecipitation, multicolor flow cytometry, and western blotting were used to explore the regulation of GDE-circPRKD3 on STAT3 signaling-mediated TME remodeling. RESULTS:We have characterized a circRNA PRKD3 in GSC exosomes, and lower circPRKD3 expression predicts a worse prognosis for glioblastoma patients. Overexpression of GDE-circPRKD3 significantly impairs the biological competence of glioma and prolongs the survival of xenograft mice. GDE-circPRKD3 binds to HNRNPC in an m6A-dependent manner, accelerates mRNA decay of IL6ST, and inhibits downstream target STAT3. Notably, GDE-circPRKD3 promotes CXCL10 secretion by reprogramming tumor-associated macrophages, which in turn recruits CD8+ tumor-infiltrating lymphocytes against GBM. Moreover, brain-targeted lipid nanoparticle delivery of circPRKD3 combined with immune checkpoint blockade therapy achieves significant combinatorial benefits. CONCLUSIONS:This study provides a novel mechanism by which GDE-circPRKD3 relies on STAT3 signaling to remodel immunosuppressive TME and offers a potential RNA immunotherapy strategy for GBM treatment.

Uric acid (UA) is a naturally antioxidant that is strongly associated with the development and progression of Parkinson’s disease (PD). The purine diet is an important exogenous pathway that modulates blood UA levels. Deep brain stimulation (DBS) is an important tool for PD treatment. This study aimed to explore the effects of preoperative purine diet on the prognosis of patients with PD after DBS. Sixty-four patients with PD who underwent DBS were included in this study, and their clinical data, blood UA levels, and daily purine intake. Patients were followed up for improvement 1 year after surgery. We found that patient higher purine intake was strongly associated with the rate of improvement after DBS and was a protective factor for patient prognosis. Daily purine intake from meat and seafood was significantly higher in the responsive patients than in the less-responsive patients. Mediation analysis showed that UA mediated 78% of the effect of purine intake on motor symptom improvement after DBS. In summary, we observed that purine intake is strongly associated with the rate of improvement in motor symptoms after subthalamic nucleus-DBS in patients with PD. This study provides a reference for preoperative diet planning in patients with PD undergoing DBS.

INTRODUCTION: Parkinson's Disease (PD) is a growing burden with varied clinical manifestations and responses to Subthalamic Nucleus Deep Brain Stimulation (STN-DBS). At present, there is no effective and simple machine learning model based on comprehensive clinical scales to predict the improvement in motor symptoms of PD treated with DBS. METHODS: A total of 647 PD patients from the First Affiliated Hospital of University of Science and Technology of China were enrolled retrospectively. Modeling with 5 kinds of machine learning algorithms (SVM, LR, RF, LigtGBM and XGBoost ), and using the Qingdao municipal hospital, 123 cases of PD patients as the external data verify the validity of the model. RESULTS: The LightGBM model outperformed others, demonstrating an internal test set AUC of 0.874 (95%CI [0.822-0.927]) and an average AUC of 0.921±0.03 during crossvalidation. The external validation yielded an AUC of 0.769 (95% CI[0.685-0.853]). Key predictive variables identified include MMSE scores, HAMA scores, years of education, medication improvement rate, and preoperative UPDRS scores. On the basis of the selected variables and the machine-learning model, we developed a web calculator for clinical use. CONCLUSIONS: The results indicate that the LightGBM model based on the top seven influencing factors is a promising tool for predicting the improvement in motor symptoms of PD after 1 year of DBS.

Parkinson’s Disease (PD) is a growing burden with varied clinical manifestations and responses to Subthalamic Nucleus Deep Brain Stimulation (STN-DBS). At present, there is no effective and simple machine learning model based on comprehensive clinical scales to predict the improvement in motor symptoms of PD treated with DBS. A total of 647 PD patients from the First Affiliated Hospital of University of Science and Technology of China were enrolled retrospectively. LightGBM machine learning algorithm was used for modeling, and 123 PD patients from Qingdao Municipal Hospital were used as external data to verify the effectiveness of the model. The study was registered in the Chinese Clinical Trial Registry with the registration number of ChiCTR2300073955. The LightGBM model outperformed others, demonstrating an internal test set AUC of 0.874 (95

Exosomes play a crucial role in regulating crosstalk between tumor and tumor stem-like cells through their cargo molecules. Circular RNAs (circRNAs) have recently been demonstrated to be critical factors in tumorigenesis. This study focuses on the molecular mechanism by which circRNAs from glioma stem-like cell (GSLC) exosomes regulate glioblastoma (GBM) tumorigenicity. In this study, we validated that GSLC exosomes accelerated the malignant phenotype of GBM. Subsequently, we found that circZNF800 was highly expressed in GSLC exosomes and was negatively associated with GBM patients. CircZNF800 promoted GBM cell proliferation and migration and inhibited GBM cell apoptosis in vitro. Silencing circZNF800 could improve the GBM xenograft model survival rate. Mechanistic studies revealed that circZNF800 activated the PIEZO1/Akt signaling pathway by sponging miR-139-5p. CircZNF800 derived from GSLC exosomes promoted GBM cell tumorigenicity and predicted poor prognosis in GBM patients. CircZNF800 has the potential to serve as a promising target for further therapeutic exploration.

The widespread utilization of lasers and optical instruments in clinical settings to evaluate and manage patients is a testament to the profound influence that light and optical methodologies have had on contemporary medicine. Advancements in the field of biomedical optics have provided opportunities for developing more advanced technologies, particularly through integrating photonics with tissues, genetic engineering, and biomaterials. This research encompasses the basic principles of interactions between light and matter, the use of light in treatment and surgery, a concise overview of their clinical applications, and the potential of developing technologies based on light.

Parkinson's disease (PD) is a common neurodegenerative disorder. The main neuropathological features of PD are the loss of dopaminergic neurons in the substantia nigra striata and the extensive accumulation of α-synuclein in different brain regions. The substantia nigra and striatum are brain regions with high expression of cannabinoid receptors, and the role of the endogenous cannabinoid system (ECS) has been explored in PD. The ECS may be a promising target for the treatment of PD. The correlation between PD and serum uric acid levels has been demonstrated in a large number of studies, but a causal relationship between the two has not been proven. Xanthine and hypoxanthine are upstream metabolite of uric acid, and they have been shown to be effective in relieving PD symptoms. Therefore, we hypothesized that xanthine or hypoxanthine or both may increase the inhibitory properties of GABAergic neurons by increasing the sensitivity of cannabinoid receptors on GABAergic neurons in the striatum, thereby activating the direct nigrostriatal pathway to ameliorate Parkinson's disease symptoms. The proposed hypothesis can be tested in animal and cellular intervention experiments. If this approach is found to be effective in improving PD symptoms with minimal adverse events, exogenous supplementation of xanthine/ hypoxanthine or the use of xanthine oxidase to block its metabolism to elevate hypoxanthine levels could potentially be an effective addition to PD therapy.

Parkinson's Disease (PD) is a progressive neurodegenerative disorder with substantial impact on patients' quality of life. Subthalamic nucleus deep brain stimulation (STN-DBS) is an effective treatment for advanced PD, but patient responses vary, necessitating predictive models for personalized care. Recent advancements in medical imaging and machine learning offer opportunities to enhance predictive accuracy, particularly through deep learning and multi-instance learning (MIL) techniques. This retrospective study included 127 PD patients undergoing STN-DBS. Medical records and imaging data were collected, and patients were categorized based on treatment outcomes. Advanced segmentation models were trained for automated region of interest (ROI) delineation. A novel 2.5D deep learning approach incorporating multi-slice representation was developed to extract detailed ROI features. Multi-instance learning fusion techniques integrated predictions across multiple slices, combining radiomics and deep learning features to enhance model performance. Various machine learning algorithms were evaluated, and model robustness was assessed using cross-validation and hyperparameter optimization. The MIL model achieved an area under the curve (AUC) of 0.846 for predicting STN-DBS outcomes, surpassing the radiomics model's AUC of 0.825. Integration of MIL and radiomics features in the DLRad model further improved discriminative ability to an AUC of 0.871. Calibration tests showed good model reliability, and decision curve analysis demonstrated clinical utility, affirming the model's predictive advantage. This study demonstrates the efficacy of integrating MIL, radiomics, and deep learning techniques to predict STN-DBS outcomes in PD patients. The multimodal fusion approach enhances predictive accuracy, supporting personalized treatment planning and advancing patient care.