AI has the capacity to strengthen diagnostic pathology by improving consistency, measurement and efficiency and moving from experimental use to routine reporting will require broad validation across centres, enhanced model transparency, strong quality‐assurance systems and close cooperation between developers and pathologists.
Abstract
Background Whole‐slide imaging (WSI) has shifted pathology toward digital workflows, creating the foundation for applying artificial intelligence (AI) to diagnostic tasks. This review summarises validated AI applications in diagnostic pathology, with an emphasis on clinical performance, regulatory developments and the practical barriers that affect implementation. Methods A structured search of PubMed, Scopus and Google Scholar identified English‐language, peer‐reviewed studies from January 2020 to May 2025. Eligible studies applied AI to diagnostic, grading or prognostic tasks in human tissue, used a pathologist‐confirmed reference standard and included external or multi‐centre validation. Results More than 150,000 digital slides were represented across the included studies. Reported performance metrics demonstrated strong diagnostic accuracy across several validated applications. Large meta‐analyses and externally validated studies reported sensitivity values exceeding 96% and specificity above 93% for selected cancer‐detection tasks, while other studies demonstrated high agreement for Gleason grading (QWK up to 0.862) and biomarker quantification (Ki‐67 ICC 0.98). Conclusion AI has the capacity to strengthen diagnostic pathology by improving consistency, measurement and efficiency. Moving from experimental use to routine reporting will require broad validation across centres, enhanced model transparency, strong quality‐assurance systems and close cooperation between developers and pathologists.
Background: Artificial intelligence (AI) has emerged as a transformative technology across dermatological practice, from automated lesion classification to whole-slide pathology analysis. Despite rapid growth in primary studies, a comprehensive synthesis of diagnostic performance, application breadth, and real-world implementation remains lacking. Methods: We conducted a PRISMA systematic review and meta-analysis of studies published from January 2000 to March 2025. We searched the PubMed, Cochrane, and ScienceDirect databases for studies reporting AI diagnostic performance in dermatology. Results: Of 30 included studies (28 valid after exclusion of two retracted publications), 60% focused on melanoma and related lesions. AI diagnostic performance improved markedly over five identified temporal eras (2003–2025), with a pooled AUROC of 0.92 (95% CI 0.87–0.96), Reitsma sensitivity of 0.88 (0.82–0.93), and Reitsma specificity of 0.85 (0.75–0.91). AI matched or surpassed specialist dermatologists in 71% of direct comparisons. Three randomized controlled trials (RCTs) were identified, with heterogeneous findings across different clinical applications: AI assistance significantly improved non-expert diagnostic accuracy in one trial (53.9% vs. 43.8%; p = 0.019), significantly reduced acne severity via personalized treatment recommendations in a second, and showed non-inferior diagnostic performance, but was not cost-effective in the third. The sole cost-effectiveness analysis found AI-assisted surveillance not cost-effective over a 2-year horizon. Conclusions: AI achieves dermatologist-level diagnostic accuracy in controlled settings; however, real-world evidence, algorithmic equity across skin phototypes, and health economic viability remain critical unresolved challenges. Prospective validation, mandatory demographic subgroup reporting, and cost-effectiveness modeling are essential prerequisites for safe and equitable clinical implementation.
Nina Ivanovic, Marius Florentin Popa, Ana-Olivia Toma et al.· Diagnostics· 0 citations
Background: Artificial intelligence (AI) is increasingly being incorporated into radiology, not only for image interpretation but also for scheduling, examination protocoling, image acquisition, reconstruction, worklist prioritisation, quantitative analysis, reporting, communication, and follow-up. The clinical value of these systems depends on more than algorithmic accuracy. It also depends on interoperability, usability, external validation, human oversight, institutional readiness, and the ability to demonstrate measurable improvement in patient care. Objective: This review examines how AI is reshaping radiology workflow, summarises clinically relevant applications across imaging modalities, evaluates evidence regarding diagnostic performance and operational efficiency, and discusses implementation, ethics, regulation, workforce, economic, and equity-related considerations. Methods: A structured narrative review framework was developed using PubMed/MEDLINE, Embase, Scopus, Web of Science, Cochrane Library, PubMed Central, major radiology journals, and publicly available regulatory and professional sources. Original clinical research articles published between January 2016 and August 2026 were prioritised. Studies were considered when they evaluated an AI application in clinical imaging, measured diagnostic or workflow outcomes, or described prospective implementation. Because the studies differed substantially in task, population, modality, endpoint, and reference standard, findings were synthesised narratively rather than pooled statistically. Results: AI has demonstrated value in selected tasks involving mammographic screening, chest-radiograph interpretation, CT triage, MRI reconstruction, segmentation, quantitative imaging, and worklist prioritisation. Prospective and randomised studies suggest that AI can preserve or improve diagnostic performance while reducing selected forms of reader workload. Nevertheless, reported workflow gains are variable. Benefits may be attenuated by false-positive alerts, additional review requirements, poor system integration, case-mix differences, and local staffing patterns. Evidence connecting AI deployment with improved patient outcomes, cost-effectiveness, and long-term equity remains less mature. Conclusions: AI should be treated as a sociotechnical intervention rather than as a stand-alone software product. The most defensible implementation strategy is to begin with narrowly defined clinical problems, conduct local validation, integrate outputs into existing systems, train users, monitor performance after deployment, and retain accountable human oversight. Sustainable transformation will require prospective multicentre research, transparent reporting, interoperable architectures, lifecycle regulation, and deliberate protection against bias and unequal access.
Neelam Rao Bharti, Deeksha Jaiswal, Nidhi Goswami et al.· Genetics and Molecular Resea...· 0 citations
The present review analyzes the existing context of AI pathology systems, particularly diagnostic precision, clinical validation, and technical systems such as convolutional neural networks and transformers and discusses the integration challenge in clinical workflows for these systems.
Abdul-Mohsen G. Alhejaily, D. Alghamdi· Biomedical Reports· 0 citations
Abstract Objective This article aims to synthesize the diagnostic accuracy of artificial intelligence (AI) for CT-based diagnosis of major respiratory diseases (COVID-19, tuberculosis [TB], COPD/ILA, lung nodules/cancer) between 2020 and 2025, and to identify barriers to clinical adoption spanning data standardization, interpretability, workflow integration, and radiation protection. Materials and Methods Following PRISMA 2020, we searched PubMed/MEDLINE, Scopus, Web of Science, IEEE Xplore, and screened preprints (January 2020 to September 2025). Eligible human studies reported the diagnostic performance of AI (ML/DL/CNN/CAD) using CT. Primary outcomes were sensitivity, specificity, and AUC; secondary outcomes included CT dose metrics, explainability, and workflow effects. Risk of bias was assessed using QUADAS-2/PROBAST-AI; reporting quality was assessed using CLAIM. Bivariate random-effects meta-analysis yielded pooled estimates with HSROC; heterogeneity ( τ 2 , I 2 ) and publication bias (Deeks) were assessed. Certainty was graded using GRADE for DTA. Results Thirty-nine studies met criteria (predominantly CT; CNN-based). Pooled sensitivity/specificity were as follows: TB 0.895/0.935, COVID-19 0.872/0.914, nodules/cancer 0.886/0.869, COPD/ILA 0.856/0.844; heterogeneity was extreme ( I 2 ≈ 99–100%). PROBAST-AI indicated highest concerns in analysis and predictors; CLAIM adherence was uneven (external validation: 33%, prospective evaluation: 12%). AI reduced reporting time by ∼20 to 40% and supported low-dose CT with ∼25 to 40% CTDIvol reductions while maintaining sensitivity. Deeks' plots suggested modest asymmetry. GRADE certainty was moderate (COVID-19, nodules/cancer), low–moderate (COPD), and low (TB). Conclusion AI demonstrates promising diagnostic performance across respiratory CT tasks but faces generalizability, bias, and reporting gaps. Prospective multicenter validation, standardized protocols and dose reporting, calibrated/transparent models with quantitatively validated explainability, and assistive workflow deployment are essential for safe, reliable clinical adoption.
M. Sadeghi, S. Sina, Mohammad-Reza Mohammadian-Behbahani et al.· Indian Journal of Radiology...· 0 citations
ABSTRACT Background and Aims Artificial intelligence (AI) is increasingly being integrated into neurosurgical practice, offering capabilities in diagnostic imaging, surgical planning, and outcome prediction. However, the “black box” nature of many AI systems generating recommendations without a transparent rationale poses fundamental challenges to adoption in a specialty defined by high‐stakes, irreversible interventions. This review critically examines whether AI is ready to reshape neurosurgical decision‐making, synthesizing current evidence while systematically analyzing technical, ethical, and regulatory barriers to clinical integration. Methods A systematic search of PubMed, Scopus, and Web of Science databases was conducted for peer‐reviewed studies published between January 2020 and March 2026. Articles reporting AI applications in neurosurgical diagnosis, prognosis, or intraoperative guidance were included. Data were synthesized thematically across clinical domains, with a focus on model interpretability, validation status, and implementation barriers. Results AI demonstrates significant capabilities: diagnostic accuracy exceeding AUC 0.90 in tumor classification, prognostic improvements up to 15% over traditional methods, and 10%–20% complication reductions with AI‐assisted planning. Currently, FDA‐cleared tools enable automated tumor segmentation, aneurysm detection, and spinal navigation. However, critical gaps persist: external validation remains rare (< 20% of studies; e.g., 10 of 60 cerebrovascular studies (16.7%) reported external validation, with pooled AUC 0.84 [95% CI, 0.79–0.88] for thrombectomy outcome), most models are trained on homogeneous single‐center datasets, and the “black box” problem limits clinician trust. From an implementation science perspective, human factors, including workflow integration and cognitive load, remain underexplored. Conclusion AI is not ready for independent decision‐making in neurosurgery but serves as a powerful augmentative tool when limitations are transparently addressed. Lessons from neurosurgery offer a blueprint for AI integration across high‐stake medical specialties. The black box must be opened before AI can truly reshape clinical practice.
Artificial intelligence (AI)-based medical imaging diagnosis has demonstrated remarkable performance across multiple clinical domains, with deep learning models frequently reporting diagnostic accuracy, sensitivity, and specificity exceeding 90% under controlled experimental conditions. However, translating these results into clinically reliable, regulatory-compliant systems remains a critical challenge. As a narrative survey rather than an original benchmark study, this paper reports no new experimental results; instead, it introduces a modality-aware analytical framework organizing the existing literature across four dimensions: imaging modality, data provenance, validation maturity, and model architecture. Using this taxonomy, the survey synthesizes unimodal and multimodal fusion approaches spanning radiology (CT, MRI, X-ray), pathology (whole-slide images), ophthalmology (fundus photography, OCT), and multi-source fusion combining imaging with electronic health records (EHR) and genomic data. The synthesis indicates that high reported accuracy is strongly contingent on data characteristics and evaluation conditions, with many models relying on low-maturity validation lacking evidence of generalization in real-world settings. To address these limitations, an engineering-oriented deployment framework is proposed, integrating modality-driven model selection, structured preprocessing pipelines, multi-level clinical validation, computational feasibility assessment, and explainability, together with a clinical deployment readiness model spanning validation maturity, data diversity, interpretability, and regulatory alignment. Key challenges include the single-site generalization gap, algorithmic bias across demographic groups, limited clinical adoption of explainable AI, insufficient alignment with regulatory frameworks including FDA 510(k), De Novo, and EU MDR/IVDR pathways, and a continuing need for prospective multicenter validation. Future directions toward federated learning, foundation models, certification-aware design, and multimodal digital biomarker integration are outlined.
Enoch Jacob Dodo, Amos Takai Yayock, Gregory Onwodi et al.· Journal of Science Research...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
