Search Results (2,266)

Lung cancer remains the leading cause of cancer-related mortality worldwide, largely due to late-stage diagnosis, biological heterogeneity, and persistent challenges in staging and treatment selection. This narrative review summarizes current and emerging applications of AI across lung cancer screening and early detection, imaging-based staging and prognostication, tissue and liquid biopsy-based tumor characterization, treatment planning, surgical and intraoperative guidance, and drug discovery. In imaging, deep learning models have demonstrated high performance in pulmonary nodule detection, risk stratification, and prediction of molecular alterations, while also showing promise in improving screening efficiency and reducing interpretive variability. In pathology and liquid biopsy domains, AI enables prediction of driver mutations, immunotherapy response, and survival outcomes directly from histopathology slides, circulating tumor DNA, and other blood-based biomarkers, facilitating minimally invasive precision oncology approaches. In treatment planning and delivery, AI systems are being developed to support clinical decision-making, surgical planning (through advanced image segmentation and delineation of operative anatomy), and intraoperative navigation through robotic and computer vision-enabled platforms. Despite these advances, significant barriers remain, including limited real-world validation, algorithmic biases, workflow integration issues, and unresolved ethical and legal concerns. Future progress will depend on the development of transparent, clinically validated, and generalizable AI systems that augment rather than replace the expertise of clinical providers and healthcare teams. Active engagement from pulmonologists, oncologists, radiologists, and thoracic surgeons will be essential in guiding safe implementation and ensuring that AI-driven innovations translate into meaningful improvements in patient outcomes. Full article

Background and Clinical Significance: Childhood-onset Sjögren’s disease (cSjD) is a rare autoimmune disorder that remains challenging to diagnose because of its heterogeneous clinical presentation and the frequent absence of classic sicca symptoms at disease onset. Recurrent parotitis and systemic manifestations often predominate in pediatric patients, contributing to diagnostic delay and potential irreversible glandular damage. Early recognition is essential to prevent complications and improve long-term outcomes. Case Presentation: We report the case of a 17-year-old female diagnosed with primary Sjögren’s disease following a prolonged history of recurrent parotid involvement and progressive glandular dysfunction. Comprehensive evaluation revealed positive anti-SSA antibodies, hypergammaglobulinemia, characteristic salivary gland ultrasonography abnormalities, and a positive minor salivary gland biopsy, resulting in fulfillment of all domains of the 2016 ACR/EULAR classification criteria. The patient also exhibited unusual vascular findings, including carotid atheromatous calcifications in the absence of traditional cardiovascular risk factors. Conclusion: This case highlights the diagnostic complexity of cSjD and underscores the value of a multimodal diagnostic approach integrating clinical assessment, serology, imaging, and histopathology. The presence of early vascular abnormalities broadens the spectrum of potential extraglandular manifestations and emphasizes the need for comprehensive evaluation and long-term monitoring in affected patients. Full article

Ovarian cancer is a gynecological cancer, which, if metastasized and not detected early, can cause death among women. Therefore, accurate prediction of drug responses to ovarian cancer is needed. A gynecological pathologist inspects abnormality in tissues and provides a report for patients; however, this diagnostic process (1) is difficult to undertake; (2) requires experience; and (3) is time-consuming. Moreover, existing tools are imperfect. Hence, we present a computational pipeline to improve predictions of drug response pertaining to ovarian cancer. First, we downloaded digital pathology images pertaining to ovarian responses to bevacizumab from the Cancer Imaging Archive Repository. We employed a histogram of oriented gradients for images, constructed feature vectors, and used Fisher’s linear discriminant analysis to alter data representations through dimensionality reduction. This reduced-dimensionality data was used for regression analysis, employing support vector regression coupled with various kernels and calculating the area under the ROC curve (AUC). Experimental results were validated using transformer-based models (ViT and Swin) and other deep learning (DL) models (VGG16, ResNet50, InceptionV3, MobileNetV2, and EfficientNetB6). Our approach using a radial kernel (named SVRD + R) improved AUC performance by 17% compared to the best-performing transformer-based model (ViT). Likewise, AUC performance improved by 14.9% when compared against the best DL-based model (MobileNetV2). These results demonstrate feasibility, showing that induced models via the presented AI-based pipeline can lead to superior performance when investigating prediction problems pertaining to gynecologic cancer studies. Full article

The early and accurate detection of lung cancer (LC) is one of the primary challenges in the clinical diagnostics process, which plays a vital role in the treatment of the disease. Although various deep learning (DL) techniques have been presented, the existing DL methods are mainly focused on single-modal images, either computed tomography (CT) or histopathological images, which are associated with poor generalization, diversity, and applicability. To mitigate the existing issues, the present work aims to develop a modality-independent ensemble DL framework that is independently evaluated on CT and histopathological image datasets for LC classification. In this work, the proposed framework was developed using the Beta Normalization Aggregation (BNA) technique, where the performance of three state-of-the-art pre-trained convolutional neural network (CNN) architectures was compared on two distinct imaging modalities images. Based on the comparative analysis of the performance metrics, Xception, DenseNet121, and MobileNetV2, are chosen to develop the Ensemble model. Predictions generated by the selected CNN models are aggregated using the proposed BNA strategy to improve classification robustness, which improves the confidence of the prediction results and discriminative capabilities. The experiments using public data sets have confirmed the excellent performance of the model. On the CT dataset, the proposed BNA Ensemble achieved a testing accuracy of 97.45%, with a precision of 97.88%, recall of 97.45%, F1-score of 97.45%, and an AUC of 0.9986. On the histopathological dataset, the framework achieved an accuracy of 99.80%, with precision, recall, and F1-score all reaching 99.80%, and an AUC of 1.0000. These results demonstrate the effectiveness, robustness, and generalizability of the proposed BNA framework. The analysis of the results using t-SNE plots, confusion matrices, ROC curves, and confidence distributions provided additional insights into feature separability, classification performance, and prediction confidence of the proposed framework. Full article

Background/Objectives: Differentiating recurrent disease from postsurgical changes on ¹⁸F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) remains challenging in malignant pleural mesothelioma (MPM). This study aimed to characterize the temporal patterns of postsurgical FDG uptake and evaluate the diagnostic performance of FDG PET/CT for detecting recurrent disease after radical surgery. Methods: We retrospectively analyzed 91 postsurgical PET/CT scans from 45 patients with MPM who underwent extrapleural pneumonectomy (EPP; n = 29) or pleurectomy/decortication (P/D; n = 16). Scans were stratified into four postoperative time intervals: <6 months, 6 to <12 months, 12 to <24 months, and ≥24 months. FDG uptake in the postsurgical bed and local recurrent lesions was quantified using maximum standardized uptake value ratios normalized to the mediastinal blood pool and liver. Recurrence was confirmed by histopathology or follow-up imaging. Results: Postsurgical FDG uptake showed a time-dependent decline, with significantly lower uptake beyond 24 months postoperatively (p < 0.05). EPP patients demonstrated significantly higher postsurgical FDG uptake than P/D patients (p < 0.01). FDG PET/CT identified occult recurrence in 23.4% of CT-negative scans. Local recurrent lesions showed significantly higher FDG uptake than postsurgical changes across all postoperative intervals (p < 0.001). Conclusions: Postsurgical FDG uptake in MPM demonstrates a time-dependent decline, and surgical extent is an important determinant of background metabolic activity. Despite this variable background, FDG PET/CT demonstrated high diagnostic accuracy for detecting recurrent disease, including CT-occult recurrences. Incorporating surgical type and postoperative interval into PET/CT interpretation may improve diagnostic accuracy in postoperative MPM surveillance. Full article

Introduction: Currently, chronic kidney disease (CKD) is detected based on glomerular filtration rate (GFR), proteinuria levels or kidney biopsy. However, the development of MRI techniques and AI algorithms gives hope to the assessment of CKD activity and kidney function with profound MRI image analysis. Methods: MRI images from healthy volunteers with no history of CKD were compared with those from CKD patients who had undergone both kidney MRI and kidney biopsy; the latter group was also divided into two subgroups based on CKD histopathological activity. Patients from both groups were scanned using either a 1.5 T or 3 T MRI scanner following sequential allocation (nine healthy controls and 28 CKD patients and 11 healthy volunteers and 43 CKD patients respectively for each scanner). Results: The final algorithm based on T1-weighted, T2-weighted and DWI images was able to distinguish patients with sensitivity ranging 77.78–87.50%, specificity 86.67–94.12% and precision 77.78–87.50%. Features of T1-weighted images and of T2-weighted images were found to correlate strongly with GFR with coefficients ranging from −0.5922 to −0.7090 and from 0.6126 to 0.6380, respectively. Conclusions: MRI image texture analysis may be suitable for assessing CKD activity, irrespective of the type of MRI scanner used. Furthermore, MRI image texture features correlate with eGFR values. Full article

Injuries of the metacarpophalangeal joint are a major cause of morbidity and catastrophic fracture in racing horses, yet early osseous pathology is often difficult to detect using conventional imaging. This pilot study aimed to correlate sodium fluoride Positron Emission Tomography (¹⁸F-NaF PET) radiopharmaceutical uptake with gross and histopathologic changes in the distal third metacarpal bone (MC3) and proximal first phalanx (P1). Five horses (three racing Thoroughbreds with fetlock injury and two non-racing controls) underwent ante-mortem ¹⁸F-NaF PET and cone-beam CT imaging (CBCT), followed by post-mortem gross and histologic examination of predefined anatomic sites. Quantitative PET measures, including maximum standardized uptake value (SUV_max), SUV_ratio, and PET grade, were compared with gross pathology and histopathologic scores for cartilage and subchondral bone. While there were significant regional correlations between PET metrics and gross or histologic scores at select sites, our results need to be considered in light of the small number of horses evaluated. Correlations between PET metrics and gross pathology score were identified on the distal metacarpus on the lateral dorsal condyle and on proximal P1 for lateral dorsal and mid-P1. Correlation of PET metrics and hyaline cartilage histopathology scores were found for dorsal medial and lateral P1, parasagittal dorsolateral P1 and the medial parasagittal groove of MC3. Correlation of PET metrics and histologic subchondral bone scores were significant for medial palmar condyle, medial parasagittal groove, and parasagittal palmar lateral of MC3. For P1, PET metrics and histologic subchondral bone scores were significantly correlated for parasagittal mid-lateral and medial dorsal regions. Overall, these findings demonstrate that ¹⁸F-NaF PET identifies localized bone remodeling that corresponds to histologic and gross pathology at specific fetlock regions, supporting its utility for detecting osseous injury, although relationships varied by anatomic location. Further work with larger numbers of horses is needed. Full article

Whole-slide imaging has transformed histopathology into a data-rich domain; however, many computational pathology models encode tissue morphology within latent representations, limiting interpretability, reproducibility, and generalization. This review positions histomics as an intermediate phenotype representation layer linking histological images with downstream clinical inference through structured descriptors of tissue morphology, spatial organization, and tissue architecture. Unlike prior reviews focused primarily on feature extraction or predictive performance, the study adopts a representation-centric perspective of histomics. A taxonomy of histomic features across biological scales is presented, and artificial intelligence frameworks, including machine learning, deep learning, weakly supervised learning, and multimodal approaches, are systematically examined. Key challenges, including segmentation dependence, feature instability, aggregation variability, and domain shift, are critically analyzed alongside emerging developments in foundation models, representation learning, and multimodal pathology. The review provides a unified framework for understanding histomic representations and identifies future directions for developing robust, interpretable, and generalizable computational pathology systems. Full article

Breast cancer is a highly prevalent malignancy among women globally and arises from the uncontrolled proliferation of abnormal cells in breast tissue. Timely and precise diagnosis is critical for effective treatment and enhanced survival. Histopathological image analysis is considered the gold standard; nevertheless, manual assessment is labor-intensive and prone to variability. Existing deep learning and transformer-based approaches demonstrate strong effectiveness; however, they incur significant computational complexity and limited efficiency in capturing multi-scale features. To address these challenges, this research presents a framework that integrates ConvMixer, multi-scale patch learning, and an Adaptive Combined Runge–Kutta–Harris Hawks Optimization (AC-RUNHHO) algorithm. The model effectively captures both fine-grained cellular patterns and global tissue structures, while adaptive optimization improves convergence and hyperparameter tuning. The framework is evaluated on a breast cancer histology dataset comprising 4000 histopathological images across four classes. Experimental results demonstrate robust performance under the evaluated experimental conditions, achieving 98.63% accuracy, 98.63% precision, 98.62% recall, and 98.62% F1-score. Ablation and cross-validation analyses further confirm the generalization capability of the model. Overall, the developed framework demonstrates promising performance in computer-aided breast histopathological image classification, achieving high predictive accuracy and providing interpretable visual explanations. Full article

Background/Objectives: Breast cancer remains one of the leading causes of cancer-related mortality among women worldwide. Early and accurate characterization of suspicious mammographic microcalcifications is essential for improving diagnostic decision-making and reducing unnecessary invasive procedures. Microcalcifications classified as BI-RADS 4 and BI-RADS 5 are clinically important radiological findings; however, differentiating benign from malignant lesions remains challenging because of overlapping morphological and distribution patterns. This study aimed to develop a structured feature-based machine learning model for predicting the pathological diagnosis of breast microcalcifications by integrating mammographic descriptors, patient age, and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) contrast enhancement findings. Methods: The dataset included 53 biopsy-confirmed cases and consisted of clinical and radiological variables, including patient age, calcification morphology, calcification size, distribution pattern, DCE-MRI contrast enhancement status, and histopathological outcome. Several conventional machine learning algorithms were evaluated, including Logistic Regression, Support Vector Machine with radial basis function kernel, K-Nearest Neighbors, Decision Tree, Random Forest, Extra Trees, Gradient Boosting, AdaBoost, and CatBoost. Hyperparameter optimization was performed using grid search with five-fold cross-validation. Model performance was assessed using accuracy, precision, recall, F1-score, ROC-AUC, and log loss. Results: Logistic Regression achieved the highest overall performance, with an accuracy of 0.909 and an F1-score of 0.889, while AdaBoost achieved a recall of 1.000 in the internal evaluation. However, given the limited sample size and lack of external validation, these findings should be interpreted as preliminary. Conclusions: The results suggest that structured radiological descriptors combined with DCE-MRI enhancement information may support malignancy risk stratification of BI-RADS 4–5 microcalcifications, although larger multicenter studies are required before clinical implementation. Full article

Prostate cancer (PCa) is the most prevalent cause of cancer-related deaths worldwide. This review aims to synthesize the contemporary literature (2018–2026) on various diagnostic approaches of PCa for clinicians. Currently, the standard approach to PCa diagnosis includes (1) prostate-specific antigen (PSA) testing, (2) its derivatives, and (3) digital rectal examination (DRE). This approach is readily available and cost-effective but lacks specificity, thus resulting in a high incidence of overdiagnosis and overtreatment of PCa patients. This review aims to compare the traditional approaches with the newly developed approaches to PCa diagnosis and investigate their suitability for research purposes, including urine-based markers, liquid biopsy, and histopathology. Each of these approaches has advantages and disadvantages. Advanced imaging techniques like multiparametric magnetic resonance imaging (mpMRI) are helpful in the localization of PCa and can prevent unnecessary biopsies. Histopathology remains the gold standard in the confirmation of PCa diagnosis. Newly developed blood-based markers and advanced imaging techniques are more sensitive and specific in the diagnosis of PCa and are likely to result in a better clinical outcome than the traditional approaches. However, the cost of these approaches remains a global challenge in developing and developed countries alike. PCa diagnosis tools and strategies are presented in a clinical and cost-effective manner and are integrated in a tiered approach with the aim of controlling and managing PCa and its diagnosis. Full article

Background/Objectives: Artificial intelligence (AI) systems for detecting lymph node metastases in histopathological images achieve near-expert classification performance but remain opaque to clinicians, limiting their clinical adoption and regulatory acceptance. This study presents the first rigorous quantitative framework for evaluating and comparing explainable AI (XAI) methods in digital pathology, providing actionable evidence-based guidance for clinical deployment. Methods: Four XAI techniques—LIME, GradCAM, GradCAM++, and SHapley Additive exPlanations (SHAP) via DeepExplainer—were applied to three convolutional neural networks (VGG19, ResNet50, and EfficientNetB3) trained on the PatchCamelyon (PCam) benchmark (220,026 patches). Quantitative evaluation employed two complementary frameworks: spatial agreement with expert pathologist annotations (Intersection over Union and Sørensen–Dice coefficient on 2847 annotated patches) and faithfulness metrics (Area Over the Perturbation Curve and insertion/deletion Area Under the Curve) independent of external annotations. Threshold sensitivity analysis was also conducted at fixed binarisation thresholds (τ = 0.3 and τ = 0.7) in addition to Otsu automatic thresholding. Results: GradCAM++ achieved the highest spatial agreement with pathologist annotations (mean IoU = 0.52 ± 0.14 for EfficientNetB3), while SHAP yielded the highest faithfulness scores (AOPC = 0.61 ± 0.08). The parameter-free squaregrid LIME variant offered a favourable trade-off (IoU = 0.44 ± 0.17) at 3.8× lower computational cost than LIME AVG. Relative method rankings were preserved across all binarisation thresholds, confirming the robustness of the evaluation framework. A Spearman correlation of ρ = 0.81 was found between model classification AUC and spatial agreement, indicating that superior classification performance systematically produces more spatially coherent explanations. Conclusions: GradCAM++ is recommended for high-throughput clinical workflows; SHAP for research contexts requiring maximal faithfulness; and squaregrid LIME as a transparent, parameter-free baseline for clinical communication and audit, preferred over LIME AVG on account of its parameter-free operation and 3.8× lower computational cost. A tiered deployment strategy integrating GradCAM++, SHAP, and squaregrid LIME is proposed. These findings provide quantitative, technical evidence of a type relevant to regulatory frameworks such as the FDA SaMD Action Plan and EU IVDR 2017/746; formal regulatory acceptance would additionally require prospective, multi-site external validation and a pathologist reader study, which lie beyond the scope of this single-benchmark study. Full article

Background: [⁶⁸Ga]Ga-FAPI PET/CT demonstrates high sensitivity for tumor detection but limited specificity because benign fibro-inflammatory conditions may also show tracer uptake. Dual-time-point imaging has been proposed to improve lesion characterization by assessing temporal changes in uptake. This study evaluated whether delayed imaging provides incremental diagnostic value over standard early imaging. Methods: This retrospective lesion-based study evaluated dual-time-point [⁶⁸Ga]Ga-FAPI PET/CT imaging. SUVmax, maximal tumor-to-background ratio (TBRmax), metabolic tumor volume (MTV), and total lesion uptake (TLU) were measured on early and delayed scans. Diagnostic performance for differentiating malignant from benign uptake findings was assessed using receiver operating characteristic analysis. Results: A total of 123 patients underwent dual-time-point imaging at approximately 26 and 65 min post-injection. Overall, 620 [⁶⁸Ga]Ga-FAPI uptake findings were analyzed, including 307 malignant uptake findings and 313 benign findings. SUVmax decreased significantly over time in both malignant and benign uptake findings, with a greater decline in benign findings (%ΔSUVmax −7.7% vs. −3.6%, p = 0.0045). TBRmax increased modestly in malignant uptake findings, while MTV and TLU remained largely stable. SUVmax and TBRmax were significantly higher in malignant uptake at both imaging time points; however, diagnostic performance remained moderate (AUCs of 0.65–0.69) due to substantial overlap between lesion types. Delayed imaging did not improve diagnostic accuracy compared with early imaging, and delta parameters demonstrated poor performance (AUC ~0.5). Similar findings were observed across tumor subgroups. Conclusions: In this lesion-based retrospective cohort, delayed dual-time-point [⁶⁸Ga]Ga-FAPI PET/CT did not demonstrate meaningful incremental diagnostic value over early imaging. A standard early acquisition appears adequate for routine practice, although lesion interpretation remains dependent on clinical, morphologic, and, when needed, histopathologic correlation. Full article

Background: Heterogeneous data integration remains a major challenge in intelligent information systems, particularly under missing-modality and cross-domain conditions. Existing multimodal fusion approaches often rely on complete datasets and weak alignment mechanisms, limiting their robustness and practical applicability. Objectives: This study aims to develop and evaluate a genomics-guided multimodal representation learning framework that enables robust heterogeneous data fusion, reliable cross-modal correspondence, and accurate prediction under incomplete-data conditions. Methods: We propose a multimodal learning architecture that models genomics as the primary biological anchor and learns conditional projections to imaging modalities, including multiparametric MRI and whole-slide histopathology (WSI). The framework formulates multimodal fusion as a genomics-guided contrastive learning problem, incorporates domain-specific optimization constraints, and learns a latent shared-state representation to support inference without requiring fully paired datasets. Evaluation was conducted using public datasets, including TCGA-PRAD and TCIA, across low-risk versus higher-risk/clinically significant prostate cancer (csPCa) discrimination, Gleason-based risk stratification, and clinically significant outcome prediction tasks under realistic multimodal and missing-modality scenarios. Results: In the adequately powered $\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{o}\mathrm{m}\mathrm{i}\mathrm{c}\mathrm{s}+\mathrm{W}\mathrm{S}\mathrm{I}$ cohort (n = 486), the framework achieved an AUROC of 0.985 ± 0.005 for low-risk versus higher-risk/csPCa discrimination (p < 0.001). Exploratory analysis in a small, matched $\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{o}\mathrm{m}\mathrm{i}\mathrm{c}\mathrm{s}+\mathrm{M}\mathrm{R}\mathrm{I}$ cohort (n = 28) yielded an AUROC of 0.980 ± 0.006 for the same endpoint; these findings are reported descriptively with bootstrap confidence intervals due to limited sample size. Because the negative reference group consisted of low-risk prostate cancer cases rather than cancer-free controls, results are interpreted as within-cancer risk discrimination rather than de novo cancer detection. The framework achieved weighted accuracy up to 92.1%, Cohen’s κ up to 0.86, and reduced critical decision errors by 58%. Calibration remained strong (ECE 0.021–0.024), and decision-curve analysis indicated improved utility with reduced unnecessary invasive workups in retrospective modeling. Robustness analysis demonstrated AUROC degradation below 0.04 under domain shifts. Single-modality inference using genomics alone maintained AUROC > 0.90. Interpretability analysis revealed feature attributions aligned with domain-relevant genomic markers. Conclusions: The proposed framework provides a scalable and generalizable solution for heterogeneous multimodal data fusion, supporting reliable prediction, robustness to missing modalities, and applicability to complex information systems beyond the studied domain. Full article

Background/Objectives: Despite hematoxylin and eosin (H&E) staining remaining the cornerstone of histopathological diagnosis, substantial intra- and inter-laboratory variability persists. This issue is increasingly relevant in Digital Pathology, where staining inconsistency may affect whole-slide image interpretation and the performance of image analysis algorithms. In the present work, we evaluated the diagnostic adequacy and non-inferiority of a novel tabs-based H&E histochemical staining method compared with conventional liquid reagents. Methods: Fifty formalin-fixed paraffin-embedded tissue samples from routine practice were sectioned in duplicate and stained either conventionally or using H&E Stain Tabs. After slide review, 14 representative tissue samples were selected, scanned at 40× magnification, and used to generate 24 matched image pairs at different magnifications. A blind online survey was completed by 13 expert pathologists using high-quality monitors. Participants assessed overall staining preference and rated stromal, epithelial, cytoplasmic, and nuclear staining quality. Non-inferiority was tested using a predefined margin of −0.10, and paired rating differences were analyzed using the Wilcoxon signed-rank test. Results: Across 312 paired evaluations, the tabs-based method was preferred in 120 cases (38.5%), conventional staining in 118 cases (37.8%), and no preference was expressed in 74 cases (23.7%). The tabs-based method met the criterion for non-inferiority compared with standard staining (z = 2.7). Rating-scale analysis showed significantly better stromal evaluation with the tablet-based method (z = 2.638; p = 0.008), whereas no significant differences were observed for epithelial, cytoplasmic, or nuclear staining. All evaluated images were considered diagnostically adequate. Conclusions: The tabs-based H&E stain was non-inferior to the conventional method and showed particularly favorable performance in the assessment of stromal components. These findings support its potential role in improving staining reproducibility and standardization, particularly in Digital Pathology workflows where pre-analytical and analytical consistency is critical. Full article