Computers in biology and medicine最新文献

{"title":"Interpretable deep learning architecture for gastrointestinal disease detection: A Tri-stage approach with PCA and XAI.","authors":"Md Faysal Ahamed, Fariya Bintay Shafi, Md Nahiduzzaman, Mohamed Arselene Ayari, Amith Khandakar","doi":"10.1016/j.compbiomed.2024.109503","DOIUrl":"https://doi.org/10.1016/j.compbiomed.2024.109503","url":null,"abstract":"<p><p>GI abnormalities significantly increase mortality rates and impose considerable strain on healthcare systems, underscoring the essential requirement for rapid detection, precise diagnosis, and efficient strategic treatment. To develop a CAD system, this study aims to automatically classify GI disorders utilizing various deep learning methodologies. The proposed system features a three-stage lightweight architecture, consisting of a feature extractor using PSE-CNN, a feature selector employing PCA, and a classifier based on DELM. The framework, designed with only 24 layers and 1.25 million parameters, is employed on the largest dataset, GastroVision, containing 8000 images of 27 GI disorders. To improve visual clarity, a sequential preprocessing strategy is implemented. The model's robustness is evaluated through 5-fold cross-validation. Additionally, several XAI methods, namely Grad-CAM, heatmaps, saliency maps, SHAP, and activation feature maps, are used to explore the model's interpretability. Statistical significance is ensured by calculating the p-value, demonstrating the framework's reliability. The proposed model PSE-CNN-PCA-DELM has achieved outstanding results in the first stage, categorizing the diseases' positions into three primary classes, with average accuracy (97.24 %), precision (97.33 ± 0.01 %), recall (97.24 ± 0.01 %), F1-score (97.33 ± 0.01 %), ROC-AUC (99.38 %), and AUC-PR (98.94 %). In the second stage, the dataset is further divided into nine separate classes, considering the overall disease characteristics, and achieves excellent outcomes with average performance rates of 90.00 %, 89.71 ± 0.11 %, 89.59 ± 0.14 %, 89.51 ± 0.12 %, 98.49 %, and 94.63 %, respectively. The third stage involves a more detailed classification into twenty-seven classes, maintaining strong performance with scores of 93.00 %, 82.69 ± 0.37 %, 83.00 ± 0.38 %, 81.54 ± 0.35 %, 97.38 %, and 88.03 %, respectively. The framework's compact size of 14.88 megabytes and average testing time of 59.17 milliseconds make it highly efficient. Its effectiveness is further validated through comparisons with several TL approaches. Practically, the framework is extremely resilient for clinical implementation.</p>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"185 ","pages":"109503"},"PeriodicalIF":7.0,"publicationDate":"2024-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142794387","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Novel Lobe-based Transformer model (LobTe) to predict emphysema progression in Alpha-1 Antitrypsin Deficiency.","authors":"Ariel Hernán Curiale, Raúl San José Estépar","doi":"10.1016/j.compbiomed.2024.109500","DOIUrl":"https://doi.org/10.1016/j.compbiomed.2024.109500","url":null,"abstract":"<p><p>Emphysema, marked by irreversible lung tissue destruction, poses challenges in progression prediction due to its heterogeneity. Early detection is particularly critical for patients with Alpha-1 Antitrypsin Deficiency (AATD), a genetic disorder reducing ATT protein levels. Heterozygous carriers (PiMS and PiMZ) have variable AAT levels thus complicating their prognosis. This study introduces a novel prognostic model, the Lobe-based Transformer encoder (LobTe), designed to predict the annual change in lung density (ΔALD [g/L-yr]) using CT scans. Utilizing a global self-attention mechanism, LobTe specifically analyzes lobar tissue destruction to forecast disease progression. In parallel, we developed and compared a second model utilizing an LSTM architecture that implements a local subject-specific attention mechanism. Our methodology was validated on a cohort of 2,019 participants from the COPDGene study. The LobTe model demonstrated a small root mean squared error (RMSE=1.73 g/L-yr) and a notable correlation coefficient (ρ=0.61), explaining over 35% of the variability in ΔALD (R²= 0.36). Notably, it achieved a higher correlation coefficient of 0.68 for PiMZ heterozygous carriers, indicating its effectiveness in detecting early emphysema progression among smokers with mild to moderate AAT deficiency. The presented models could serve as a tool for monitoring disease progression and informing treatment strategies in carriers and subjects with AATD. Our code is available at github.com/acil-bwh/LobTe.</p>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"185 ","pages":"109500"},"PeriodicalIF":7.0,"publicationDate":"2024-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142791196","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ToxSTK: A multi-target toxicity assessment utilizing molecular structure and stacking ensemble learning.","authors":"Surapong Boonsom, Panisara Chamnansil, Sarote Boonseng, Tarapong Srisongkram","doi":"10.1016/j.compbiomed.2024.109480","DOIUrl":"https://doi.org/10.1016/j.compbiomed.2024.109480","url":null,"abstract":"<p><p>Drug registration requires risk assessment of new active pharmaceutical ingredients or excipients to ensure they are safe for human health and the environment. However, traditional risk assessment is expensive and relies heavily on animal testing. Machine learning (ML) has been used as a risk assessment tool, providing less time, money, and involved animals than in vivo experiments. Despite that, the ML models often rely on a single model, which may introduce bias and unreliable prediction. Stacking ensemble learning is an ML framework that makes predictions based on multimodal outcomes. This framework performs well in quantitative structure-activity relationship (QSAR) studies. In this study, we developed ToxSTK, a multi-target toxicity assessment using stacking ensemble learning. We aimed to create an ML tool that facilitates toxicity assessments more affordably with reduced reliance on animal models. We focused on four key targets generally assessed in early-stage drug development: hERG toxicity, mTOR toxicity, PBMCs toxicity, and mutagenicity. Our model integrated 12 molecular fingerprints with 3 ML algorithms, generating 36 novel predictive features (PFs). These PFs were then combined to construct the final meta-decision model. Our results demonstrated that the ToxSTK model surpasses standard regression and classification metrics, ensuring it is highly reliable and accurate in predicting chemical toxicities within its application domain. This model passed the y-randomization test, confirming that the identified QSAR is robust and not due to random chance. Additionally, this model outperforms the existing ML methods for these endpoints, suggesting its effectiveness for risk assessment applications. We recommend incorporating this stacking ensemble learning framework into the chemical risk assessment pipeline to improve model generalization, accuracy, robustness, and reliability.</p>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"185 ","pages":"109480"},"PeriodicalIF":7.0,"publicationDate":"2024-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142791259","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A computational approach to matching multiple sclerosis-related, IGH CDR3s with a MBP epitope.","authors":"Justin M Cole, Jacob T Treanor, Cassondra M Lyman, Diep Nguyen, Andrea Chobrutskiy, Boris I Chobrutskiy, George Blanck","doi":"10.1016/j.compbiomed.2024.109482","DOIUrl":"https://doi.org/10.1016/j.compbiomed.2024.109482","url":null,"abstract":"<p><p>In multiple sclerosis (MS), T-cell receptors (TCRs) and antibodies specifically target the main structural proteins of myelin, including myelin basic protein (MBP), especially a specific, canonical, immunoglobulin (IG)-targeted MBP epitope. Efficient computational analyses to diagnose or monitor autoimmune conditions, which could have broad applicability in clinical trials or in diagnoses, remains a challenge. As such, we considered the possibility that focusing on the immunoglobin heavy chain (IGH) complementarity determining region-3 (CDR3) amino acid sequences could support the development of an efficient, convenient, and user-friendly approach to detecting or assessing IGH targets in MS. Thus, we applied a chemical complementarity scoring algorithm, extensively benchmarked in many cancer settings, to assess the combined electrostatic and hydrophobic attractiveness of large numbers of (individual patient) IGH CDR3s and the canonical IG MBP epitope. Samples and controls were filtered to only include CDR3s above a baseline chemical complementarity score. Then, the frequency of each unique IGH CDR3 (with the minimum MBP epitope complementarity) in the MS samples was compared to the same parameter for the control sample. Specifically, a greater number of high frequency IGH CDR3s, with chemically complementary to the canonical MBP epitope, was detected in 47 out of 48 MS-control comparisons, in most cases representing a p < 0.0001. With continued development, this approach has the potential to lead to a user-friendly computational screening tool for patients at risk for developing MS. Additional results indicate that the methodology could also be applied to antigen epitope discovery.</p>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"185 ","pages":"109482"},"PeriodicalIF":7.0,"publicationDate":"2024-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142791185","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reliable peak detection and feature extraction for wireless electrocardiograms.","authors":"Sajad Farrokhi, Waltenegus Dargie, Christian Poellabauer","doi":"10.1016/j.compbiomed.2024.109478","DOIUrl":"https://doi.org/10.1016/j.compbiomed.2024.109478","url":null,"abstract":"<p><p>The electrocardiogram (ECG) is a vital device to examine the electrical activities of the heart. It is useful for diagnosing cardiovascular diseases, which often manifest themselves through alterations in the ECG signals' characteristics. These alterations are primarily observed in the signals' key components: the Q, R, S, T, and P peaks. At present, cardiologists typically rely on manual inspection of ECG measurements taken in controlled environments, such as hospitals and clinics, but most cardiac conditions reveal themselves outside clinical settings, when patients freely move and exert. In this paper, we dynamically identify and extract prominent ECG features in measurements taken outside clinical settings by subjects who have no medical training. The activities we consider are typical activities cardiac patients carry out in residential and rehabilitation environments, such as sitting, climbing up and down stairs, and standing up. To achieve accurate feature extraction, we employ adaptive thresholding and localization techniques. Our approach achieves promising results, with an average% for R peak detection and 92% for Q and S peaks detection. Similarly, our approach enables the detection of T and P peaks with an average accuracy of 87% and 84%, respectively.</p>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"185 ","pages":"109478"},"PeriodicalIF":7.0,"publicationDate":"2024-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142791237","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Unraveling new avenues in pancreatic cancer treatment: A comprehensive exploration of drug repurposing using transcriptomic data.","authors":"Naina Sunildutt, Faheem Ahmed, Abdul Rahim Chethikkattuveli Salih, Hyung Chul Kim, Kyung Hyun Choi","doi":"10.1016/j.compbiomed.2024.109481","DOIUrl":"https://doi.org/10.1016/j.compbiomed.2024.109481","url":null,"abstract":"<p><p>Pancreatic cancer, a malignancy notorious for its late-stage diagnosis and low patient survival rates, remains a formidable global health challenge. The currently available FDA-approved treatments for pancreatic cancer, notably chemotherapeutic agents, exhibit suboptimal efficacy, often accompanied by concerns regarding toxicity. Given the intricate nature of pancreatic cancer pathogenesis and the time-intensive nature of in silico drug discovery approaches, drug repurposing emerges as a compelling strategy to expedite the development of novel therapeutic interventions. In our study, we harnessed transcriptomic data from an exhaustive exploration of four diverse databases, ensuring a rigorous and unbiased analysis of differentially expressed genes, with a particular focus on upregulated genes associated with pancreatic cancer. Leveraging these pancreatic cancer-associated host protein targets, we employed a battery of cutting-edge bioinformatics tools, including Cytoscape STRING, GeneMANIA, Connectivity Map, and NetworkAnalyst, to identify potential small molecule drug candidates and elucidate their interactions. Subsequently, we conducted meticulous docking and redocking simulations for the selected drug-protein target pairs. This rigorous computational approach culminated in the identification of two promising broad-spectrum drug candidates against four pivotal host genes implicated in pancreatic cancer. Our findings strongly advocate for further investigation and preclinical validation of these candidates. Specifically, we propose prioritizing Dasatinib for evaluation against MMP3, MMP9, and EGFR due to their remarkable binding affinities, as well as Pioglitazone against MMP3, MMP2 and MMP9. These discoveries hold great promise in advancing the therapeutic landscape for pancreatic cancer, offering new avenues for improving patient outcomes.</p>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"185 ","pages":"109481"},"PeriodicalIF":7.0,"publicationDate":"2024-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142791263","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MDMNI-DGD: A novel graph neural network approach for druggable gene discovery based on the integration of multi-omics data and the multi-view network.","authors":"Jianwei Li, Bing Li, Xukun Zhang, Xuxu Ma, Ziyu Li","doi":"10.1016/j.compbiomed.2024.109511","DOIUrl":"https://doi.org/10.1016/j.compbiomed.2024.109511","url":null,"abstract":"<p><p>Accurately predicting druggable genes is of paramount importance for enhancing the efficacy of targeted therapies, reducing drug-related toxicities and improving patients' survival rates. Nevertheless, accurately predicting candidate cancer-druggable genes remains a critical challenge in translational medicine due to the high heterogeneity and complexity of cancer data. In this study, we proposed a novel graph neural approach called Druggable Gene Discovery based on the Integration of Multi-omics Data and the Multi-view Network (MDMNI-DGD), aiming to predict and evaluate cancer-druggable genes. MDMNI-DGD integrated a comprehensive set of multi-omics data, including copy number variations, DNA methylation, somatic mutations, and gene expression profiles. Simultaneously, it constructed the multi-view gene association network based on protein-protein interactions (PPI), protein structural domains, gene co-expression, pathway co-occurrence, gene sequence and gene ontology. Compared to other state-of-the-art approaches, MDMNI-DGD exhibits excellent performance in key evaluation metrics such as AUROC and AUPR. Moreover, the case study has also demonstrated the efficacy of our approach in discovering potentially druggable genes. Among more than 20,000 protein-coding genes, MDMNI-DGD successfully identified 872 potentially druggable genes. The findings from this investigation may serve to bolster the assessment of pan-cancer druggable genes, potentially catalyzing the development of more personalized and efficacious therapeutic interventions.</p>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"185 ","pages":"109511"},"PeriodicalIF":7.0,"publicationDate":"2024-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142791194","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Protein-protein interaction detection using deep learning: A survey, comparative analysis, and experimental evaluation.","authors":"Kamal Taha","doi":"10.1016/j.compbiomed.2024.109449","DOIUrl":"https://doi.org/10.1016/j.compbiomed.2024.109449","url":null,"abstract":"<p><p>This survey paper provides a comprehensive analysis of various Deep Learning (DL) techniques and algorithms for detecting protein-protein interactions (PPIs). It examines the scalability, interpretability, accuracy, and efficiency of each technique, offering a detailed empirical and experimental evaluation. Empirically, the techniques are assessed based on four key criteria, while experimentally, they are ranked by specific algorithms and broader methodological categories. Deep Neural Networks (DNNs) demonstrated high accuracy but faced limitations such as overfitting and low interpretability. Convolutional Neural Networks (CNNs) were highly efficient at extracting hierarchical features from biological sequences, while Generative Stochastic Networks (GSNs) excelled in handling uncertainty. Long Short-Term Memory (LSTM) networks effectively captured temporal dependencies within PPI sequences, though they presented scalability challenges. This paper concludes with insights into potential improvements and future directions for advancing DL techniques in PPI identification, highlighting areas where further optimization can enhance performance and applicability.</p>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"185 ","pages":"109449"},"PeriodicalIF":7.0,"publicationDate":"2024-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142791219","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Securing healthcare medical image information using advance morphological component analysis, information hiding systems, and hybrid convolutional neural networks on IoMT.","authors":"Binay Kumar Pandey, Digvijay Pandey","doi":"10.1016/j.compbiomed.2024.109499","DOIUrl":"https://doi.org/10.1016/j.compbiomed.2024.109499","url":null,"abstract":"<p><p>Health care images contain a variety of imaging information that has specific features, which can make it challenging to assess and decide on the methods necessitated to safeguard the highly classified visuals from unauthorized exposure during transmission in a communication channel. As a result, this proposed approach utilizes a variety of techniques that will enhance the quality of textual healthcare images, communicate information securely, and interpret textual data from healthcare visuals without difficulty. Natural interference, primarily on the receiver side, reduces text-based healthcare image contrast, and numerous artifacts and adjacent picture element values impede diagnosis. Therefore, at the transmission end, the suggested method uses morphological component analysis to improve the contrast of textual healthcare images. Subsequently, it masks this textual healthcare image behind the cover image using steganography, maintaining the secrecy of private information during transmission on the Internet of Medical Things (IoMT) network. After obtaining the stego-image, reverse steganography is used to separate the textual health care image from the cover image. Following that, pre-processing had been performed utilizing weighted guided image filters to ensure that a text-based health care image would not be altered when data had been sent through an IoMT. After that, the Gabor Transform (GT) and stroke width transform are then used to extract the features required for a weighted classification approach that distinguishes between healthcare images with and without text content. Employing the cultural emperor penguin optimization strategy strengthened the performance of the weighted naive Bayes classifier. Later, a hybrid convolutional neural network with enhanced cuckoo search optimization is utilized to detect textual information in healthcare images. A variety of indicators are utilized to evaluate each cover picture and text-based healthcare image. These are accuracy, precision, recall, sensitivity, specificity, structural similarity index, peak signal-to-noise ratio, number of bytes of embedded and recovered input health-care textual pictures, and mean square error. The findings show that the proposed strategy outperforms all of the existing approaches. The suggested method successfully retrieves content at the receiver end. However, a few characters may be misplaced or recovered many times due to weighted guided image filtration halo artifacts, which impair image quality and provide inaccurate textual data.</p>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"185 ","pages":"109499"},"PeriodicalIF":7.0,"publicationDate":"2024-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142791252","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Challenges and solutions of deep learning-based automated liver segmentation: A systematic review.","authors":"Vahideh Ghobadi, Luthffi Idzhar Ismail, Wan Zuha Wan Hasan, Haron Ahmad, Hafiz Rashidi Ramli, Nor Mohd Haziq Norsahperi, Anas Tharek, Fazah Akhtar Hanapiah","doi":"10.1016/j.compbiomed.2024.109459","DOIUrl":"https://doi.org/10.1016/j.compbiomed.2024.109459","url":null,"abstract":"<p><p>The liver is one of the vital organs in the body. Precise liver segmentation in medical images is essential for liver disease treatment. The deep learning-based liver segmentation process faces several challenges. This research aims to analyze the challenges of liver segmentation in prior studies and identify the modifications made to network models and other enhancements implemented by researchers to tackle each challenge. In total, 88 articles from Scopus and ScienceDirect databases published between January 2016 and January 2022 have been studied. The liver segmentation challenges are classified into five main categories, each containing some subcategories. For each challenge, the proposed technique to overcome the challenge is investigated. The provided report details the authors, publication years, dataset types, imaging technologies, and evaluation metrics of all references for comparison. Additionally, a summary table outlines the challenges and solutions.</p>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"185 ","pages":"109459"},"PeriodicalIF":7.0,"publicationDate":"2024-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142791189","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}