Computational and systems oncology最新文献

{"title":"Cross-Cancer Profiling of CDH1 Reveals Context-Dependent EMT Decoupling, Immune Heterogeneity, and Prognostic Variability in Epithelial Cancers","authors":"Md. Abdur Rahman,&nbsp;Sm Faysal Bellah,&nbsp;Md. Mustafizur Rahman","doi":"10.1002/cso2.70023","DOIUrl":"https://doi.org/10.1002/cso2.70023","url":null,"abstract":"<p>Cadherin 1 (<i>CDH1</i>) encodes E-cadherin, a key epithelial adhesion molecule traditionally associated with tumor suppression and epithelial–mesenchymal transition (EMT). However, its roles across cancers remain incompletely understood, particularly within multilayer regulatory contexts involving genomic, epigenetic, transcriptional, and immune mechanisms. <i>CDH1</i> expression, survival associations, EMT-correlated gene profiles (<i>VIM</i>, <i>SNAI1</i>, and <i>ZEB1</i>), immune infiltration patterns, immune checkpoint correlations (<i>PDCD1</i>, <i>CD274</i>, and <i>CTLA4</i>), promoter methylation, and genomic alterations were assessed across five epithelial cancers, breast invasive carcinoma (BRCA), colon adenocarcinoma (COAD), lung adenocarcinoma (LUAD), ovarian serous cystadenocarcinoma (OV), and stomach adenocarcinoma (STAD). Cross-platform analysis was performed using The Cancer Genome Atlas (TCGA)/Genomic Data Commons (GDC) datasets, Gene Expression Profiling Interactive Analysis 2 (GEPIA2), UALCAN, TIMER, KM Plotter, cBioPortal, and g:Profiler. <i>CDH1</i> was overexpressed but showed variable prognostic significance; TCGA/GDC analyses were nonsignificant, whereas KM Plotter showed better survival in COAD, LUAD, and STAD, worse survival in BRCA, and no association in OV. Classic inverse relationships between <i>CDH1</i> and <i>VIM</i> or <i>ZEB1</i> were evident only in STAD, and <i>SNAI1</i> showed no consistent association. Immune infiltration patterns were tumor-specific, ranging from cytotoxic T-cell dominance in LUAD to macrophage-rich profiles in OV; immune checkpoint correlations were similarly context-dependent. Cancer-specific enrichment revealed distinct trafficking- and adhesion-related profiles. Promoter methylation patterns varied by cancer, whereas genomic alterations of <i>CDH1</i> were rare. <i>CDH1</i> does not function as a universal epithelial or EMT marker across epithelial cancers. Instead, its associations with EMT, immune contexture, methylation, and prognosis are context-dependent, supporting a model of <i>CDH1</i> as a heterogeneous regulator of epithelial plasticity. These findings challenge single-function interpretations and support cancer-specific <i>CDH1</i> evaluation in translational research.</p>","PeriodicalId":72658,"journal":{"name":"Computational and systems oncology","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cso2.70023","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148752947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrated Analysis of lncRNA–miRNA–mRNA ceRNA Network in Order to Discover Potential Novel Biomarkers in Stomach Cancer","authors":"Habib MotieGhader,&nbsp;Seyed Mehdi Jazayeri,&nbsp;Reyhaneh Sadat Jazayeri","doi":"10.1002/cso2.70021","DOIUrl":"https://doi.org/10.1002/cso2.70021","url":null,"abstract":"<p>Stomach cancer (SC) or gastric cancer (GC) is one of the most common gastrointestinal malignancies. Currently, some studies based on competing endogenous RNAs (ceRNA) network analysis have been assessed for this cancer. The construction and analysis of ceRNA network is a novel approach for identification of RNA-based biomarkers in cancer studies. In the current study, our aim is to discover novel potential biomarker based on lncRNA, miRNA, and mRNA ceRNA network analysis. To do this, differentially expressed miRNAs (DEmiRNAs) between SC and normal samples have been collected from the DBDEMC website. Then, a list of lncRNAs regulating DEmiRNA is collected from the LncACTdb 3.0 database and the lncRNA–miRNA bipartite network has been constructed. Subsequently, target mRNAs of the DEmiRNAs have been collected from the miRTarBase website and a protein–protein interaction (PPI) network was constructed thanks to the STRING website. Afterwards, five gene modules were discovered from the PPI network. Thus, miRNA–mRNA bipartite network has been constructed for DEmiRNAs and modules' genes. Finally, lncRNA–miRNA–mRNA ceRNA network is constructed and analyzed. After analyzing the ceRNA network, it is observed that <i>DANCR</i> and <i>MEG3</i> with degree three and <i>XIST</i>, <i>MALAT1</i>, <i>TTN-AS1</i>, <i>HOTAIR</i>, <i>HCAL</i>, <i>GAS5</i> with degree two, have the important role in the ceRNA network regulation. As well as, <i>miR-145-5p, miR-145-3p, miR-183-5p, miR-21-5p</i> are hub miRNAs in the ceRNA network and they are targeting more genes of the PPI modules' genes. Among the genes evaluated, miR-96-5p and miR-183-5p were the only candidates that exhibited significant differential expression between the normal and SC groups and showed a significant association with overall survival (OS) in patients with stomach cancer (log-rank <i>p</i> &lt; 0.05), underscoring their potential prognostic significance.</p>","PeriodicalId":72658,"journal":{"name":"Computational and systems oncology","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cso2.70021","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148616354","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"HierAnom-Path: Zero-Shot Rare Cancer Detection in Histopathology via Hierarchical Foundation Model Adaptation","authors":"Noora Saleem Jumaah,&nbsp;Mohanad Sameer Jabbar,&nbsp;Saif H. Alwaeli,&nbsp;Noor Ayesha,&nbsp;Mohamed Babiker A. Mohamed,&nbsp;Muhammad I. Khan","doi":"10.1002/cso2.70022","DOIUrl":"https://doi.org/10.1002/cso2.70022","url":null,"abstract":"<p>Rare cancers collectively represent 20%–25% of all malignancies yet yield fewer than 50 annotated whole-slide images (WSIs) per subtype in publicly available archives, rendering supervised deep learning inapplicable. We present HierAnom-Path, to the best of our knowledge, the first framework for zero-shot rare cancer anomaly detection in gigapixel histopathological WSIs that operates without any labeled rare cancer examples at any stage of training. HierAnom-Path integrates four novel contributions: (1) a hierarchical two-level BiomedCLIP adapter with cross-cancer contrastive LoRA (Rank 16, 4.7 million trainable parameters) that learns morphological anomaly representations invariant across cancer subtypes without labeled rare cancer data; (2) RarePath-Diffusion, a text-conditioned latent diffusion model synthesizing photorealistic rare cancer pathology tiles from structured clinical prompts, providing training signal without any real labeled rare cancer slides; (3) the cross-cancer-subtype zero-shot (CCSZ) benchmark protocol, the first evaluation framework enforcing strictly disjoint training and test cancer subtypes; and (4) a clinician-aligned spatial anomaly heatmap with text-grounded morphological explanations and a structured, auditable inference log. Evaluated on 30 rare brain tumor subtypes from the EBRAINS Digital Brain Tumour Atlas, HierAnom-Path achieves a mean slide-level AUROC of 0.841 ± 0.024, a +17.0 percentage point improvement over the strongest CLIP-based baseline (AnomalyCLIP: 0.671 ± 0.031; DeLong test, <i>p</i> &lt; 0.001, Bonferroni-corrected), with tile-level AUROC 0.783 ± 0.031 and AUPRO 0.762 ± 0.028. Multicenter validation on CAMELYON17 confirms scanner-robust generalization (Slide-AUROC 0.812 ± 0.019 across five independent sites; inter-site range: 3.8 percentage points). Grade correlation on the PANDA prostate dataset (<i>n</i> = 10,616 WSIs) yields Spearman <i>ρ</i> = 0.741 (<i>p</i> &lt; 0.001), and cross-dataset evaluation on five non-brain rare TCGA subtypes achieves a mean Slide-AUROC of 0.798 ± 0.029. Blinded pathologist validation yields Fleiss' <i>κ</i> = 0.67 (substantial agreement) with explanation-localization Dice 0.64 ± 0.11, confirming clinical interpretability. HierAnom-Path closes the majority of the gap to a fully supervised upper bound (0.910 ± 0.018) while requiring zero rare cancer labels, establishing a new state of the art for zero-shot rare cancer anomaly detection.</p>","PeriodicalId":72658,"journal":{"name":"Computational and systems oncology","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cso2.70022","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148616313","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"HierAnom-Path: Zero-Shot Rare Cancer Detection in Histopathology via Hierarchical Foundation Model Adaptation","authors":"Noora Saleem Jumaah,&nbsp;Mohanad Sameer Jabbar,&nbsp;Saif H. Alwaeli,&nbsp;Noor Ayesha,&nbsp;Mohamed Babiker A. Mohamed,&nbsp;Muhammad I. Khan","doi":"10.1002/cso2.70022","DOIUrl":"https://doi.org/10.1002/cso2.70022","url":null,"abstract":"<p>Rare cancers collectively represent 20%–25% of all malignancies yet yield fewer than 50 annotated whole-slide images (WSIs) per subtype in publicly available archives, rendering supervised deep learning inapplicable. We present HierAnom-Path, to the best of our knowledge, the first framework for zero-shot rare cancer anomaly detection in gigapixel histopathological WSIs that operates without any labeled rare cancer examples at any stage of training. HierAnom-Path integrates four novel contributions: (1) a hierarchical two-level BiomedCLIP adapter with cross-cancer contrastive LoRA (Rank 16, 4.7 million trainable parameters) that learns morphological anomaly representations invariant across cancer subtypes without labeled rare cancer data; (2) RarePath-Diffusion, a text-conditioned latent diffusion model synthesizing photorealistic rare cancer pathology tiles from structured clinical prompts, providing training signal without any real labeled rare cancer slides; (3) the cross-cancer-subtype zero-shot (CCSZ) benchmark protocol, the first evaluation framework enforcing strictly disjoint training and test cancer subtypes; and (4) a clinician-aligned spatial anomaly heatmap with text-grounded morphological explanations and a structured, auditable inference log. Evaluated on 30 rare brain tumor subtypes from the EBRAINS Digital Brain Tumour Atlas, HierAnom-Path achieves a mean slide-level AUROC of 0.841 ± 0.024, a +17.0 percentage point improvement over the strongest CLIP-based baseline (AnomalyCLIP: 0.671 ± 0.031; DeLong test, <i>p</i> &lt; 0.001, Bonferroni-corrected), with tile-level AUROC 0.783 ± 0.031 and AUPRO 0.762 ± 0.028. Multicenter validation on CAMELYON17 confirms scanner-robust generalization (Slide-AUROC 0.812 ± 0.019 across five independent sites; inter-site range: 3.8 percentage points). Grade correlation on the PANDA prostate dataset (<i>n</i> = 10,616 WSIs) yields Spearman <i>ρ</i> = 0.741 (<i>p</i> &lt; 0.001), and cross-dataset evaluation on five non-brain rare TCGA subtypes achieves a mean Slide-AUROC of 0.798 ± 0.029. Blinded pathologist validation yields Fleiss' <i>κ</i> = 0.67 (substantial agreement) with explanation-localization Dice 0.64 ± 0.11, confirming clinical interpretability. HierAnom-Path closes the majority of the gap to a fully supervised upper bound (0.910 ± 0.018) while requiring zero rare cancer labels, establishing a new state of the art for zero-shot rare cancer anomaly detection.</p>","PeriodicalId":72658,"journal":{"name":"Computational and systems oncology","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cso2.70022","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148616317","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrated Analysis of lncRNA–miRNA–mRNA ceRNA Network in Order to Discover Potential Novel Biomarkers in Stomach Cancer","authors":"Habib MotieGhader,&nbsp;Seyed Mehdi Jazayeri,&nbsp;Reyhaneh Sadat Jazayeri","doi":"10.1002/cso2.70021","DOIUrl":"https://doi.org/10.1002/cso2.70021","url":null,"abstract":"<p>Stomach cancer (SC) or gastric cancer (GC) is one of the most common gastrointestinal malignancies. Currently, some studies based on competing endogenous RNAs (ceRNA) network analysis have been assessed for this cancer. The construction and analysis of ceRNA network is a novel approach for identification of RNA-based biomarkers in cancer studies. In the current study, our aim is to discover novel potential biomarker based on lncRNA, miRNA, and mRNA ceRNA network analysis. To do this, differentially expressed miRNAs (DEmiRNAs) between SC and normal samples have been collected from the DBDEMC website. Then, a list of lncRNAs regulating DEmiRNA is collected from the LncACTdb 3.0 database and the lncRNA–miRNA bipartite network has been constructed. Subsequently, target mRNAs of the DEmiRNAs have been collected from the miRTarBase website and a protein–protein interaction (PPI) network was constructed thanks to the STRING website. Afterwards, five gene modules were discovered from the PPI network. Thus, miRNA–mRNA bipartite network has been constructed for DEmiRNAs and modules' genes. Finally, lncRNA–miRNA–mRNA ceRNA network is constructed and analyzed. After analyzing the ceRNA network, it is observed that <i>DANCR</i> and <i>MEG3</i> with degree three and <i>XIST</i>, <i>MALAT1</i>, <i>TTN-AS1</i>, <i>HOTAIR</i>, <i>HCAL</i>, <i>GAS5</i> with degree two, have the important role in the ceRNA network regulation. As well as, <i>miR-145-5p, miR-145-3p, miR-183-5p, miR-21-5p</i> are hub miRNAs in the ceRNA network and they are targeting more genes of the PPI modules' genes. Among the genes evaluated, miR-96-5p and miR-183-5p were the only candidates that exhibited significant differential expression between the normal and SC groups and showed a significant association with overall survival (OS) in patients with stomach cancer (log-rank <i>p</i> &lt; 0.05), underscoring their potential prognostic significance.</p>","PeriodicalId":72658,"journal":{"name":"Computational and systems oncology","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cso2.70021","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148616291","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Big Data in Cancer Genomics: Computational Foundations and Emerging Pathways for Precision Oncology","authors":"Nur Vanu,&nbsp;Nur Mohammad,&nbsp;Fahad Ahmed,&nbsp;Towsif Alam,&nbsp;Koushik Saha,&nbsp;Durga Shahi,&nbsp;Adib Hossain,&nbsp;Mia Md. Tofayel Gonee Manik,&nbsp;Mst. Rina Parvin","doi":"10.1002/cso2.70020","DOIUrl":"https://doi.org/10.1002/cso2.70020","url":null,"abstract":"<p>This review aims to explore the computational foundations of big data in cancer genomics and examine emerging pathways that support precision oncology and personalized cancer care. A narrative review approach was adopted to synthesize evidence from PubMed, Scopus, Web of Science, and IEEE Xplore. The literature search was conducted between January 10 and February 25, 2026, and 68 relevant studies were included in the final synthesis. Relevant studies were selected on the basis of their focus on computational methods, data integration strategies, and artificial intelligence (AI) applications in cancer genomics. Extracted data were organized into thematic categories and analyzed using an iterative synthesis framework. The findings indicate that high-throughput sequencing and multi-omics technologies have significantly expanded the volume and complexity of cancer-related data. Advanced infrastructures, including cloud platforms, improve storage and access but raise privacy and interoperability concerns. Machine learning and AI support tumor classification, biomarker discovery, and treatment prediction. Integrative multi-omics enhances biological insight and predictive accuracy. However, challenges such as data heterogeneity, limited model generalizability, and gaps in clinical integration remain. Big data in cancer genomics offer substantial potential to advance precision oncology by enabling more accurate and personalized treatment strategies. However, overcoming technical, ethical, and infrastructural barriers is essential to ensure effective translation into clinical practice and equitable healthcare outcomes.</p>","PeriodicalId":72658,"journal":{"name":"Computational and systems oncology","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cso2.70020","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148466675","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Big Data in Cancer Genomics: Computational Foundations and Emerging Pathways for Precision Oncology","authors":"Nur Vanu,&nbsp;Nur Mohammad,&nbsp;Fahad Ahmed,&nbsp;Towsif Alam,&nbsp;Koushik Saha,&nbsp;Durga Shahi,&nbsp;Adib Hossain,&nbsp;Mia Md. Tofayel Gonee Manik,&nbsp;Mst. Rina Parvin","doi":"10.1002/cso2.70020","DOIUrl":"https://doi.org/10.1002/cso2.70020","url":null,"abstract":"<p>This review aims to explore the computational foundations of big data in cancer genomics and examine emerging pathways that support precision oncology and personalized cancer care. A narrative review approach was adopted to synthesize evidence from PubMed, Scopus, Web of Science, and IEEE Xplore. The literature search was conducted between January 10 and February 25, 2026, and 68 relevant studies were included in the final synthesis. Relevant studies were selected on the basis of their focus on computational methods, data integration strategies, and artificial intelligence (AI) applications in cancer genomics. Extracted data were organized into thematic categories and analyzed using an iterative synthesis framework. The findings indicate that high-throughput sequencing and multi-omics technologies have significantly expanded the volume and complexity of cancer-related data. Advanced infrastructures, including cloud platforms, improve storage and access but raise privacy and interoperability concerns. Machine learning and AI support tumor classification, biomarker discovery, and treatment prediction. Integrative multi-omics enhances biological insight and predictive accuracy. However, challenges such as data heterogeneity, limited model generalizability, and gaps in clinical integration remain. Big data in cancer genomics offer substantial potential to advance precision oncology by enabling more accurate and personalized treatment strategies. However, overcoming technical, ethical, and infrastructural barriers is essential to ensure effective translation into clinical practice and equitable healthcare outcomes.</p>","PeriodicalId":72658,"journal":{"name":"Computational and systems oncology","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cso2.70020","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148466676","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The Evolution of Breast Cancer Detection: A Review of Imaging, Machine Learning, and Multimodal Strategies","authors":"Likhon Chandra Sarkar,&nbsp;Farida Siddiqi Prity,&nbsp;Farhad Hussain,&nbsp;Anup Talukder,&nbsp;Mirza Raquib,&nbsp;Sharmina Rahman,&nbsp;MD Jiabul Hoque","doi":"10.1002/cso2.70018","DOIUrl":"https://doi.org/10.1002/cso2.70018","url":null,"abstract":"<p>Breast cancer is still a serious problem in the world arena, where its early and prompt detection is the most important factor in improving patient prognosis and survival. The use of traditional diagnostic techniques, such as imaging (e.g., mammography and ultrasound) and subsequent histopathological examination, is the mainstay, which, however, is hindered by a variety of limitations and is less sensitive in certain instances (e.g., dense breast tissue). Computer‑aided diagnosis (CAD) systems based on machine learning (ML) and deep learning (DL) approaches have become revolutionary technologies to tackle these issues. The review is a comparative analysis of all the various methodologies involved in detecting breast cancer. It first analyzes the conventional imaging methods and their shortcomings. It then reviews the use of classical ML algorithms (e.g., support vector machines, random forests) mostly on structured clinical data. The main idea of the review is the popularity of DL to evaluate a number of medical images, such as histopathology slides, ultrasound, and mammograms, in particular, convolutional neural networks (CNNs). Finally, the review examines the next generation of hybrid approaches, such as ensemble learning, data fusion architectures, and combinations of multimodal data, such as genomics, pathomics, and metabolomics. This paper assesses the performance, strengths, and limitations of these coexisting strategies by synthesizing recent findings, noting that there are still some persistent challenges in terms of data scarcity, model interpretability, and clinical translation. The review has ended with a summary of the present situation and the future of breast cancer diagnostics.</p>","PeriodicalId":72658,"journal":{"name":"Computational and systems oncology","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cso2.70018","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148324146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The Evolution of Breast Cancer Detection: A Review of Imaging, Machine Learning, and Multimodal Strategies","authors":"Likhon Chandra Sarkar,&nbsp;Farida Siddiqi Prity,&nbsp;Farhad Hussain,&nbsp;Anup Talukder,&nbsp;Mirza Raquib,&nbsp;Sharmina Rahman,&nbsp;MD Jiabul Hoque","doi":"10.1002/cso2.70018","DOIUrl":"https://doi.org/10.1002/cso2.70018","url":null,"abstract":"<p>Breast cancer is still a serious problem in the world arena, where its early and prompt detection is the most important factor in improving patient prognosis and survival. The use of traditional diagnostic techniques, such as imaging (e.g., mammography and ultrasound) and subsequent histopathological examination, is the mainstay, which, however, is hindered by a variety of limitations and is less sensitive in certain instances (e.g., dense breast tissue). Computer‑aided diagnosis (CAD) systems based on machine learning (ML) and deep learning (DL) approaches have become revolutionary technologies to tackle these issues. The review is a comparative analysis of all the various methodologies involved in detecting breast cancer. It first analyzes the conventional imaging methods and their shortcomings. It then reviews the use of classical ML algorithms (e.g., support vector machines, random forests) mostly on structured clinical data. The main idea of the review is the popularity of DL to evaluate a number of medical images, such as histopathology slides, ultrasound, and mammograms, in particular, convolutional neural networks (CNNs). Finally, the review examines the next generation of hybrid approaches, such as ensemble learning, data fusion architectures, and combinations of multimodal data, such as genomics, pathomics, and metabolomics. This paper assesses the performance, strengths, and limitations of these coexisting strategies by synthesizing recent findings, noting that there are still some persistent challenges in terms of data scarcity, model interpretability, and clinical translation. The review has ended with a summary of the present situation and the future of breast cancer diagnostics.</p>","PeriodicalId":72658,"journal":{"name":"Computational and systems oncology","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cso2.70018","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148324193","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"National Trends and Disparities in Cancer Mortality Among Individuals With Alzheimer's Disease and Related Dementias in the United States, 1999–2020","authors":"Aasim Ali,&nbsp;Javed Iqbal,&nbsp;Aimen Anwaar Pannun,&nbsp;Muhammad Talha,&nbsp;Fiza Nisar,&nbsp;Ufaq Reyaz,&nbsp;Muhammad Saad Ansari,&nbsp;Sameen Sarfaraz,&nbsp;Ahmad Butt,&nbsp;Hammad Azam,&nbsp;Muhammad Shamoon,&nbsp;Labeeba Abdul Ghafoor,&nbsp;Armaghana Abdullah,&nbsp;Muhammad Asad Shabbir,&nbsp;Mukesh Kumar Sharma","doi":"10.1002/cso2.70019","DOIUrl":"https://doi.org/10.1002/cso2.70019","url":null,"abstract":"<p>As population aging accelerates, the coexistence of cancer and Alzheimer's disease and related dementias (ADRD) is becoming increasingly common. While each condition independently contributes to mortality, national trends in cancer-related deaths occurring alongside ADRD—and their demographic and geographic disparities—remain poorly defined. We conducted a retrospective population-based analysis using mortality data from the CDC WONDER (1999–2020). Deaths listing both malignant neoplasms (ICD-10: C00–C97) and ADRD (F01, F03, G30) were identified. Age-adjusted mortality rates (AAMRs) per 100,000 population were calculated using the 2000 U.S. standard population. Temporal trends were assessed using joinpoint regression to estimate annual percent change (APC) with 95% confidence intervals. Stratified analyses by sex, race, region, and state were performed to evaluate disparities. A total of 89,514 deaths with concurrent cancer and ADRD were identified. The AAMR more than doubled from 2.0 per 100,000 in 1999 to 4.6 in 2020, peaking in 2012. Overall mortality increased significantly from 1999 to 2009 (APC 9.11%, <i>p</i> &lt; 0.05), followed by stabilization. Sex-specific analyses showed continued increases among females, whereas males demonstrated a significant decline after 2010. By race, Black individuals experienced a steeper early rise followed by a significant decline, while trends among White individuals plateaued after 2009. Regionally, the South exhibited a significant post-2010 decline, whereas other regions stabilized. Marked state-level variation was observed, with the highest rates in the Midwest and the lowest in the Southwest. Cancer mortality with concurrent ADRD increased substantially in the early 2000s, followed by divergent trends across demographic and geographic subgroups. Persistent disparities highlight the need for integrated oncologic and cognitive care strategies and targeted public health interventions addressing inequities in aging populations.</p>","PeriodicalId":72658,"journal":{"name":"Computational and systems oncology","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cso2.70019","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148282280","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}