Computer methods and programs in biomedicine update最新文献

Computer methods and programs in biomedicine update

Pub Date : 2026-06-01 Epub Date: 2026-01-07 DOI: 10.1016/j.cmpbup.2025.100229

Sarah L. Alzamili , Salwa Shakir Baawi , Mustafa Noaman Kadhim , Dhiah Al-Shammary , Ayman Ibaida , Khandakar Ahmed

This paper aims to introduce a novel clustering method for electroencephalogram (EEG) based on Ruzicka mathematical similarity and incorporates Particle Swarm Optimization (PSO) to enhance feature selection. Medical datasets often contain both convergent and divergent features, making feature selection a crucial step for accurate disease diagnosis and public health applications. The proposed Ruzicka-based clustering method groups EEG records into non-overlapping subgroups according to a defined similarity metric. Cluster centers are determined using a polynomial-based calculation, after which EEG records are assigned to clusters based on the Ruzicka similarity measure. After clustering the EEG records into highly coherent groups, PSO algorithm is employed to identify the most effective subset of features. This process enhances classification accuracy and contributes to more reliable diagnostic outcomes by combining clustering with feature selection. The selected features are then evaluated using multiple classifiers, including Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), K-Nearest Neighbors (KNN), and Naive Bayes (NB). Accuracy, recall, f1-score and precision measures are conducted to evaluate the model’s performance. Experimental validation is carried out on the Bonn University EEG dataset. With both RF and NB classifiers, the proposed model has achieved up to 100% accuracy compared to other models. The proposed method can be implemented in medical organizations as a decision-support system to assist healthcare professionals in analyzing EEG patterns. Its integration can enhance the accuracy and efficiency of disease diagnosis, leading to improved patient care.

本文提出了一种基于Ruzicka数学相似度的脑电图聚类方法，并结合粒子群算法（PSO）增强特征选择。医疗数据集通常包含收敛和发散特征，这使得特征选择成为准确疾病诊断和公共卫生应用的关键步骤。提出的基于ruzicka的聚类方法根据定义的相似度度量将EEG记录划分为不重叠的子组。使用基于多项式的计算确定聚类中心，然后根据Ruzicka相似性度量将EEG记录分配到聚类中。将EEG记录聚类成高度相干的组后，采用粒子群算法识别最有效的特征子集。该过程通过将聚类与特征选择相结合，提高了分类精度，并有助于获得更可靠的诊断结果。然后使用多个分类器对所选特征进行评估，包括支持向量机（SVM）、决策树（DT）、随机森林（RF）、k近邻（KNN）和朴素贝叶斯（NB）。采用准确性、召回率、f1分数和精度指标来评估模型的性能。在波恩大学EEG数据集上进行了实验验证。使用RF和NB分类器，与其他模型相比，所提出的模型达到了100%的准确率。该方法可作为决策支持系统在医疗机构中实现，以帮助医疗保健专业人员分析脑电图模式。它的整合可以提高疾病诊断的准确性和效率，从而改善患者的护理。

{"title":"Ruzicka similarity-based brain EEG clustering for improved intelligent epilepsy diagnosis","authors":"Sarah L. Alzamili , Salwa Shakir Baawi , Mustafa Noaman Kadhim , Dhiah Al-Shammary , Ayman Ibaida , Khandakar Ahmed","doi":"10.1016/j.cmpbup.2025.100229","DOIUrl":"10.1016/j.cmpbup.2025.100229","url":null,"abstract":"<div><div>This paper aims to introduce a novel clustering method for electroencephalogram (EEG) based on Ruzicka mathematical similarity and incorporates Particle Swarm Optimization (PSO) to enhance feature selection. Medical datasets often contain both convergent and divergent features, making feature selection a crucial step for accurate disease diagnosis and public health applications. The proposed Ruzicka-based clustering method groups EEG records into non-overlapping subgroups according to a defined similarity metric. Cluster centers are determined using a polynomial-based calculation, after which EEG records are assigned to clusters based on the Ruzicka similarity measure. After clustering the EEG records into highly coherent groups, PSO algorithm is employed to identify the most effective subset of features. This process enhances classification accuracy and contributes to more reliable diagnostic outcomes by combining clustering with feature selection. The selected features are then evaluated using multiple classifiers, including Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), K-Nearest Neighbors (KNN), and Naive Bayes (NB). Accuracy, recall, f1-score and precision measures are conducted to evaluate the model’s performance. Experimental validation is carried out on the Bonn University EEG dataset. With both RF and NB classifiers, the proposed model has achieved up to 100% accuracy compared to other models. The proposed method can be implemented in medical organizations as a decision-support system to assist healthcare professionals in analyzing EEG patterns. Its integration can enhance the accuracy and efficiency of disease diagnosis, leading to improved patient care.</div></div>","PeriodicalId":72670,"journal":{"name":"Computer methods and programs in biomedicine update","volume":"9 ","pages":"Article 100229"},"PeriodicalIF":0.0,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145977757","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Towards personalised biomechanical assessment of child birth safety; Automatic generation of personalised bony pelvis geometry by template mesh morphing 面向个性化的分娩安全生物力学评估通过模板网格变形自动生成个性化骨骨盆几何形状

Computer methods and programs in biomedicine update

Pub Date : 2026-06-01 Epub Date: 2026-01-29 DOI: 10.1016/j.cmpbup.2026.100233

Luděk Hynčík , Adam Wittek , Magdalena Jansová , Vít Nováček , Hana Čechová , Lucie Hájková Hympánová , Ladislav Krofta , Karol Miller

Background and objective

Pelvic floor muscle injuries associated with vaginal childbirth can result in pelvic organ disorders. Personalised biomechanical models offer a tool for predicting the risk of complications during childbirth. An important component of such computational models is a geometrically precise description of a pelvis. In this study, we developed an algorithm that automatically generates a discretised surface to define the personalised bony pelvis geometry of individual women from a template mesh.

Methods

We developed and implemented in 3D Slicer, a free open-source image computing software platform, the algorithm that applies radial basis function to morph the template mesh to the personalised geometry based on the bony pelvis landmarks identified in the target magnetic resonance image that depicts the analysed pelvis.

Results

We demonstrated the performance of our methods by automatically generating personalised bony pelvis meshes for six women. For quantitative evaluation, we used the Hausdorff distance (HD) and birth canal dimensions. The median HD was within two times the voxel dimension of the pelvis magnetic resonance (MR) images. The dimensions of the birth canal determined from the personalised meshes and from manually segmented MR images were, for practical purposes, undistinguishable.

Conclusions

Our algorithm generates a personalised bony pelvis model by mesh-morphing based on a template model and bony landmarks. Accuracy and performance of the algorithm were evaluated by morphing six bony pelves. Differences between the key birth canal dimensions derived from the personalised pelvis models and those obtained from pelvis MR images were within the inter-observer variation reported in the literature for MR pelvimetry measurements.

背景与目的阴道分娩相关的盆底肌肉损伤可导致盆腔器官紊乱。个性化的生物力学模型为预测分娩并发症的风险提供了一种工具。这种计算模型的一个重要组成部分是对骨盆的几何精确描述。在这项研究中，我们开发了一种算法，可以自动生成一个离散的表面，从模板网格中定义个体女性的个性化骨盆几何形状。方法在3D Slicer（免费的开源图像计算软件平台）中开发并实现了该算法，该算法基于在描述分析骨盆的目标磁共振图像中识别的骨盆骨地标，应用径向基函数将模板网格变形为个性化几何形状。结果我们通过对6名女性自动生成个性化骨盆骨网来展示我们方法的性能。为了进行定量评价，我们使用了Hausdorff距离（HD）和产道尺寸。中位HD是骨盆磁共振（MR）图像体素尺寸的两倍以内。从个性化网格和人工分割的MR图像中确定的产道尺寸实际上是无法区分的。结论sour算法基于模板模型和骨标记进行网格化，生成个性化骨盆骨模型。通过对6个骨瓣的变形来评估算法的准确性和性能。从个性化骨盆模型得出的关键产道尺寸与从骨盆MR图像获得的产道尺寸之间的差异在MR骨盆测量的文献中报道的观察者之间的差异。

{"title":"Towards personalised biomechanical assessment of child birth safety; Automatic generation of personalised bony pelvis geometry by template mesh morphing","authors":"Luděk Hynčík , Adam Wittek , Magdalena Jansová , Vít Nováček , Hana Čechová , Lucie Hájková Hympánová , Ladislav Krofta , Karol Miller","doi":"10.1016/j.cmpbup.2026.100233","DOIUrl":"10.1016/j.cmpbup.2026.100233","url":null,"abstract":"<div><h3>Background and objective</h3><div>Pelvic floor muscle injuries associated with vaginal childbirth can result in pelvic organ disorders. Personalised biomechanical models offer a tool for predicting the risk of complications during childbirth. An important component of such computational models is a geometrically precise description of a pelvis. In this study, we developed an algorithm that automatically generates a discretised surface to define the personalised bony pelvis geometry of individual women from a template mesh.</div></div><div><h3>Methods</h3><div>We developed and implemented in 3D Slicer, a free open-source image computing software platform, the algorithm that applies radial basis function to morph the template mesh to the personalised geometry based on the bony pelvis landmarks identified in the target magnetic resonance image that depicts the analysed pelvis.</div></div><div><h3>Results</h3><div>We demonstrated the performance of our methods by automatically generating personalised bony pelvis meshes for six women. For quantitative evaluation, we used the Hausdorff distance (HD) and birth canal dimensions. The median HD was within two times the voxel dimension of the pelvis magnetic resonance (MR) images. The dimensions of the birth canal determined from the personalised meshes and from manually segmented MR images were, for practical purposes, undistinguishable.</div></div><div><h3>Conclusions</h3><div>Our algorithm generates a personalised bony pelvis model by mesh-morphing based on a template model and bony landmarks. Accuracy and performance of the algorithm were evaluated by morphing six bony pelves. Differences between the key birth canal dimensions derived from the personalised pelvis models and those obtained from pelvis MR images were within the inter-observer variation reported in the literature for MR pelvimetry measurements.</div></div>","PeriodicalId":72670,"journal":{"name":"Computer methods and programs in biomedicine update","volume":"9 ","pages":"Article 100233"},"PeriodicalIF":0.0,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188806","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

NuDetect: A point annotation-based framework for nuclei detection using density estimation and conformal thresholding NuDetect：一个基于点注释的核检测框架，使用密度估计和保形阈值

Computer methods and programs in biomedicine update

Pub Date : 2026-06-01 Epub Date: 2025-12-18 DOI: 10.1016/j.cmpbup.2025.100225

Khaled Al-Thelaya , Nauman Ullah Gilal , Fahad Majeed , Mahmood Alzubaidi , Sabri Boughorbel , William Mifsud , Marco Agus , Jens Schneider

Whole Slide Imaging (WSI) generates vast data sets in histopathology. Manual annotation is impractical and time consuming. There is, thus, a dire need for effective analysis tools. However, a lack of annotated data hampers supervised learning of models that generalize well across domains. Point annotations have emerged as a practical remedy. Motivated by the fact that the randomness of the tissue slice angle and depth renders size measurements of nuclei — such as it would be provided by segmentation — meaningless (unlike in other medical tasks), point annotations are efficient and useful due to their sparseness. In this paper, we formulate the task of nuclei detection as a density estimation problem. We use a U-Net architecture with PoolFormer encoders as the basis to compute point-annotations for nuclei detection. Specifically, we use Gaussian kernels to generate target density masks from a segmented data set and use isocontouring to separate overlapping nuclei. We show that conformal prediction can compute a near-optimal threshold for contouring. This significantly enhances our detection rate. To address cross-domain generalization issues, our framework uses color normalization. As a result, our framework sets a new state-of-the-art in nucleus localization on both the PanNuke and MoNuSeg data sets, and we demonstrate our cross-domain generalization capabilities using samples of the TCGA data set.

全玻片成像（WSI）在组织病理学中产生大量数据集。手动注释是不切实际的，而且耗时。因此，迫切需要有效的分析工具。然而，缺乏带注释的数据阻碍了模型的监督学习，这些模型可以很好地跨领域泛化。点注释作为一种实用的补救措施出现了。由于组织切片角度和深度的随机性使得核的尺寸测量（如分割所提供的）变得毫无意义（与其他医学任务不同），点注释由于其稀疏性而变得高效和有用。在本文中，我们将核检测任务表述为密度估计问题。我们使用带有PoolFormer编码器的U-Net架构作为计算核检测点注释的基础。具体来说，我们使用高斯核从分割的数据集生成目标密度掩模，并使用等轮廓来分离重叠的核。我们证明保形预测可以计算出轮廓的近最优阈值。这大大提高了我们的检出率。为了解决跨域泛化问题，我们的框架使用颜色归一化。因此，我们的框架在PanNuke和MoNuSeg数据集上设置了新的核定位技术，并且我们使用TCGA数据集的样本展示了我们的跨域泛化能力。

{"title":"NuDetect: A point annotation-based framework for nuclei detection using density estimation and conformal thresholding","authors":"Khaled Al-Thelaya , Nauman Ullah Gilal , Fahad Majeed , Mahmood Alzubaidi , Sabri Boughorbel , William Mifsud , Marco Agus , Jens Schneider","doi":"10.1016/j.cmpbup.2025.100225","DOIUrl":"10.1016/j.cmpbup.2025.100225","url":null,"abstract":"<div><div>Whole Slide Imaging (WSI) generates vast data sets in histopathology. Manual annotation is impractical and time consuming. There is, thus, a dire need for effective analysis tools. However, a lack of annotated data hampers supervised learning of models that generalize well across domains. Point annotations have emerged as a practical remedy. Motivated by the fact that the randomness of the tissue slice angle and depth renders size measurements of nuclei — such as it would be provided by segmentation — meaningless (unlike in other medical tasks), point annotations are efficient and useful due to their sparseness. In this paper, we formulate the task of nuclei detection as a density estimation problem. We use a U-Net architecture with PoolFormer encoders as the basis to compute point-annotations for nuclei detection. Specifically, we use Gaussian kernels to generate target density masks from a segmented data set and use isocontouring to separate overlapping nuclei. We show that conformal prediction can compute a near-optimal threshold for contouring. This significantly enhances our detection rate. To address cross-domain generalization issues, our framework uses color normalization. As a result, our framework sets a new state-of-the-art in nucleus localization on both the PanNuke and MoNuSeg data sets, and we demonstrate our cross-domain generalization capabilities using samples of the TCGA data set.</div></div>","PeriodicalId":72670,"journal":{"name":"Computer methods and programs in biomedicine update","volume":"9 ","pages":"Article 100225"},"PeriodicalIF":0.0,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

USE-MiT: Attention-based model for breast ultrasound images segmentation USE-MiT：基于注意力的乳腺超声图像分割模型

Computer methods and programs in biomedicine update

Pub Date : 2026-06-01 Epub Date: 2026-01-02 DOI: 10.1016/j.cmpbup.2025.100226

Nadia Brancati, Maria Frucci

Early detection of breast cancer disease is crucial to enhancing patient outcomes through effective treatment. Ultrasound imaging, a simple, low-cost, and non-invasive technique, can help differentiate cystic from solid masses, mainly on the basis of the analysis of the detected anomalies’ boundaries. Automatic detection methods of mass boundaries in ultrasound images can reduce the dependence on the radiologist’s experience for this analysis. We propose USE-MiT, a segmentation method for breast ultrasound images, based on a UNet architecture in which the encoder and decoder modules are interfaced through a configuration based on Squeeze and Excitation Attention modules, and the encoder structure is represented by a Mix Transformer. The model was trained and validated, with a 4-fold cross-validation, on the Breast Ultrasound Image Dataset, and was tested on the independent dataset, namely Breast-Lesions-USG. The experiments have demonstrated the efficiency of the model, achieving an overall Dice of 0.88 and an IoU of 0.64, outperforming the state-of-the-art. The source code is available at https://github.com/nbrancati/USE-MiT.

早期发现乳腺癌疾病对于通过有效治疗提高患者预后至关重要。超声成像是一种简单、低成本、无创的技术，主要基于对检测到的异常边界的分析，可以帮助区分囊性肿块和实性肿块。超声图像中质量边界的自动检测方法可以减少对放射科医生经验的依赖。我们提出了一种基于UNet架构的乳腺超声图像分割方法USE-MiT，其中编码器和解码器模块通过基于挤压和激励注意模块的配置进行接口，编码器结构由Mix Transformer表示。在乳腺超声图像数据集上对模型进行4倍交叉验证和训练，并在独立数据集Breast- lesions - usg上进行测试。实验证明了该模型的效率，实现了0.88的总体Dice和0.64的IoU，优于最先进的技术。源代码可从https://github.com/nbrancati/USE-MiT获得。

{"title":"USE-MiT: Attention-based model for breast ultrasound images segmentation","authors":"Nadia Brancati, Maria Frucci","doi":"10.1016/j.cmpbup.2025.100226","DOIUrl":"10.1016/j.cmpbup.2025.100226","url":null,"abstract":"<div><div>Early detection of breast cancer disease is crucial to enhancing patient outcomes through effective treatment. Ultrasound imaging, a simple, low-cost, and non-invasive technique, can help differentiate cystic from solid masses, mainly on the basis of the analysis of the detected anomalies’ boundaries. Automatic detection methods of mass boundaries in ultrasound images can reduce the dependence on the radiologist’s experience for this analysis. We propose USE-MiT, a segmentation method for breast ultrasound images, based on a UNet architecture in which the encoder and decoder modules are interfaced through a configuration based on Squeeze and Excitation Attention modules, and the encoder structure is represented by a Mix Transformer. The model was trained and validated, with a 4-fold cross-validation, on the Breast Ultrasound Image Dataset, and was tested on the independent dataset, namely Breast-Lesions-USG. The experiments have demonstrated the efficiency of the model, achieving an overall Dice of 0<em>.</em>88 and an IoU of 0<em>.</em>64, outperforming the state-of-the-art. The source code is available at <span><span>https://github.com/nbrancati/USE-MiT</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":72670,"journal":{"name":"Computer methods and programs in biomedicine update","volume":"9 ","pages":"Article 100226"},"PeriodicalIF":0.0,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146037808","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

MR Optimum: A web-based open-source tool for standardized signal-to-noise ratio evaluation in MRI MR Optimum：一个基于网络的开源工具，用于MRI中标准化的信噪比评估

Computer methods and programs in biomedicine update

Pub Date : 2026-06-01 Epub Date: 2026-02-02 DOI: 10.1016/j.cmpbup.2026.100235

Eros Montin , Xuan Thao Nguyen , Riccardo Lattanzi

Signal-to-noise ratio (SNR) is a key performance metric in magnetic resonance imaging (MRI) to evaluate pulse sequences, receive coils, and image reconstruction algorithms. A variety of methods have been proposed to estimate SNR. However, the lack of consistent and broadly available open-source implementations has been a challenge for reliable SNR comparisons in clinical and research settings. To address this gap, this work introduces MR Optimum, a cloud-native, open-source platform for standardized SNR analysis. MR Optimum integrates established SNR estimation techniques within a flexible, modular software architecture. A web-based user interface supports data upload, task configuration, cloud computations, and real-time results visualization. MR Optimum leverages serverless computing technologies (AWS Lambda and Fargate) to perform scalable, event-driven processing of MRI rawdata and allow users to calculate SNR using established methods: multiple replicas, pseudo multiple replicas, generalized pseudo multiple replicas, and analytic methods. Results include SNR maps, noise covariance and noise coefficient matrices, coil sensitivity profiles, and g factor maps. The web interface enables interactive visualization and histogram analysis based on regions of interest. Results can be exported in MATLAB, NIfTI, and JSON formats. By providing a unified computational environment, MR Optimum ensures reproducibility, and democratizes access to state-of-the-art SNR estimation, promoting multi-center harmonization and quality assurance.

信噪比（SNR）是磁共振成像（MRI）中评估脉冲序列、接收线圈和图像重建算法的关键性能指标。已经提出了各种方法来估计信噪比。然而，缺乏一致和广泛可用的开源实现一直是临床和研究环境中可靠信噪比比较的挑战。为了解决这一差距，这项工作引入了MR Optimum，这是一个用于标准化信噪比分析的云原生开源平台。MR Optimum将已建立的信噪比估计技术集成在灵活的模块化软件架构中。基于web的用户界面支持数据上传、任务配置、云计算和实时结果可视化。MR Optimum利用无服务器计算技术（AWS Lambda和Fargate）对MRI原始数据执行可扩展的、事件驱动的处理，并允许用户使用既定方法计算信噪比：多个副本、伪多个副本、广义伪多个副本和分析方法。结果包括信噪比图、噪声协方差和噪声系数矩阵、线圈灵敏度曲线和g因子图。web界面支持基于感兴趣区域的交互式可视化和直方图分析。结果可以以MATLAB、NIfTI和JSON格式导出。通过提供统一的计算环境，MR Optimum确保再现性，并使最先进的信噪比估计民主化，促进多中心协调和质量保证。

{"title":"MR Optimum: A web-based open-source tool for standardized signal-to-noise ratio evaluation in MRI","authors":"Eros Montin , Xuan Thao Nguyen , Riccardo Lattanzi","doi":"10.1016/j.cmpbup.2026.100235","DOIUrl":"10.1016/j.cmpbup.2026.100235","url":null,"abstract":"<div><div>Signal-to-noise ratio (SNR) is a key performance metric in magnetic resonance imaging (MRI) to evaluate pulse sequences, receive coils, and image reconstruction algorithms. A variety of methods have been proposed to estimate SNR. However, the lack of consistent and broadly available open-source implementations has been a challenge for reliable SNR comparisons in clinical and research settings. To address this gap, this work introduces MR Optimum, a cloud-native, open-source platform for standardized SNR analysis. MR Optimum integrates established SNR estimation techniques within a flexible, modular software architecture. A web-based user interface supports data upload, task configuration, cloud computations, and real-time results visualization. MR Optimum leverages serverless computing technologies (AWS Lambda and Fargate) to perform scalable, event-driven processing of MRI rawdata and allow users to calculate SNR using established methods: multiple replicas, pseudo multiple replicas, generalized pseudo multiple replicas, and analytic methods. Results include SNR maps, noise covariance and noise coefficient matrices, coil sensitivity profiles, and g factor maps. The web interface enables interactive visualization and histogram analysis based on regions of interest. Results can be exported in MATLAB, NIfTI, and JSON formats. By providing a unified computational environment, MR Optimum ensures reproducibility, and democratizes access to state-of-the-art SNR estimation, promoting multi-center harmonization and quality assurance.</div></div>","PeriodicalId":72670,"journal":{"name":"Computer methods and programs in biomedicine update","volume":"9 ","pages":"Article 100235"},"PeriodicalIF":0.0,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188807","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Towards objective In-Vitro wound healing assessment with segment anything: A large evaluation of interactive and automated pipelines 面向客观的体外切口愈合评估：交互式和自动化管道的大规模评估

Computer methods and programs in biomedicine update

Pub Date : 2026-06-01 Epub Date: 2025-12-16 DOI: 10.1016/j.cmpbup.2025.100224

Katja Löwenstein , Johanna Rehrl , Anja Schuster , Michael Gadermayr

The in vitro scratch assay is a widely used assay in cell biology to assess the rate of wound closure related to a variety of therapeutic interventions. While manual measurement is subjective and vulnerable to intra- and interobserver variability, computer-based tools are theoretically objective, but in practice often contain parameters which are manually adjusted (individually per image or data set) and thereby provide a source for subjectivity. Modern deep learning approaches typically require large annotated training data which complicates instant applicability. In this paper, we deeply investigate the Segment Anything Model (SAM), a deep foundation model based on interactive point-prompts, which enables class-agnostic segmentation without tuning the network’s parameters based on any domain specific training data. With respect to segmentation accuracy, the interactive method significantly outperformed a semi-objective baseline that required manual inspection and, when necessary, parameter adjustments for each image. Experiments were conducted to evaluate the impact of variability due to interactive prompting. The results exhibited remarkably low intra- and interobserver variability, clearly surpassing the consistency of manual segmentation by domain experts. In addition, a fully automated zero-shot approach was explored, incorporating the self-supervised learning model DINOv2 as a preprocessing step before sampling input points for SAM, with various sampling methods systematically investigated.

体外划痕实验是细胞生物学中广泛使用的一种实验，用于评估与各种治疗干预相关的伤口愈合率。虽然人工测量是主观的，容易受到观察者内部和观察者之间变化的影响，但基于计算机的工具在理论上是客观的，但在实践中往往包含手动调整的参数（每个图像或数据集单独），从而提供了主观性的来源。现代深度学习方法通常需要大量带注释的训练数据，这使得即时适用性变得复杂。本文深入研究了基于交互式点提示的深度基础模型SAM (Segment Anything Model)，该模型无需根据任何特定领域的训练数据调整网络参数，即可实现与类别无关的分割。在分割精度方面，交互式方法显著优于半客观基线，后者需要人工检查，并在必要时对每个图像进行参数调整。我们进行了实验来评估交互提示引起的变异性的影响。结果显示观察者内部和观察者之间的可变性非常低，明显超过了领域专家手工分割的一致性。此外，探索了一种全自动零采样方法，将自监督学习模型DINOv2作为SAM输入点采样前的预处理步骤，并对各种采样方法进行了系统研究。

{"title":"Towards objective In-Vitro wound healing assessment with segment anything: A large evaluation of interactive and automated pipelines","authors":"Katja Löwenstein , Johanna Rehrl , Anja Schuster , Michael Gadermayr","doi":"10.1016/j.cmpbup.2025.100224","DOIUrl":"10.1016/j.cmpbup.2025.100224","url":null,"abstract":"<div><div>The <em>in vitro</em> scratch assay is a widely used assay in cell biology to assess the rate of wound closure related to a variety of therapeutic interventions. While manual measurement is subjective and vulnerable to intra- and interobserver variability, computer-based tools are theoretically objective, but in practice often contain parameters which are manually adjusted (individually per image or data set) and thereby provide a source for subjectivity. Modern deep learning approaches typically require large annotated training data which complicates instant applicability. In this paper, we deeply investigate the Segment Anything Model (SAM), a deep foundation model based on interactive point-prompts, which enables class-agnostic segmentation without tuning the network’s parameters based on any domain specific training data. With respect to segmentation accuracy, the interactive method significantly outperformed a semi-objective baseline that required manual inspection and, when necessary, parameter adjustments for each image. Experiments were conducted to evaluate the impact of variability due to interactive prompting. The results exhibited remarkably low intra- and interobserver variability, clearly surpassing the consistency of manual segmentation by domain experts. In addition, a fully automated zero-shot approach was explored, incorporating the self-supervised learning model DINOv2 as a preprocessing step before sampling input points for SAM, with various sampling methods systematically investigated.</div></div>","PeriodicalId":72670,"journal":{"name":"Computer methods and programs in biomedicine update","volume":"9 ","pages":"Article 100224"},"PeriodicalIF":0.0,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145766132","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Investigation into respiratory sound classification for an imbalanced data set using hybrid LSTM-KAN architectures 基于混合LSTM-KAN架构的不平衡数据集呼吸声分类研究

Computer methods and programs in biomedicine update

Pub Date : 2026-06-01 Epub Date: 2025-12-26 DOI: 10.1016/j.cmpbup.2025.100227

Nithinkumar K.V., Anand R.

Respiratory sounds captured via auscultation contain critical clues for diagnosing pulmonary conditions. Automated classification of these sounds faces the dual challenge of distinguishing subtle acoustic patterns and addressing the severe class imbalance inherent in clinical datasets. This study investigates methods for classifying respiratory sounds into multiple disease categories, with a specific focus on mitigating pronounced class imbalances. In this study, we developed and evaluated a hybrid deep learning model incorporating a Long Short-Term Memory (LSTM) network as a feature sequence encoder, followed by a Kolmogorov–Arnold Network (KAN) for classification. This architecture was combined with a comprehensive feature extraction pipeline and targeted imbalance mitigation techniques. The model was evaluated using a public respiratory sound database comprising six classes with a highly skewed distribution. Strategies such as focal loss, class-specific data augmentation, and Synthetic Minority Over-sampling Technique (SMOTE) are employed to improve minority class recognition. Our results demonstrate that the proposed Hybrid LSTM-KAN model achieves a high overall accuracy of 94.6% and a macro-averaged

F_{1}

-score of 0.703. This performance is notable, given that the dominant class (COPD) constitutes over 86% of the data. While challenges persist for the rarest classes (Bronchiolitis and URTI, with

F_{1}

-scores of approximately 0.45 and 0.44, respectively), the approach shows significant improvement in their detection compared to naive baselines and performs strongly on other minority classes, such as bronchiectasis (

F_{1}

-score

\approx

0.84). This study contributes to the development of intelligent auscultation tools for the early detection of respiratory diseases, highlighting the potential of combining recurrent neural networks with advanced KAN architectures and focused imbalance handling.

通过听诊捕获的呼吸声音包含诊断肺部疾病的关键线索。这些声音的自动分类面临着双重挑战，即区分细微的声学模式和解决临床数据集中固有的严重类别不平衡。本研究探讨了将呼吸音分类为多种疾病类别的方法，特别侧重于减轻明显的类别不平衡。在本研究中，我们开发并评估了一种混合深度学习模型，该模型将长短期记忆（LSTM）网络作为特征序列编码器，然后使用Kolmogorov-Arnold网络（KAN）进行分类。该架构结合了全面的特征提取管道和有针对性的失衡缓解技术。该模型使用公共呼吸声数据库进行评估，该数据库包含6个高度倾斜分布的类别。采用焦点丢失、特定类别数据增强和合成少数派过采样技术（SMOTE）等策略来提高少数派类别识别。结果表明，本文提出的混合LSTM-KAN模型总体准确率达到94.6%，宏观平均f1得分为0.703。考虑到占主导地位的类别（COPD）占数据的86%以上，这一表现值得注意。虽然对于最罕见的类别（细支气管炎和尿路感染，f1分数分别约为0.45和0.44）仍然存在挑战，但与初始基线相比，该方法在检测它们方面显示出显着改善，并且在其他少数类别，如支气管扩张（f1分数≈0.84）上表现强劲。这项研究有助于智能听诊工具的发展，用于早期发现呼吸系统疾病，突出了将递归神经网络与先进的KAN架构和集中不平衡处理相结合的潜力。

{"title":"Investigation into respiratory sound classification for an imbalanced data set using hybrid LSTM-KAN architectures","authors":"Nithinkumar K.V., Anand R.","doi":"10.1016/j.cmpbup.2025.100227","DOIUrl":"10.1016/j.cmpbup.2025.100227","url":null,"abstract":"<div><div>Respiratory sounds captured via auscultation contain critical clues for diagnosing pulmonary conditions. Automated classification of these sounds faces the dual challenge of distinguishing subtle acoustic patterns and addressing the severe class imbalance inherent in clinical datasets. This study investigates methods for classifying respiratory sounds into multiple disease categories, with a specific focus on mitigating pronounced class imbalances. In this study, we developed and evaluated a hybrid deep learning model incorporating a Long Short-Term Memory (LSTM) network as a feature sequence encoder, followed by a Kolmogorov–Arnold Network (KAN) for classification. This architecture was combined with a comprehensive feature extraction pipeline and targeted imbalance mitigation techniques. The model was evaluated using a public respiratory sound database comprising six classes with a highly skewed distribution. Strategies such as focal loss, class-specific data augmentation, and Synthetic Minority Over-sampling Technique (SMOTE) are employed to improve minority class recognition. Our results demonstrate that the proposed Hybrid LSTM-KAN model achieves a high overall accuracy of 94.6% and a macro-averaged <span><math><msub><mrow><mi>F</mi></mrow><mrow><mn>1</mn></mrow></msub></math></span>-score of 0.703. This performance is notable, given that the dominant class (COPD) constitutes over 86% of the data. While challenges persist for the rarest classes (Bronchiolitis and URTI, with <span><math><msub><mrow><mi>F</mi></mrow><mrow><mn>1</mn></mrow></msub></math></span>-scores of approximately 0.45 and 0.44, respectively), the approach shows significant improvement in their detection compared to naive baselines and performs strongly on other minority classes, such as bronchiectasis (<span><math><msub><mrow><mi>F</mi></mrow><mrow><mn>1</mn></mrow></msub></math></span>-score <span><math><mo>≈</mo></math></span> 0.84). This study contributes to the development of intelligent auscultation tools for the early detection of respiratory diseases, highlighting the potential of combining recurrent neural networks with advanced KAN architectures and focused imbalance handling.</div></div>","PeriodicalId":72670,"journal":{"name":"Computer methods and programs in biomedicine update","volume":"9 ","pages":"Article 100227"},"PeriodicalIF":0.0,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940013","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A new architecture based on the Xception algorithm for pneumonia detection using medical image datasets 一种基于异常算法的医学图像肺炎检测新架构

Computer methods and programs in biomedicine update

Pub Date : 2026-06-01 Epub Date: 2026-01-29 DOI: 10.1016/j.cmpbup.2026.100234

Chaymae Taib, Otman Abdoun, El Khatir Haimoudi

This study aims to improve the reliability of pneumonia detection from chest X-ray images by addressing the instability and performance variability observed in conventional CNNs, particularly the original Xception architecture, under different training conditions. An improved Xception-based model (IXCEP) is proposed, in which the Entry, Middle, and Exit flows are redesigned using enhanced separable convolutions and skip connections. The model is evaluated on a single pneumonia dataset using both a train/validation split and 5-fold cross-validation, considering different learning rates and numbers of epochs, with additional ablation experiments to assess the contribution of each modified flow.

Experimental results show that IXCEP_full consistently outperforms the original Xception, achieving accuracy ranges of 88.1–98.1% (30 epochs, LR = 0.01), 94.5–97.9% (30 epochs, LR = 0.001), 96.0–99.8% (100 epochs, LR = 0.01), and 87.9–99.0% (100 epochs, LR = 0.001), with markedly reduced variability across folds. Ablation analysis reveals that the optimized Entry and Middle flows yield the most stable performance, reaching accuracies of 98.6–99.3%, whereas the Exit-only configuration shows higher sensitivity to training conditions. In contrast, the original Xception exhibits strong instability, with accuracy ranging from 51.4% to 93.8% across folds.

Additional results, including F1-score values of up to 99.8% and AUC values between 98.8% and 100%, supported by Friedman and Iman–Davenport statistical tests, confirm the statistical significance of the improvements. Grad-CAM visualizations further demonstrate that IXCEP focuses on clinically relevant lung regions. Overall, these findings recommend IXCEP as a more stable and reliable alternative to the original Xception for pneumonia detection from chest radiographs.

本研究旨在通过解决传统cnn（特别是原始异常架构）在不同训练条件下观察到的不稳定性和性能变异性，提高从胸部x线图像检测肺炎的可靠性。提出了一种改进的基于异常的模型（IXCEP），其中使用增强的可分离卷积和跳过连接重新设计入口、中间和出口流。该模型在单个肺炎数据集上进行评估，使用训练/验证分割和5倍交叉验证，考虑不同的学习率和epoch数量，并通过额外的消融实验来评估每个修改流量的贡献。实验结果表明，IXCEP_full的准确率范围分别为88.1-98.1%（30个epoch， LR = 0.01）、94.5-97.9%（30个epoch， LR = 0.001）、96.0-99.8%（100个epoch， LR = 0.01）和87.9-99.0%(100个epoch， LR = 0.001)，显著降低了不同折叠间的变异性。烧蚀分析表明，优化后的入口流和中间流的性能最稳定，达到98.6-99.3%的精度，而只有出口流的配置对训练条件的敏感性更高。相比之下，原始异常具有较强的不稳定性，其准确率在51.4% ~ 93.8%之间。在Friedman和Iman-Davenport统计检验的支持下，f1得分值高达99.8%，AUC值在98.8% - 100%之间，进一步证实了改善的统计学意义。Grad-CAM可视化进一步证明IXCEP侧重于临床相关的肺区域。总的来说，这些发现建议IXCEP作为一种更稳定和可靠的替代方案，用于胸片肺炎检测。

{"title":"A new architecture based on the Xception algorithm for pneumonia detection using medical image datasets","authors":"Chaymae Taib, Otman Abdoun, El Khatir Haimoudi","doi":"10.1016/j.cmpbup.2026.100234","DOIUrl":"10.1016/j.cmpbup.2026.100234","url":null,"abstract":"<div><div>This study aims to improve the reliability of pneumonia detection from chest X-ray images by addressing the instability and performance variability observed in conventional CNNs, particularly the original Xception architecture, under different training conditions. An improved Xception-based model (IXCEP) is proposed, in which the Entry, Middle, and Exit flows are redesigned using enhanced separable convolutions and skip connections. The model is evaluated on a single pneumonia dataset using both a train/validation split and 5-fold cross-validation, considering different learning rates and numbers of epochs, with additional ablation experiments to assess the contribution of each modified flow.</div><div>Experimental results show that IXCEP_full consistently outperforms the original Xception, achieving accuracy ranges of 88.1–98.1% (30 epochs, LR = 0.01), 94.5–97.9% (30 epochs, LR = 0.001), 96.0–99.8% (100 epochs, LR = 0.01), and 87.9–99.0% (100 epochs, LR = 0.001), with markedly reduced variability across folds. Ablation analysis reveals that the optimized Entry and Middle flows yield the most stable performance, reaching accuracies of 98.6–99.3%, whereas the Exit-only configuration shows higher sensitivity to training conditions. In contrast, the original Xception exhibits strong instability, with accuracy ranging from 51.4% to 93.8% across folds.</div><div>Additional results, including F1-score values of up to 99.8% and AUC values between 98.8% and 100%, supported by Friedman and Iman–Davenport statistical tests, confirm the statistical significance of the improvements. Grad-CAM visualizations further demonstrate that IXCEP focuses on clinically relevant lung regions. Overall, these findings recommend IXCEP as a more stable and reliable alternative to the original Xception for pneumonia detection from chest radiographs.</div></div>","PeriodicalId":72670,"journal":{"name":"Computer methods and programs in biomedicine update","volume":"9 ","pages":"Article 100234"},"PeriodicalIF":0.0,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188809","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Machine learning-based analysis of ECG and PCG signals for rheumatic heart disease detection: A scoping review (2015–2025) 基于机器学习的ECG和PCG信号分析用于风湿性心脏病检测：范围综述（2015-2025）

Computer methods and programs in biomedicine update

Pub Date : 2026-06-01 Epub Date: 2025-12-23 DOI: 10.1016/j.cmpbup.2025.100228

Damilare Emmanuel Olatunji , Julius Dona Zannu, Carine Pierrette Mukamakuza, Godbright Nixon Uiso, Chol Buol, John Bosco Thuo, Nchofon Tagha Ghogomu, Mona Mamoun Mubarak Aman, Evelyne Umubyeyi

AI-powered stethoscopes offer a promising alternative for screening rheumatic heart disease (RHD), particularly in regions with limited diagnostic infrastructure. Early detection is vital, yet echocardiography, the gold standard tool, remains largely inaccessible in low-resource settings due to cost and workforce constraints. This review systematically examines machine learning (ML) applications from 2015 to 2025 that analyze electrocardiogram (ECG) and phonocardiogram (PCG) data to support accessible, scalable screening of all RHD variants in relation to the World Heart Federation's "25 by 25" goal to reduce RHD mortality. Using PRISMA-ScR guidelines, 37 peer-reviewed studies were selected from PubMed, IEEE Xplore, Scopus, and Embase. Convolutional neural networks (CNNs) dominate recent efforts, achieving a median accuracy of 97.75 %, F1-score of 0.95, and AUROC of 0.89. However, challenges remain: 73 % of studies used single-center datasets, 81.1 % relied on private data, only 10.8 % were externally validated, and none assessed cost-effectiveness. Although 45.9 % originated from endemic regions, few addressed demographic diversity or implementation feasibility. These gaps underscore the disconnect between model performance and clinical readiness. Bridging this divide requires standardized benchmark datasets, prospective trials in endemic areas, and broader validation. If these issues are addressed, AI-augmented auscultation could transform cardiovascular diagnostics in underserved populations, thereby aiding early detection. This review also offers practical recommendations for building accessible ML-based RHD screening tools, aiming to close the diagnostic gap in low-resource settings where conventional auscultation may miss up to 90 % of cases and echocardiography remains out of reach.

人工智能听诊器为筛查风湿性心脏病（RHD）提供了一种有希望的替代方法，特别是在诊断基础设施有限的地区。早期检测至关重要，但由于成本和劳动力限制，超声心动图作为一种金标准工具，在资源匮乏的环境中仍然难以获得。本综述系统地研究了2015年至2025年机器学习（ML）的应用，这些应用分析了心电图（ECG）和心音图（PCG）数据，以支持与世界心脏联合会（World Heart Federation）降低RHD死亡率的“25 by 25”目标相关的所有RHD变异的可访问、可扩展的筛查。使用PRISMA-ScR指南，从PubMed、IEEE explore、Scopus和Embase中选择了37项同行评议的研究。卷积神经网络（cnn）在最近的研究中占主导地位，实现了97.75%的中位数准确率，f1得分为0.95，AUROC为0.89。然而，挑战仍然存在：73%的研究使用单中心数据集，81.1%依赖于私人数据，只有10.8%的研究经过外部验证，没有评估成本效益。虽然45.9%来自流行地区，但很少涉及人口多样性或实施可行性。这些差距强调了模型性能和临床准备之间的脱节。弥合这一鸿沟需要标准化的基准数据集、流行地区的前瞻性试验和更广泛的验证。如果这些问题得到解决，人工智能增强听诊可以改变服务不足人群的心血管诊断，从而有助于早期发现。本综述还为建立可访问的基于ml的RHD筛查工具提供了实用建议，旨在缩小资源匮乏地区的诊断差距，在这些地区，传统听诊可能错过高达90%的病例，超声心动图仍然遥不可及。

{"title":"Machine learning-based analysis of ECG and PCG signals for rheumatic heart disease detection: A scoping review (2015–2025)","authors":"Damilare Emmanuel Olatunji , Julius Dona Zannu, Carine Pierrette Mukamakuza, Godbright Nixon Uiso, Chol Buol, John Bosco Thuo, Nchofon Tagha Ghogomu, Mona Mamoun Mubarak Aman, Evelyne Umubyeyi","doi":"10.1016/j.cmpbup.2025.100228","DOIUrl":"10.1016/j.cmpbup.2025.100228","url":null,"abstract":"<div><div>AI-powered stethoscopes offer a promising alternative for screening rheumatic heart disease (RHD), particularly in regions with limited diagnostic infrastructure. Early detection is vital, yet echocardiography, the gold standard tool, remains largely inaccessible in low-resource settings due to cost and workforce constraints. This review systematically examines machine learning (ML) applications from 2015 to 2025 that analyze electrocardiogram (ECG) and phonocardiogram (PCG) data to support accessible, scalable screening of all RHD variants in relation to the World Heart Federation's \"25 by 25\" goal to reduce RHD mortality. Using PRISMA-ScR guidelines, 37 peer-reviewed studies were selected from PubMed, IEEE Xplore, Scopus, and Embase. Convolutional neural networks (CNNs) dominate recent efforts, achieving a median accuracy of 97.75 %, F1-score of 0.95, and AUROC of 0.89. However, challenges remain: 73 % of studies used single-center datasets, 81.1 % relied on private data, only 10.8 % were externally validated, and none assessed cost-effectiveness. Although 45.9 % originated from endemic regions, few addressed demographic diversity or implementation feasibility. These gaps underscore the disconnect between model performance and clinical readiness. Bridging this divide requires standardized benchmark datasets, prospective trials in endemic areas, and broader validation. If these issues are addressed, AI-augmented auscultation could transform cardiovascular diagnostics in underserved populations, thereby aiding early detection. This review also offers practical recommendations for building accessible ML-based RHD screening tools, aiming to close the diagnostic gap in low-resource settings where conventional auscultation may miss up to 90 % of cases and echocardiography remains out of reach.</div></div>","PeriodicalId":72670,"journal":{"name":"Computer methods and programs in biomedicine update","volume":"9 ","pages":"Article 100228"},"PeriodicalIF":0.0,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940014","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A fully automated, data-driven approach for dimensionality reduction and clustering in single-cell RNA-seq analysis 在单细胞RNA-seq分析中用于降维和聚类的全自动，数据驱动的方法

Computer methods and programs in biomedicine update

Pub Date : 2026-06-01 Epub Date: 2026-01-26 DOI: 10.1016/j.cmpbup.2026.100232

Hyun Kim , Faeyza Rishad Ardi , Kévin Spinicci , Jae Kyoung Kim

Single-cell RNA sequencing (scRNA-seq) provides deep insights into cellular heterogeneity but demands robust dimensionality reduction (DR) and clustering to handle high-dimensional, noisy data. Many DR and clustering approaches rely on user-defined parameters, undermining reliability. Even automated clustering methods like ChooseR and MultiK still employ fixed principal component defaults, limiting their full automation. To overcome this limitation, we propose a fully automated clustering approach by integrating scLENS—a method for optimal PC selection—with these tools. Our fully automated approach improves clustering performance by ∼14 % for ChooseR and ∼10 % for MultiK and identifies additional cell subtypes, highlighting the advantages of adaptive, data-driven DR.

单细胞RNA测序（scRNA-seq）提供了对细胞异质性的深入了解，但需要强大的降维（DR）和聚类来处理高维，嘈杂的数据。许多容灾和聚类方法依赖于用户定义的参数，从而降低了可靠性。即使是像ChooseR和MultiK这样的自动化集群方法仍然使用固定的主成分默认值，限制了它们的完全自动化。为了克服这一限制，我们提出了一种完全自动化的聚类方法，通过将sclen -一种最佳PC选择方法-与这些工具集成在一起。我们的全自动方法将ChooseR和MultiK的聚类性能分别提高了14%和10%，并识别了额外的细胞亚型，突出了自适应数据驱动DR的优势。

引用次数: 0