Hasil untuk "Cytology"

W. Wolberg, O. Mangasarian

J. Goldman, A. Murr, R. Cooper

Y. Nikiforov, S. Carty, S. Chiosea et al.

J. Dillner, M. Rebolj, P. Birembaut et al.

Objective To obtain large scale and generalisable data on the long term predictive value of cytology and human papillomavirus (HPV) testing for development of cervical intraepithelial neoplasia grade 3 or cancer (CIN3+). Design Multinational cohort study with joint database analysis. Setting Seven primary HPV screening studies in six European countries. Participants 24 295 women attending cervical screening enrolled into HPV screening trials who had at least one cervical cytology or histopathology examination during follow-up. Main outcome measure Long term cumulative incidence of CIN3+. Results The cumulative incidence rate of CIN3+ after six years was considerably lower among women negative for HPV at baseline (0.27%, 95% confidence interval 0.12% to 0.45%) than among women with negative results on cytology (0.97%, 0.53% to 1.34%)). By comparison, the cumulative incidence rate for women with negative cytology results at the most commonly recommended screening interval in Europe (three years) was 0.51% (0.23% to 0.77%). The cumulative incidence rate among women with negative cytology results who were positive for HPV increased continuously over time, reaching 10% at six years, whereas the rate among women with positive cytology results who were negative for HPV remained below 3%. Conclusions A consistently low six year cumulative incidence rate of CIN3+ among women negative for HPV suggests that cervical screening strategies in which women are screened for HPV every six years are safe and effective.

H. Katki, W. Kinney, B. Fetterman et al.

Zhen Qiu, Kaiwen Xiao, Zhengwei Lu et al.

We present Singpath-VL, a vision-language large model, to fill the vacancy of AI assistant in cervical cytology. Recent advances in multi-modal large language models (MLLMs) have significantly propelled the field of computational pathology. However, their application in cytopathology, particularly cervical cytology, remains underexplored, primarily due to the scarcity of large-scale, high-quality annotated datasets. To bridge this gap, we first develop a novel three-stage pipeline to synthesize a million-scale image-description dataset. The pipeline leverages multiple general-purpose MLLMs as weak annotators, refines their outputs through consensus fusion and expert knowledge injection, and produces high-fidelity descriptions of cell morphology. Using this dataset, we then fine-tune the Qwen3-VL-4B model via a multi-stage strategy to create a specialized cytopathology MLLM. The resulting model, named Singpath-VL, demonstrates superior performance in fine-grained morphological perception and cell-level diagnostic classification. To advance the field, we will open-source a portion of the synthetic dataset and benchmark.

Khushboo Singh, Roon Banerjee, Chandrakanta Potdar et al.

Mutations in leucine-rich repeat kinase 2 (LRRK2) are among the most common genetic causes of Parkinson’s disease (PD), yet substantial heterogeneity exists among pathogenic variants. How mutations in distinct functional domains of LRRK2 differentially perturb cellular homeostasis remains incompletely understood. Here, we compared two pathogenic LRRK2 mutations—G2019S in the kinase domain and I1371V in the GTPase domain—across multiple cellular models, including SH-SY5Y and U87 cells, and healthy human iPSC-derived floor plate cells. We demonstrate that the I1371V mutation induces markedly more severe cellular dysfunction than G2019S. I1371V-expressing cells exhibited elevated LRRK2 autophosphorylation at S1292 and robust hyperphosphorylation of Rab8A and Rab10, indicating enhanced downstream signaling. These alterations impaired sterol trafficking, leading to selective depletion of membrane cholesterol without changes in total cellular cholesterol. Consequently, I1371V cells displayed increased membrane fluidity, disrupted microdomain organization, altered membrane topology, reduced caveolin-1 expression, and impaired dopamine transporter surface expression and dopamine uptake. Lipidomic profiling further revealed a broad disruption of lipid homeostasis, including reductions in cholesteryl esters, sterols, sphingolipids, and glycerophospholipids, whereas G2019S cells showed comparatively modest changes. Pharmacological intervention revealed mutation-specific responses, with the non-selective LRRK2 modulator GW5074 outperforming the kinase-selective inhibitor MLi-2 in restoring Rab8A phosphorylation, membrane integrity, and dopaminergic function. Collectively, these findings identify membrane lipid dysregulation as a central cell biological mechanism in LRRK2-associated PD and underscore the importance of variant-specific therapeutic strategies.

F. Nardi, F. Basolo, A. Crescenzi et al.

Hao Jiang, Cheng Jin, Huangjing Lin et al.

Cervical cancer is a leading malignancy in female reproductive system. While AI-assisted cytology offers a cost-effective and non-invasive screening solution, current systems struggle with generalizability in complex clinical scenarios. To address this issue, we introduced Smart-CCS, a generalizable Cervical Cancer Screening paradigm based on pretraining and adaptation to create robust and generalizable screening systems. To develop and validate Smart-CCS, we first curated a large-scale, multi-center dataset named CCS-127K, which comprises a total of 127,471 cervical cytology whole-slide images collected from 48 medical centers. By leveraging large-scale self-supervised pretraining, our CCS models are equipped with strong generalization capability, potentially generalizing across diverse scenarios. Then, we incorporated test-time adaptation to specifically optimize the trained CCS model for complex clinical settings, which adapts and refines predictions, improving real-world applicability. We conducted large-scale system evaluation among various cohorts. In retrospective cohorts, Smart-CCS achieved an overall area under the curve (AUC) value of 0.965 and sensitivity of 0.913 for cancer screening on 11 internal test datasets. In external testing, system performance maintained high at 0.950 AUC across 6 independent test datasets. In prospective cohorts, our Smart-CCS achieved AUCs of 0.947, 0.924, and 0.986 in three prospective centers, respectively. Moreover, the system demonstrated superior sensitivity in diagnosing cervical cancer, confirming the accuracy of our cancer screening results by using histology findings for validation. Interpretability analysis with cell and slide predictions further indicated that the system's decision-making aligns with clinical practice. Smart-CCS represents a significant advancement in cancer screening across diverse clinical contexts.

Swarn Singh Warshaneyan, Maksims Ivanovs, Blaž Cugmas et al.

This study explores the application of deep learning to improve and automate pollen grain detection and classification in both optical and holographic microscopy images, with a particular focus on veterinary cytology use cases. We used YOLOv8s for object detection and MobileNetV3L for the classification task, evaluating their performance across imaging modalities. The models achieved 91.3% mAP50 for detection and 97% overall accuracy for classification on optical images, whereas the initial performance on greyscale holographic images was substantially lower. We addressed the performance gap issue through dataset expansion using automated labeling and bounding box area enlargement. These techniques, applied to holographic images, improved detection performance from 2.49% to 13.3% mAP50 and classification performance from 42% to 54%. Our work demonstrates that, at least for image classification tasks, it is possible to pair deep learning techniques with cost-effective lensless digital holographic microscopy devices.

Muhammed Furkan Dasdelen, Hyesu Lim, Michele Buck et al.

Sparse autoencoders (SAEs) emerged as a promising tool for mechanistic interpretability of transformer-based foundation models. Very recently, SAEs were also adopted for the visual domain, enabling the discovery of visual concepts and their patch-wise attribution to tokens in the transformer model. While a growing number of foundation models emerged for medical imaging, tools for explaining their inferences are still lacking. In this work, we show the applicability of SAEs for hematology. We propose CytoSAE, a sparse autoencoder which is trained on over 40,000 peripheral blood single-cell images. CytoSAE generalizes to diverse and out-of-domain datasets, including bone marrow cytology, where it identifies morphologically relevant concepts which we validated with medical experts. Furthermore, we demonstrate scenarios in which CytoSAE can generate patient-specific and disease-specific concepts, enabling the detection of pathognomonic cells and localized cellular abnormalities at the patch level. We quantified the effect of concepts on a patient-level AML subtype classification task and show that CytoSAE concepts reach performance comparable to the state-of-the-art, while offering explainability on the sub-cellular level. Source code and model weights are available at https://github.com/dynamical-inference/cytosae.

Rayyah R. Alkhanjari, Maitha M. Alhajeri, Poorna Manasa Bhamidimarri et al.

Abstract Septins are GTP-binding cytoskeletal proteins primarily known to be involved in cell division, membrane remodeling, and cytoskeletal organization. In the nervous system, septins are suggested as key regulators of neural development, including neurite outgrowth, spine morphology, and axon initial segment formation. Septins are localized to specialized membrane domains, such as dendritic spines, axon initial segments, and synaptic terminals, where they function as scaffolding components and diffusion barriers. They are abundant in neurons, oligodendrocytes, Schwann cells, and astrocytes, regulating processes like myelination and synaptic organization. In neuronal cells, specific septin isoforms such as SEPT3, SEPT5, and SEPT7 contribute to dendritic spine formation, neurotransmitter vesicle trafficking, and axonal integrity. Alterations in septin expression or assembly can disrupt synaptic architecture and neuroplasticity, emphasizing their role in neuronal homeostasis. Dysregulation of septin expression and function has been implicated in a range of neurological disorders, including demyelinating diseases like Multiple Sclerosis and Hereditary Neuralgic Amyotrophy. Abnormal septin aggregation has been observed in neurodegenerative diseases such as Alzheimer's and Parkinson's disease. Moreover, septins can modulate inflammatory responses, where antibodies for septins 5 and 7 were associated with autoimmune encephalitis conditions. This review will provide a comprehensive overview of the role of septins in the nervous system, focusing on their molecular mechanisms, cellular functions, and implications in neurological disorders.

Jiaoyun Zheng, Jian Gong

Due to the difficulty in early diagnosis and the lack of treatment for advanced disease, the mortality rate of hepatocellular carcinoma (HCC) is high, and the 5-year overall survival rate is low at present. SLC1A4 is a neutral amino acid transporter, but its regulatory role and mechanism in HCC are still unclear. Through analyzing the TCGA database and clinical tissue specimens, this study uncovered the high expression of SLC1A4 in tumor tissues of HCC. Worse more, a high level of SLC1A4 may lead to a poor prognosis of HCC. Mechanically, silencing SLC1A4 inhibited the phosphorylation activation of AKT by suppressing the ubiquitin modification of AKT at lysine 63 and amino acid influx represented by D-serine, decreasing the protein level of β-catenin in the cell nucleus and suppressing the transcriptional activity of c-Myc and EpCAM promoters. As a result, silencing SLC1A4 inhibited the proliferation, migration, and stemness of hepatic cancer cells, which was successfully reversed by the introduction of exogenous AKT. Moreover, epithelial–mesenchymal transition (EMT) in vitro and metastasis potential in vivo of hepatic cancer cells was suppressed by the downregulated SLC1A4 level. In conclusion, SLC1A4 promotes the malignant transformation of HCC through activating signal transduction mediated by AKT. The findings in this study suggested that SLC1A4 may be a diagnostic indicator for the early HCC and a therapeutic target for the advanced HCC.

Onella Athnaiel, Nicholas Davidson, Jaskaran Mangat et al.

Chronic pain, pain that lasts beyond three months, is a common finding in the elderly. It is often due to musculoskeletal conditions but can be precipitated by other factors as well. While physiological systems decline with aging, chronic pain is influenced by changes in hormone profiles as men and women enter into andropause and menopause, respectively. Research on gonadal hormones is limited, especially when it comes to their relationship with chronic pain. Women tend to experience less pain with aging compared to their premenopausal years, and this is partially explained by the fact that estrogen enhances pain sensitivity and its decline during menopause decreases it. However, hormone replacement therapy (HRT) seems to increase pain tolerance post-menopause. There is some evidence that testosterone plays a protective factor in pain perception. Men on the other hand, have higher pain tolerance as testosterone is considered to be a protective factor. With aging and decreasing testosterone, older men tend to be less tolerant to pain. This paper explores how hormonal changes with aging impact pain perception in both men and women, highlighting several pain conditions influenced by hormones. Although research remains limited, the potential of HRT as a treatment for common pain conditions is examined.

María Cristina Franco-Arellanes, Perla Xóchitl Toledo-Valdes, Cynthia Díaz-Hernández et al.

INTRODUCTION: O-GlcNAcylation is a post-translational modification in which a single N-Acetyl-D-Glucosamine (GlcNAc) molecule is added to Ser or Thr residues of proteins. The O-N-acetylglucosaminyl transferase (OGT) enzyme is responsible for adding GlcNAc to the target proteins and N-acetyl-β-D-glucosaminidase (OGA) that removes the GlcNAc residue. O-GlcNAcylation has been described in the pathophysiology of several diseases; however, little has been studied in dental tissue. The aim of the present work is to characterise the product of O-GlcNAcylation and its enzymes at the tissue level in the dental pulp, as well as its expression in dental pulp stem cells (DPSCs) both in situ and in vitro. This enables the recognition of the behaviour of O-GlcNAcylation in pulp tissue without pathology. MATERIAL AND METHODS: Pulp tissue was obtained from 10 healthy donors, and the expression of O-GlcNAc, OGT, and OGA was analysed using immunofluorescence with specific antibodies in different regions of the dental pulp. DPSCs were isolated, cultured, and identified with anti-STRO1 (antibody specific for human CD34+ cells, useful for DPSC identification). The expression of O-GlcNAc in DPSCs was confirmed in vitro through Western blot. Results. Different regions of the dental pulp and DPSCs express O-GlcNAc and the enzymes OGT and OGA. O-GlcNAc and OGT expression was more prominent in the odontoblastic layer, cell-rich zone, and in the central core. OGA was distributed throughout the pulp tissue with lower immunoreactivity compared to OGT. CONCLUSIONS: Our results suggest that O-GlcNAcylation may play a relevant role in human dental pulp homeostasis and in physiology of DPSCs.

Hao Jiang, Runsheng Liu, Yanning Zhou et al.

Cytology screening from Papanicolaou (Pap) smears is a common and effective tool for the preventive clinical management of cervical cancer, where abnormal cell detection from whole slide images serves as the foundation for reporting cervical cytology. However, cervical cell detection remains challenging due to 1) hazily-defined cell types (e.g., ASC-US) with subtle morphological discrepancies caused by the dynamic cancerization process, i.e., cell class ambiguity, and 2) imbalanced class distributions of clinical data may cause missed detection, especially for minor categories, i.e., cell class imbalance. To this end, we propose a holistic and historical instance comparison approach for cervical cell detection. Specifically, we first develop a holistic instance comparison scheme enforcing both RoI-level and class-level cell discrimination. This coarse-to-fine cell comparison encourages the model to learn foreground-distinguishable and class-wise representations. To emphatically improve the distinguishability of minor classes, we then introduce a historical instance comparison scheme with a confident sample selection-based memory bank, which involves comparing current embeddings with historical embeddings for better cell instance discrimination. Extensive experiments and analysis on two large-scale cytology datasets including 42,592 and 114,513 cervical cells demonstrate the effectiveness of our method. The code is available at https://github.com/hjiangaz/HERO.

Kanghyun Kim, Amey Chaware, Clare B. Cook et al.

Optical microscopy has long been the standard method for diagnosis in cytopathology. Whole slide scanners can image and digitize large sample areas automatically, but are slow, expensive and therefore not widely available. Clinical diagnosis of cytology specimens is especially challenging since these samples are both spread over large areas and thick, which requires 3D capture. Here, we introduce a new parallelized microscope for scanning thick specimens across extremely wide fields-of-view (54x72 mm^2) at 1.2 and 0.6 μm resolutions, accompanied by machine learning software to rapidly assess these 16 gigapixel scans. This Multi-Camera Array Scanner (MCAS) comprises 48 micro-cameras closely arranged to simultaneously image different areas. By capturing 624 megapixels per snapshot, the MCAS is significantly faster than most conventional whole slide scanners. We used this system to digitize entire cytology samples (scanning three entire slides in 3D in just several minutes) and demonstrate two machine learning techniques to assist pathologists: first, an adenocarcinoma detection model in lung specimens (0.73 recall); second, a slide-level classification model of lung smears (0.969 AUC).

Townim Faisal Chowdhury, Kewen Liao, Vu Minh Hieu Phan et al.

Deep Neural Networks (DNNs) are widely used for visual classification tasks, but their complex computation process and black-box nature hinder decision transparency and interpretability. Class activation maps (CAMs) and recent variants provide ways to visually explain the DNN decision-making process by displaying 'attention' heatmaps of the DNNs. Nevertheless, the CAM explanation only offers relative attention information, that is, on an attention heatmap, we can interpret which image region is more or less important than the others. However, these regions cannot be meaningfully compared across classes, and the contribution of each region to the model's class prediction is not revealed. To address these challenges that ultimately lead to better DNN Interpretation, in this paper, we propose CAPE, a novel reformulation of CAM that provides a unified and probabilistically meaningful assessment of the contributions of image regions. We quantitatively and qualitatively compare CAPE with state-of-the-art CAM methods on CUB and ImageNet benchmark datasets to demonstrate enhanced interpretability. We also test on a cytology imaging dataset depicting a challenging Chronic Myelomonocytic Leukemia (CMML) diagnosis problem. Code is available at: https://github.com/AIML-MED/CAPE.

Cheng Jin, Luyang Luo, Huangjing Lin et al.

Fine-grained classification of whole slide images (WSIs) is essential in precision oncology, enabling precise cancer diagnosis and personalized treatment strategies. The core of this task involves distinguishing subtle morphological variations within the same broad category of gigapixel-resolution images, which presents a significant challenge. While the multi-instance learning (MIL) paradigm alleviates the computational burden of WSIs, existing MIL methods often overlook hierarchical label correlations, treating fine-grained classification as a flat multi-class classification task. To overcome these limitations, we introduce a novel hierarchical multi-instance learning (HMIL) framework. By facilitating on the hierarchical alignment of inherent relationships between different hierarchy of labels at instance and bag level, our approach provides a more structured and informative learning process. Specifically, HMIL incorporates a class-wise attention mechanism that aligns hierarchical information at both the instance and bag levels. Furthermore, we introduce supervised contrastive learning to enhance the discriminative capability for fine-grained classification and a curriculum-based dynamic weighting module to adaptively balance the hierarchical feature during training. Extensive experiments on our large-scale cytology cervical cancer (CCC) dataset and two public histology datasets, BRACS and PANDA, demonstrate the state-of-the-art class-wise and overall performance of our HMIL framework. Our source code is available at https://github.com/ChengJin-git/HMIL.

Mengqing Ma, Rui Hu, Qi Huang et al.

Abstract Hepatocellular carcinoma (HCC) is recognized as a highly malignant tumor. Targeted combination immunotherapy, the initially approved regimen, is compromised by adverse side effects and low response rates during clinical treatment. Traditional Chinese medicine and its derived natural compounds, known for their anticancer effects, offer advantages of low toxicity and cost. In this study, we performed high-throughput phenotypic screening in vitro to identify promising anti-HCC drugs. Among 1,444 bioactive compounds, digoxigenin (DIG) was found to significantly impede HCC cell progression. We validated DIG’s therapeutic effects through assays such as cell counting by CCK8, lactate dehydrogenase, and colony formation. Analyses including transmission electron microscopy, western blotting, and immunofluorescence demonstrated that DIG inhibits HCC cell proliferation via autophagy. Network pharmacology and molecular docking studies suggest that DIG targets the PI3K/AKT/mTOR signaling pathway. Comparative treatments of Hep3B and Huh7 cells with DIG or mTOR inhibitors revealed similar inhibitory impacts, indicating that DIG induces autophagy by inhibiting the PI3K/AKT/mTOR pathway. In vivo studies confirmed that DIG halts the growth of subcutaneous xenograft tumors. In conclusion, DIG represents a potential HCC treatment by modulating the PI3K/AKT/mTOR pathway to induce autophagy. This research, via phenotypic screening, accelerates drug discovery and the development of novel therapies targeting the underlying mechanisms of liver cancer.