J Neurointerv Surg. 2025 Mar 5:jnis-2025-023068. doi: 10.1136/jnis-2025-023068. Online ahead of print.

ABSTRACT

BACKGROUND: Newly diagnosed glioblastoma (ndGBM) remains one of the most challenging malignancies to treat. Since the majority of patients experience tumor recurrence (rGBM) after first-line therapy, advancements in both initial and salvage treatments are essential.

OBJECTIVE: We report our single-center experience on the feasibility and safety of superselective intra-arterial cerebral infusion (SIACI) with bevacizumab or cetuximab after osmotic blood-brain barrier disruption (oBBBd).

METHODS: Partial results of three distinct trials (anonymized for blinded review) were analyzed. All patients were histopathologically confirmed to have either ndGBM or previously diagnosed ndGBM that progressed to rGBM despite standard therapy and had aKarnofsky Performance Status (KPS)≥70. All patients were admitted on the same day of the surgery, and the intervention followed similar steps in all included patients. Under general anesthesia, after oBBBd with mannitol, patients received SIACI.

RESULTS: Between October 2014 and March 2024, 70 patients with a mean age of 56.2±12.4 years (range: 19-78) underwent successful treatment, encompassing 139 SIACIs and 246 infusions. All planned SIACIs were completed successfully. Forty-one patients with rGBM received bevacizumab-SIACI, 7 with ndGBM bevacizumab-SIACI, and 22 with ndGBM cetuximab-SIACI. In 133 of 139 SIACIs (95.7%), patients were discharged home with a length of stay of 1 day. The incidence of patients who experienced procedure-related and drug-related adverse events was 11.4% and 8.6%, respectively. No procedure-related deaths occurred.

CONCLUSION: In our single-center experience, comprising the largest cohort of bevacizumab or cetuximab SIACI treatment for rGBM and ndGBM, this promising and cutting-edge intervention is highly feasible and safe.

PMID:40044413 | DOI:10.1136/jnis-2025-023068

J Affect Disord. 2025 Mar 3:S0165-0327(25)00345-3. doi: 10.1016/j.jad.2025.03.007. Online ahead of print.

ABSTRACT

This study evaluated adverse events (AEs) associated with Vortioxetine by analyzing extensive data from the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS). We collected data from the FAERS database spanning ten years, from the first quarter of 2014 to the second quarter of 2024, focusing on drug-related AEs involving Vortioxetine. A comprehensive analysis was performed using multiple signal detection methods, including reporting odds ratio (ROR), proportional reporting ratio (PRR), Bayesian Confidence Propagation Neural Network (BCPNN), and Multi-item Gamma Poisson Shrinker (MGPS). Among 13,116 reports where Vortioxetine was identified as the primary suspect drug, AEs were categorized into 27 system organ classes (SOCs) and 146 preferred terms (PTs). The results highlighted significant signals for common AEs, such as psychiatric disorders, gastrointestinal disorders, and nervous system disorders. Notably, feeling guilty exhibited the strongest signal strength; however, its clinical relevance requires cautious interpretation. Additionally, the study identified novel signals not listed in the drug label but potentially of clinical value, such as hyperarousal and alcoholic, which were significantly associated with Vortioxetine. Of particular note, AEs related to sexual dysfunction were the most diverse, while suicidal ideation was the most frequently reported. The study also uncovered rare but noteworthy signals, including hallucination and olfactory disorders, dermatillomania, and bruxism, which warrant further attention. In conclusion, while Vortioxetine demonstrates multifaceted benefits in alleviating symptoms of depression, its clinical use requires a comprehensive evaluation of potential risks. Developing safe and rational treatment strategies is essential to optimize therapeutic outcomes.

PMID:40044085 | DOI:10.1016/j.jad.2025.03.007

Parasite. 2025;32:18. doi: 10.1051/parasite/2025009. Epub 2025 Mar 5.

ABSTRACT

Leishmaniases are vector-borne parasitic diseases that pose a threat to over 1 billion people worldwide. The parasites target cells of the reticulohistiocytic system, such as macrophages, where they replicate. The disease manifests in various forms, ranging from localized cutaneous leishmaniasis to life-threatening visceral forms, which are fatal in 95% of cases without treatment. Current treatments rely on the invasive administration of toxic and expensive drugs that are increasingly encountering resistance. Therefore, finding alternative treatments for this disease is imperative. This literature review focuses on recent advancements in alternative treatments and aims to present the various strategies designed to address current limitations, including cost, toxicity, off-target effects, administration routes, and the emergence of drug resistance. Starting with an overview of the existing approved treatments and their specific limitations, we categorize treatment development strategies into five key sections: (i) combination therapies using existing approved treatments to enhance efficacy and reduce resistance; (ii) nanoparticle formulations, which enable targeted delivery to infected organs and improved therapeutic efficiency; (iii) drug repositioning, a strategy that has already contributed to the approval of over half of current therapeutic compounds; (iv) immunomodulation, used in conjunction with standard chemotherapies to enhance treatment efficacy and lower relapse rates; and (v) ethnobotanicals, which have demonstrated promising in vitro results by combining low toxicity, immunomodulatory properties, and potent anti-parasitic effects. In summary, this review outlines current strategies in treatment development, emphasizing their advantages over conventional therapies while acknowledging their limitations.

PMID:40043198 | DOI:10.1051/parasite/2025009

Ann Am Thorac Soc. 2025 Jan;22(1):39-40. doi: 10.1513/AnnalsATS.202411-1148ED.

NO ABSTRACT

PMID:40043192 | DOI:10.1513/AnnalsATS.202411-1148ED

PLoS One. 2025 Mar 5;20(3):e0319589. doi: 10.1371/journal.pone.0319589. eCollection 2025.

NO ABSTRACT

PMID:40043015 | DOI:10.1371/journal.pone.0319589

ACS Nano. 2025 Mar 5. doi: 10.1021/acsnano.4c18035. Online ahead of print.

ABSTRACT

Small interfering RNA (siRNA) targeting the NOD-like receptor family pyrin domain-containing 3 (NLRP3) inflammasome has emerged as a promising therapeutic strategy to mitigate infarct volume and brain injury following ischemic stroke. However, the clinical translation of siRNA-based therapies is significantly hampered by the formidable blood-brain barrier (BBB), which restricts drug penetration into the central nervous system. To address this challenge, we have developed an innovative long-circulating near-infrared II (NIR-II) nanoparticle platform YWFC NPs, which is meticulously engineered to enhance BBB transcytosis and enable efficient delivery of siRNA targeting NLRP3 (siNLRP3@YWFC NPs) in preclinical models of ischemic stroke. Furthermore, we integrated advanced deep learning neural network algorithms to optimize in vivo NIR-II imaging of the cerebral infarct penumbra, achieving an improved signal-to-background ratio at 72 h poststroke. In vivo studies employing middle cerebral artery occlusion (MCAO) mouse models demonstrated that image-guided therapy with siNLRP3@YWFC NPs, guided by prolonged NIR-II imaging, resulted in significant therapeutic benefits.

PMID:40042964 | DOI:10.1021/acsnano.4c18035

IEEE Trans Pattern Anal Mach Intell. 2025 Mar 5;PP. doi: 10.1109/TPAMI.2025.3548620. Online ahead of print.

ABSTRACT

Understanding the effect of hyperparameters of the network structure on the performance of Convolutional Neural Networks (CNNs) remains the most fundamental and urgent issue in deep learning, and we attempt to address this issue based on the piecewise linear (PWL) function nature of CNNs in this paper. Firstly, the operations of convolutions, ReLUs and Max pooling in a CNN are represented as the multiplication of multiple matrices for a fixed sample in order to obtain an algebraic expression of CNNs, this expression clearly suggests that CNNs are PWL functions. Although such representation has high time complexity, it provides a more convenient and intuitive way to study the mathematical properties of CNNs. Secondly, we develop a tight bound of the number of linear regions and the upper bounds of generalization error for CNNs, both taking into account factors such as the number of layers, dimension of pooling, and the width in the network. The above research results provide a possible guidance for designing and training CNNs.

PMID:40042958 | DOI:10.1109/TPAMI.2025.3548620

IEEE Trans Biomed Eng. 2025 Mar 5;PP. doi: 10.1109/TBME.2025.3548297. Online ahead of print.

ABSTRACT

In multispectral fluorescence lifetime imaging (FLIm), achieving consistent imaging quality across all spectral channels is crucial for accurately identifying a wide range of fluorophores. However, these essential measurements are frequently compromised by saturation artifacts due to the inherently limited dynamic range of detection systems. To address this issue, we present SatCompFLImNet, a deep learning-based network specifically designed to correct saturation artifacts in multispectral FLIm, facilitating high dynamic range applications. Leveraging generative adversarial networks, SatCompFLImNet effectively compensates for saturated fluorescence signals, ensuring accurate lifetime measurements across various levels of saturation. Extensively validated with simulated and real-world data, SatCompFLImNet demonstrates remarkable capability in correcting saturation artifacts, improving signal-to-noise ratios, and maintaining fidelity of lifetime measurements. By enabling reliable fluorescence lifetime measurements under a variety of saturation conditions, SatCompFLImNet paves the way for improved diagnostic tools and a deeper understanding of biological processes, making it a pivotal advancement for research and clinical diagnostics in tissue characterization and disease pathogenesis.

PMID:40042955 | DOI:10.1109/TBME.2025.3548297

IEEE Trans Biomed Eng. 2025 Mar 5;PP. doi: 10.1109/TBME.2025.3548092. Online ahead of print.

ABSTRACT

This study investigates a novel approach for analyzing Sit-to-Stand (STS) movements using millimeterwave (mmWave) radar technology, aiming to develop a noncontact, privacy-preserving, and all-day operational solution for healthcare applications. A 60GHz mmWave radar system was employed to collect radar point cloud data from 45 participants performing STS motions. Using a deep learning-based pose estimation model and Inverse Kinematics (IK), we calculated joint angles, segmented STS motions, and extracted clinically relevant features for fall risk assessment. The extracted features were compared with those obtained from Kinect and wearable sensors. While Kinect provided a reference for motion capture, we acknowledge its limitations compared to the gold-standard VICON system, which is planned for future validation. The results demonstrated that mmWave radar effectively captures general motion patterns and large joint movements (e.g., trunk), though challenges remain for more finegrained motion analysis. This study highlights the unique advantages and limitations of mmWave radar and other sensors, emphasizing the potential of integrated sensor technologies to enhance the accuracy and reliability of motion analysis in clinical and biomedical research. Future work will expand the scope to more complex movements and incorporate high-precision motion capture systems to further validate the findings.

PMID:40042953 | DOI:10.1109/TBME.2025.3548092

IEEE J Biomed Health Inform. 2025 Mar 5;PP. doi: 10.1109/JBHI.2025.3548048. Online ahead of print.

ABSTRACT

Applying deep learning to predict patient prognostic survival outcomes using histological whole-slide images (WSIs) and genomic data is challenging due to the morphological and transcriptomic heterogeneity present in the tumor microenvironment. Existing deep learning-enabled methods often exhibit learning biases, primarily because the genomic knowledge used to guide directional feature extraction from WSIs may be irrelevant or incomplete. This results in a suboptimal and sometimes myopic understanding of the overall pathological landscape, potentially overlooking crucial histological insights. To tackle these challenges, we propose the CounterFactual Bidirectional Co-Attention Transformer framework. By integrating a bidirectional co-attention layer, our framework fosters effective feature interactions between the genomic and histology modalities and ensures consistent identification of prognostic features from WSIs. Using counterfactual reasoning, our model utilizes causality to model unimodal and multimodal knowledge for cancer risk stratification. This approach directly addresses and reduces bias, enables the exploration of 'what-if' scenarios, and offers a deeper understanding of how different features influence survival outcomes. Our framework, validated across eight diverse cancer benchmark datasets from The Cancer Genome Atlas (TCGA), represents a major improvement over current histology-genomic model learning methods. It shows an average 2.5% improvement in c-index performance over 18 state-of-the-art models in predicting patient prognoses across eight cancer types. Our code is released at https://github.com/BusyJzy599/CFBCT-main.

PMID:40042950 | DOI:10.1109/JBHI.2025.3548048

IEEE J Biomed Health Inform. 2025 Mar 5;PP. doi: 10.1109/JBHI.2025.3548263. Online ahead of print.

ABSTRACT

Bronchoalveolar lavage fluid (BALF) is a liquid obtained from the alveoli and bronchi, often used to study pulmonary diseases. So far, proteomic analyses have identified over three thousand proteins in BALF. However, the comprehensive characterization of these proteins remains challenging due to their complexity and technological limitations. This paper presented a novel deep learning framework called SecProGNN, designed to predict secretory proteins in BALF. Firstly, SecProGNN represented proteins as graph-structured data, with amino acids connected based on their interactions. Then, these graphs were processed through graph neural networks (GNNs) model to extract graph features. Finally, the extracted feature vectors were fed into a multi-layer perceptron (MLP) module to predict BALF secreted proteins. Additionally, by utilizing SecProGNN, we investigated potential biomarkers for lung adenocarcinoma and identified 16 promising candidates that may be secreted into BALF.

PMID:40042949 | DOI:10.1109/JBHI.2025.3548263

STAR Protoc. 2025 Mar 4;6(1):103665. doi: 10.1016/j.xpro.2025.103665. Online ahead of print.

ABSTRACT

Here, we present a step-by step protocol to visualize intracellular aggregations in Arabidopsis mutants with cell wall secretion defects using FM4-64, a lipophilic styryl dye. We describe steps for growing seedlings, staining them with FM4-64, and identifying intracellular aggregates in cell wall synthesis and/or secretion mutants in root and hypocotyl epidermal cells via confocal microscopy. Additionally, we provide troubleshooting suggestions for common pitfalls. For complete details on the use and execution of this protocol, please refer to Hoffmann and McFarlane.1.

PMID:40042968 | DOI:10.1016/j.xpro.2025.103665

BMJ Open Respir Res. 2023 Apr 6;10(1):e001342. doi: 10.1136/bmjresp-2022-001342.

ABSTRACT

INTRODUCTION: Immune checkpoint inhibitors (ICI) have improved outcomes for patients with many malignancies. However, these treatments are associated with immune-related adverse events, including pulmonary toxicity (pneumonitis). Pneumonitis is associated with significant short-term morbidity and mortality, but long-term outcomes are not well described.

METHODS: We used the Vanderbilt Synthetic Derivative, a deidentified electronic health record database of >2.5 million patients seen at Vanderbilt, to identify patient charts that included treatment with pembrolizumab, nivolumab, ipilimumab, ipilimumab and nivolumab, atezolizumab or durvalumab by keyword search and ICD-10 codes for acute respiratory failure and/or bronchoalveolar lavage. We manually reviewed these charts and identified 78 subjects who met criteria for probable pneumonitis which included patients presenting with symptoms (dyspnoea, hypoxia, cough) and/or CT imaging consistent with this diagnosis. We collected data on demographics, ICI regimen, hospital admissions and long-term survival.

RESULTS: Of the 78 patients (48 males; median age 64 (range 28-81)), 52 patients required at least 1 hospital admission related to pneumonitis. A total of 25 patients experienced poor short-term outcomes (including 6 referred to hospice, 11 discharged to rehabilitation and 9 deaths). There was no association with these outcomes by patient age (p=0.96), sex (p=0.60), smoking status (p=0.63) or cancer type (p=0.13). Median duration of follow-up was 8.3 months (range 0.2-110.6 months), and 29 patients (37%) were alive at last follow-up. Patients admitted to the hospital were more likely to die (p=0.002) and less likely to receive additional treatment (p<0.0001) or survive for ≥12 months with no evidence of disease (p=0.02). There were no differences in long-term outcomes for patients with underlying pulmonary comorbidities.

DISCUSSION: ICI-pneumonitis has a high likelihood of causing hospitalisation and poor outcomes, including death. While there appears to be no difference in outcomes for patients with underling pulmonary comorbidities, those requiring admission have worse outcomes.

PMID:40042935 | DOI:10.1136/bmjresp-2022-001342

Endocr Connect. 2025 Feb 1:EC-24-0082. doi: 10.1530/EC-24-0082. Online ahead of print.

ABSTRACT

INTRODUCTION: We previously observed that continuous subcutaneous gonadotropin infusion (CSGI) in infants with congenital hypogonadotropic hypogonadism (CHH) can mimic minipuberty.

OBJECTIVE: to describe the early adolescence outcome of boys treated during the first year of life.

METHODS: In this retrospective cohort study, we describe 11 CHH boys aged 12 years [11.5-14.6] treated at the age of 4.5 months [2.0-11] with CSGI. To caompare we report testicular function of 12 untreated CHH boys aged 12 years [12-15.9].

RESULTS: In response to CSGI, serum testosterone and inhibinB levels increased from 0.03 ng/mL [0-0.07] to 2.25 ng/mL [1.12-3.86] and from 73 [11-173] to 401 [185-727] pg/mL, respectively. Testicular volume increased from 0.50 mL [0.5-1] to 1.50 mL [0.7-3]. Between end of CSGI and early adolescence, testicular volume in the treated group decreased from 1.5 mL [0.7-3] to 1.05 ml [0.7-2.36] (p=0.024) and differed from that in untreated boys (0.3 mL [0.13-1.3]). Hormone levels were higher in the treated group : serum AMH and inhibin B levels in treated patients decreased from 1028 pmol/l [550-1750] and 356 [185-727] pg/mL at neonatal period to 331 pmol/l [85-479] and 68 pg/ml [19-239] respectively at early adolescence and differed from those in untreated patients (57.5 [30-169] and 8 pg/ml [<5-37] (p<0.001)).

CONCLUSION: We report the first long-term follow-up of CHH boys treated with CSGI in infancy. Our results shown that the CSGI treatment resulted in higher inhibin B, AMH levels and testicular volume at early adolescence age. Follow-up should be continued until the end of puberty to assess spermatogenesis.

PMID:40042209 | DOI:10.1530/EC-24-0082

Ann Am Thorac Soc. 2025 Mar 5. doi: 10.1513/AnnalsATS.202408-831OC. Online ahead of print.

ABSTRACT

RATIONALE: Since the approval of elexacaftor/tezacaftor/ivacaftor (ETI), data suggests there have been changes in the management of pulmonary exacerbations (PEx) of Cystic Fibrosis (CF).

OBJECTIVE: Given the subjective nature of PEx diagnosis and management, we sought to characterize provider PEx management practices in people with CF (pwCF) prescribed highly effective modulator therapy (HEMT) and to identify practice changes that may impact clinical outcomes.

METHODS: We conducted semi-structured qualitative interviews amongst clinicians in the United States (US) in late 2021 to 2022 to investigate changes in the management of PEx in pwCF prescribed ETI. Inductive coding of transcripts was utilized in a thematic analysis.

RESULTS: We conducted 19 qualitative interviews with providers at 15 CF centers. Thematic analysis identified five themes regarding the presentation, diagnosis, and management of PEx in pwCF prescribed ETI: (1) PEx have changed in the era of HEMT to become a more subtle pathology that may result in providers questioning PEx diagnosis; (2) providers feel less anxious about clinical outcomes after PEx; (3) providers are expanding their assessment of PEx in the era of HEMT to identify more subtle PEx phenotypes; (4) pwCF are driving their care during PEx more than in the pre-HEMT era, with interviewees reporting that some "patients don't really contact us [with mild PEx symptoms]…we hear about it in retrospect." Interviewees expressed concern that this may result in more severe PEx; (5) provider management is less aggressive in the post-HEMT era, reflecting reduced PEx severity. Participants emphasized that their approach to PEx in general is unchanged and that "[providers] treat depending on severity…and the background of the patient." Interviewees reported they increasingly recommend maintenance therapies for PEx treatment before prescribing antibiotics.

CONCLUSIONS: Participants report that PEx in pwCF prescribed ETI appear milder, resulting in less anxiety about outcomes and a more conservative approach to management. Providers express uncertainty regarding the diagnosis of PEx given its evolving presentation and reduced in-person evaluation. Further research is necessary to identify sensitive markers of PEx and to assess the impact of conservative management on clinical outcomes.

PMID:40042492 | DOI:10.1513/AnnalsATS.202408-831OC

mBio. 2025 Mar 5:e0388324. doi: 10.1128/mbio.03883-24. Online ahead of print.

ABSTRACT

Cystic fibrosis (CF) is a genetic disease caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene, resulting in CFTR protein dysfunction. CFTR dysfunction has multi-organ consequences, leading to dehydrated mucus that is adherent to epithelia. In the lungs, this leads to recalcitrant infections with bacteria such as Pseudomonas aeruginosa. In the gut, mucus-laden feces can adhere to the intestines, resulting in distal intestinal obstruction syndrome (DIOS). There is limited information on how lung colonization and DIOS are correlated in people with CF (pwCF). In this novel work, we describe the development of spontaneous lung colonization of CF pathogens in young (<3 months old) CF rats, preceding the development of DIOS. Once DIOS is established, the lung microbiome becomes predominated by taxa also observed in the feces. Induced infection with P. aeruginosa in the CF rats reflects data found in pwCF, as once CF rats are infected, they retain a higher relative abundance of P. aeruginosa than their healthy agemates. Finally, we found that ivacaftor treatment favors a healthier gut microbiome in CF rats, decreasing the relative abundance of Escherichia coli. These results indicate that the CF rat model is recapitulative of human CF disease with the spontaneous lung colonization of traditional CF pathogens and maintenance of P. aeruginosa after induced infection. Furthermore, these results indicate a possible role for the gut-lung axis in lung colonization and DIOS in CF.IMPORTANCEThese data describe for the first time the development of spontaneous lung colonization in the cystic fibrosis (CF) rat model, a hallmark aspect of human CF disease. We also find that CF rats infected with Pseudomonas aeruginosa maintain higher relative abundance following chronic infection as compared to healthy rats, similar to those is seen in people with CF. Additionally, we describe the possible contribution of the gut-lung axis linking lung health with distal intestinal obstruction syndrome, a relationship largely unexplored in the context of CF.

PMID:40042272 | DOI:10.1128/mbio.03883-24

Pediatr Pulmonol. 2025 Mar;60(3):e71035. doi: 10.1002/ppul.71035.

NO ABSTRACT

PMID:40042147 | DOI:10.1002/ppul.71035

Pediatr Pulmonol. 2025 Mar;60(3):e71024. doi: 10.1002/ppul.71024.

ABSTRACT

BACKGROUND: Systemic antibiotics can impact all microbes inhabiting patients, regardless of the intended target organism(s). We studied the simultaneous effects on respiratory and fecal microbiomes of β-lactam antibiotics administered for respiratory symptoms in infants with cystic fibrosis (IWCF).

OBJECTIVE: To compare the magnitude and duration of intended (respiratory) and unintended (fecal) antimicrobial action by analyzing oropharyngeal (OP) and fecal microbiota in IWCF.

DESIGN: Shotgun metagenomic sequencing and qPCR were performed on OP and fecal samples collected longitudinally from 14 IWCF (ages 1-17 months) during ("On Antibiotics") and after ("Off Antibiotics") β-lactam therapy, and from 5 IWCF (3-16 months) never treated with antibiotics.

RESULTS: Total bacterial loads (TBL) for On Antibiotics samples were lower than for both Never (OP and fecal) and Off Antibiotics samples (fecal only). α-diversities (within-sample) for OP On Antibiotics samples were lower than for Never and Off Antibiotics samples but did not differ between fecal sample groups. β-diversity (between-sample) differed between all OP sample groups and between fecal On and Never Antibiotics and Off and Never antibiotics samples; however, fecal On and Off Antibiotics sample β-diversities did not differ. Patterns of change in antibiotic resistance gene abundances reflected shifts in microbial community composition.

CONCLUSIONS: β-lactam antibiotic exposure was followed by marked alterations in both OP and fecal microbiota. While microbiota appeared to rebound after treatment in both sample types, our results suggest that fecal microbiota recovered less than OP. The clinical consequences of these findings should be studied in IWCF and other populations frequently treated with antibiotics.

PMID:40042126 | DOI:10.1002/ppul.71024

Elife. 2025 Mar 5;13:RP98485. doi: 10.7554/eLife.98485.

ABSTRACT

Quantitative information about synaptic transmission is key to our understanding of neural function. Spontaneously occurring synaptic events carry fundamental information about synaptic function and plasticity. However, their stochastic nature and low signal-to-noise ratio present major challenges for the reliable and consistent analysis. Here, we introduce miniML, a supervised deep learning-based method for accurate classification and automated detection of spontaneous synaptic events. Comparative analysis using simulated ground-truth data shows that miniML outperforms existing event analysis methods in terms of both precision and recall. miniML enables precise detection and quantification of synaptic events in electrophysiological recordings. We demonstrate that the deep learning approach generalizes easily to diverse synaptic preparations, different electrophysiological and optical recording techniques, and across animal species. miniML provides not only a comprehensive and robust framework for automated, reliable, and standardized analysis of synaptic events, but also opens new avenues for high-throughput investigations of neural function and dysfunction.

PMID:40042890 | DOI:10.7554/eLife.98485

MAGMA. 2025 Mar 5. doi: 10.1007/s10334-025-01233-7. Online ahead of print.

ABSTRACT

OBJECTIVE: This study presents a novel deep learning-based framework for precise brain MR region segmentation, aiming to identify the location and the shape details of different anatomical structures within the brain.

MATERIALS AND METHODS: The approach uses a two-stage 3D segmentation technique on a dataset of adult subjects, including cognitively normal participants and individuals with cognitive decline. Stage 1 employs a 3D U-Net to segment 13 brain regions, achieving a mean DSC of 0.904 ± 0.060 and a mean HD95 of 1.52 ± 1.53 mm (a mean DSC of 0.885 ± 0.065 and a mean HD95 of 1.57 ± 1.35 mm for smaller parts). For challenging regions like hippocampus, thalamus, cerebrospinal fluid, amygdala, basal ganglia, and corpus callosum, Stage 2 with SegResNet refines segmentation, improving mean DSC to 0.921 ± 0.048 and HD95 to 1.17 ± 0.69 mm.

RESULTS: Statistical analysis reveals significant improvements (p-value < 0.001) for these regions, with DSC increases ranging from 1.3 to 3.2% and HD95 reductions of 0.06-0.33 mm. Comparisons with recent studies highlight the superior performance of the performed method.

DISCUSSION: The inclusion of a second stage for refining the segmentation of smaller regions demonstrates substantial improvements, establishing the framework's potential for precise and reliable brain region segmentation across diverse cognitive groups.

PMID:40042762 | DOI:10.1007/s10334-025-01233-7