Artificial Intelligence and Failure Mode and Effect Analysis Framework to Reduce Diagnostic Errors in Pathology

Artificial Intelligence and Failure Mode and Effect Analysis Framework to Reduce Diagnostic Errors in Pathology | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Artificial Intelligence and Failure Mode and Effect Analysis Framework to Reduce Diagnostic Errors in Pathology FNU Poombal, Raees Lail, Swati Satturwar, Rajendra Singh, Ehsan Ullah, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9369303/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Diagnostic errors in pathology remain an important contributor to patient morbidity and healthcare inefficiency despite increasing laboratory automation and quality management initiatives. Failure Mode and Effects Analysis (FMEA) offers a structured framework for prospective identification and prioritization of workflow risks across the total testing process. Concurrently, artificial intelligence (AI) has demonstrated substantial capability in image analysis, data validation, and workflow optimisation in both anatomical and clinical pathology. We present our findings from a survey among 43 pathologists who identified 266 errors i.e., (183, 69%) within pre-analytical phase, 59 (22%) errors within analytical phase and 22 (9%) errors within post-analytical phase. Then we present the results of the FMEA along with the most suitable AI approaches to mitigate the errors identified. In the end, we present a conceptual framework to integrate AI particularly agentic AI with FMEA-driven quality governance. Agentic AI enables dynamic monitoring, predictive error detection, and autonomous workflow orchestration across pre-analytical, analytical, and post-analytical phases. Integrating agentic AI with FMEA may enable continuous recalibration of risk models, transition pathology quality programs from reactive to predictive paradigms, and support scalable diagnostic safety ecosystems. Implementation considerations including algorithmic transparency, human–AI interaction, and regulatory governance are discussed. This integrated approach represents a potential next-generation model for improving diagnostic accuracy, operational efficiency, and patient safety across pathology services. Diagnostic errors pathology failure mode and effect analysis agentic artificial intelligence Figures Figure 1 Figure 2 Introduction Pathology services underpin most of the modern medical decision-making, with laboratory results influencing an estimated 60–70% of clinical management decisions. 1 Despite advances in laboratory automation, quality management systems, and accreditation standards, diagnostic errors remain a persistent challenge across both anatomical pathology (AP) 2 and clinical pathology (CP) 3 . Errors in pathology occur across the entire spectrum of testing process, encompassing pre-analytical, analytical, and post-analytical phases. The pre-analytical phase represents the most frequent source of laboratory errors. 3 – 6 Specimen misidentification, incorrect labelling, incomplete clinical information, and suboptimal specimen preservation are consistently identified as high-risk failure modes. Multiple studies have demonstrated that pre-analytical errors account for approximately 60–70% of total laboratory testing errors. 3 , 4 Within anatomical pathology, specimen identification failures remain among the most consequential error categories due to their potential to produce catastrophic diagnostic and treatment consequences. 7 – 9 Transport delays, fixation variability, and incomplete requisition data further contribute to diagnostic uncertainty and may adversely affect downstream molecular and immunohistochemical analyses. 10 Analytical errors in anatomical pathology include tissue processing artefacts, slide preparation defects, immunohistochemical interpretation variability, sampling bias during gross examination, and diagnostic interpretation discrepancies. 11 , 12 Clinical pathology analytical errors frequently involve instrument calibration failures, reagent instability, and automated flagging limitations. 13 Post-analytical failures include reporting delays, transcription errors, ambiguous report language, and miscommunication of critical results. 14 , 15 Increasing complexity of integrated molecular and multimodal diagnostic reporting has further expanded post-analytical vulnerability. Traditional safety interventions have relied heavily on retrospective root cause analysis and incident reporting. 5 However, increasing diagnostic complexity, expanding testing modalities, workforce pressures, and growing data volumes necessitate proactive, systems-based approaches to quality and patient safety. FMEA has emerged as a widely adopted prospective risk-assessment methodology capable of identifying and prioritising workflow vulnerabilities in high risk industries such as airline and healthcare, in latter it can avert harm by minimizing errors. 16 , 17 Simultaneously, digital transformation has enabled the rapid development of artificial intelligence technologies capable of automating image analysis, detecting workflow anomalies, and supporting diagnostic interpretation. 18 – 21 More recently, the emergence of agentic AI, defined as autonomous, goal-directed AI capable of multi-step reasoning and workflow orchestration in similar fields, presents new opportunities for integrated safety management across anatomical and clinical pathology. 22 This paper presents results from our survey on the diagnostic errors, findings from the FMEA analysis on diagnostic errors; proposes a conceptual model for an agentic AI–FMEA integrated safety architecture for error reduction in anatomical pathology; and outlines key implementation considerations. Methods and Findings Firstly, we surveyed 43 pathologists each reporting up to 10 errors in November 2023 to February 2024. 23 We received 266 counts of errors (62%) as the number of errors reported varied among respondents. Out of the total, we tracked back over two-third of errors (183, 69%) within pre-analytical phase, 59 (22%) errors within analytical phase and 22 (9%) errors within post-analytical phase. Figure 1 summarizes the common error types reported at each phase. Failure Mode and Effects Analysis After analysing the survey, we performed FMEA as a team of authors with author EU acting as FMEA process lead and other authors acted as pathologist domain experts. The FMEA was undertaken over three virtual meetings during May to October 2025 to identify the potential failure modes within pathology workflow. Table 1 summarizes the results of the most common failure modes identified by our FMEA process within pre-analytical, analytical and post-analytical phases of pathology workflow along with their severity, occurring and detectability ratings, all of which are multiplied to calculate the RPN for each failure mode. In the last column, we present the most likely AI approach that can be used as a mitigation strategy for each failure mode. Table 1 FMEA Failure Modes Across Pathology Workflow Phase Failure Mode Severity (1–10) Occurrence (1–10) Detectability (1–10) RPN AI Mitigation Strategy Pre-Analytical Specimen mislabeling 10 6 8 480 Computer vision for barcode verification and patient identification matching at collection. Fixation artifacts 7 5 6 210 Predictive analytics to monitor fixation time and temperature conditions. Analytical Tissue processing defect 8 4 5 160 Automated image analysis to detect folds, bubbles, and microtome artifacts. Diagnostic interpretation error 10 3 7 210 AI-based diagnostic decision support and second-read systems. Post-Analytical Transcription error in report 9 4 4 144 NLP for automated report proofreading and consistency checking against structured data. AI Applications as Error Mitigation Strategies After completing FMEA on the pathology workflow and identifying the most imperative failure modes, we identified the key domains of artificial intelligence and their practical applications in pathology workflows and the mechanisms through which these technologies can mitigate the risk and thus avert the failure modes to manifest themselves. Each application contributes to improved system reliability by either reducing the likelihood of errors (occurrence) or increasing the probability that errors are detected before affecting patient care (detectability) within the Failure Mode and Effects Analysis (FMEA) framework, as summarized in Table 2 below. Table 2 AI Technologies and Corresponding Risk Reduction Mechanisms AI Technology Domain Specific Application Risk Reduction Mechanism Target FMEA Domain Computer Vision Automated Slide Quality Control Detects pre-analytical and analytical artifacts before pathologist review. Detectability (Increases likelihood of catching errors) Diagnostic Decision Support (e.g., tumor detection) Highlights areas of interest, reducing search errors and inter-observer variability. Occurrence (Reduces likelihood of diagnostic misses) Natural Language Processing Automated Report Validation Identifies inconsistencies, missing information, and critical value omissions in reports. Detectability (Ensures errors in reporting are caught) Requisition Data Extraction Automates data entry from clinical forms, reducing transcription errors. Occurrence (Reduces likelihood of data entry errors) Predictive Analytics Instrument Failure Prediction Monitoring equipment performance to predict and prevent breakdowns. Occurrence (Reduces instrument-related failures) Workflow Bottleneck Analysis Identifies and predicts delays, allowing for proactive resource allocation. Occurrence (Reduces delays and associated risks) Discussion FMEA is a structured, team-based methodology designed to identify potential failure modes within complex processes and prioritise mitigation strategies. 16 , 17 Each failure mode is evaluated across three domains i.e., severity of potential harm; probability of occurrence; and detectability prior to patient impact. These parameters generate Risk Priority Numbers (RPNs) used to prioritise quality interventions. Application of FMEA within histopathology workflows has successfully identified high-risk process steps including specimen handling, accessioning, gross examination, tissue processing, slide preparation, and reporting. 24 , 25 Studies demonstrate that FMEA-driven interventions, including barcode-based specimen tracking and standardised fixation protocols, significantly reduce specimen identification errors and improve workflow reliability. 26 FMEA has been extensively applied in clinical laboratory testing environments and is also supported by regulators in addressing risks for medical devices. 25 – 28 Implementation within emergency hematology workflows has demonstrated significant reductions in specimen rejection rates, turnaround times, and instrument failure events. 27 Similar applications have been reported in transfusion medicine, molecular diagnostics, and microbiology laboratories. 29 , 30 In Anatomical Pathology, AI has found myriads of useful applications since the advent of digital scanners. From a perspective of using AI to reduce errors, these applications can be broadly categorized into three themes. In image quality assessment, deep learning algorithms have demonstrated high performance in detecting focus errors, tissue folds, staining artefacts, and scanning failures. 31 , 32 Automated quality control systems improve slide adequacy prior to diagnostic interpretation. 33 , 34 In terms of clinical decision support, AI models have demonstrated effectiveness in tumor detection 35 – 38 , histological classification 39 , 40 , and biomarker quantification and prediction. 41 , 42 AI models also provide quantitative and spatially resolved outputs that reduce interobserver variability and support diagnostic consistency. 43 , 44 In addition, NLP tools can detect report inconsistencies, missing diagnostic elements, and contradictory statements. 45 AI systems can also validate clinical metadata completeness and identify diagnostic outliers. 46 Outside of anatomical pathology, AI has been extensively applied with other specialities of clinical laboratory medicine, such as AI tools have been applied to automated haematology interpretation, microbiological image analysis, predictive instrument maintenance, and anomaly detection within laboratory information systems. 47 These tools enhance diagnostic accuracy and operational efficiency. As outlined above, FMEA provides a structured framework for identifying high-risk process steps 24 – 26 , while AI enhances failure detection and process monitoring capabilities. 3 , 16 , 21 , 32 Agentic AI systems represent an evolution beyond static decision-support tools by incorporating autonomous multi-agent workflows capable of dynamic process monitoring and adaptive intervention across industries 48 , 49 including healthcare. 50 , 51 We identify the application of agentic AI functional domains within pathology safety ecosystems below and then discuss how a combined FMEA-Agentic AI framework could play a pivotal role in reducing the errors in pathology practice to near-elimination scale. We propose a robust engineering framework whereby carefully selected AI applications are strategically deployed to address the high-RPN failure modes to improve detectability through automated quality monitoring; to reduce failure occurrence through predictive analytics; and to provide real-time decision support as shown in Fig. 2 . Continuous performance data generated by AI systems can then inform iterative FMEA recalibration. A Specimen Integrity Agent can be deployed to monitor specimen identification, tracking, and transport conditions. An Analytical Quality Agent can autonomously perform slide quality assessment, staining validation, and instrument performance monitoring. A Diagnostic Consistency Agent can be positioned to identify discrepancies between gross, microscopic, molecular, and clinical findings, ideally prior to pathologist reviewing the case. A Reporting Validation Agent can perform NLP-based report verification and ensure compliance with structured reporting standards i.e., synoptic reporting. Once these four types of Agentic AI systems are deployed, this model can autonomously generate continuous operational data that can update FMEA risk scoring in real-time. This closed-loop safety model enables transition from static risk assessment to real-time adaptive and proactive safety governance, as shown in Fig. 2 . This diagram illustrates how Agentic AI and FMEA can be integrated to create a continuously learning, predictive safety ecosystem. Authors would also like to outline key considerations for a meaningful implementation of this model. Automation bias and over-reliance on AI outputs may introduce new safety risks 52 , 53 . Therefore, maintaining human oversight and multidisciplinary governance is essential. Secondly, validation standards, algorithm transparency, and auditability remain key regulatory priorities with every AI deployment. 54 A successful deployment of this model also requires interoperability between digital pathology platforms, laboratory information systems, and enterprise health data ecosystems. Conclusions Integration of artificial intelligence with FMEA provides a powerful framework for comprehensive diagnostic error reduction across pathology and is supported by regulators. The addition of agentic AI workflows offers the potential to create continuously learning, predictive safety ecosystems capable of transforming pathology quality governance. We encourage prospective evaluation of this proposed integrated agentic AI–FMEA safety model, development of standardised performance metrics, and assessment of clinical outcome improvements. Abbreviations AI Artificial Intelligence AP Anatomical Pathology CP Clinical Pathology FMEA Failure Mode and Effect Analysis NLP Natural Language Processing RPN Risk Priority Number Declarations Author Contributions Conceptualization: EU, RS, Writing – Original Draft: P, RL, Writing – Review & Editing: SS, EU, RL; Supervision: RS, AP Funding The authors received no specific funding for this work. Conflicts of Interest The authors declare that they have no competing interests. Ethics Approval Not applicable. This article is an editorial and does not involve human participants, animal subjects, or identifiable patient data. Consent for Publication Not applicable. Data Availability No datasets were generated or analyzed during the preparation of this editorial. References Sox HC, Higgins MC, Owens DK, Schmidler GS. Medical decision making . John Wiley & Sons; 2024. Nakhleh RE, Nosé V, Colasacco C, et al. Interpretive diagnostic error reduction in surgical pathology and cytology: guideline from the College of American Pathologists Pathology and Laboratory Quality Center and the Association of Directors of Anatomic and Surgical Pathology. Archives of Pathology & Laboratory Medicine . 2016;140(1):29-40. Nordin N, Ab Rahim SN, Omar WFAW, et al. Preanalytical errors in clinical laboratory testing at a glance: source and control measures. Cureus . 2024;16(3) Lima-Oliveira G, Volanski W, Lippi G, Picheth G, Guidi GC. Pre-analytical phase management: a review of the procedures from patient preparation to laboratory analysis. Scandinavian journal of clinical and laboratory investigation . 2017;77(3):153-163. Morias C, Palmer G, Santhakumar A. Pre-analytical errors and patient outcomes in the total testing process. Australian Journal of Medical Science . 2018;39(1/2):14-27. Plebani M, Sciacovelli L, Aita A, Padoan A, Chiozza ML. Quality indicators to detect pre-analytical errors in laboratory testing. Clinica chimica acta . 2014;432:44-48. Brunyé TT, Balla A, Drew T, et al. From image to diagnosis: characterizing sources of error in histopathologic interpretation. Modern Pathology . 2023;36(7):100162. Raab SS, Grzybicki DM, Janosky JE, et al. Clinical impact and frequency of anatomic pathology errors in cancer diagnoses. Cancer: Interdisciplinary International Journal of the American Cancer Society . 2005;104(10):2205-2213. Zarbo RJ, Meier FA, Raab SS. Error detection in anatomic pathology. Archives of pathology & laboratory medicine . 2005;129(10):1237-1245. Bauer DR, Otter M, Chafin DR. A new paradigm for tissue diagnostics: tools and techniques to standardize tissue collection, transport, and fixation. Current pathobiology reports . 2018;6(2):135-143. Correia HMV, Bernardo S, Esteves F, Garcia C. Application of strategies to minimize the error in Pathological Anatomy. Millenium-Journal of Education, Technologies, and Health . 2017:95-106. Murphy CE, Knierim RH, Benjamin Pardue J. Anatomic and Clinical Pathology. Understanding The Principles and Practice of Legal Oncology . 2022:145. Schultze A, Irizarry A. Recognizing and reducing analytical errors and sources of variation in clinical pathology data in safety assessment studies. Toxicologic pathology . 2017;45(2):281-287. Plebani M. Harmonizing the post-analytical phase: focus on the laboratory report. Clinical Chemistry and Laboratory Medicine (CCLM) . 2024;62(6):1053-1062. Walz SE, Darcy TP. Patient safety & post-analytical error. Clinics in laboratory medicine . 2013;33(1):183-194. Faiella G, Parand A, Franklin BD, et al. Expanding healthcare failure mode and effect analysis: A composite proactive risk analysis approach. Reliability Engineering & System Safety . 2018;169:117-126. Ullah E, Baig MM, GholamHosseini H, Lu J. Failure mode and effect analysis (FMEA) to identify and mitigate failures in a hospital rapid response system (RRS). Heliyon . 2022;8(2) Baxi V, Edwards R, Montalto M, Saha S. Digital pathology and artificial intelligence in translational medicine and clinical practice. Modern Pathology . 2022;35(1):23-32. Brodsky V, Ullah E, Bychkov A, et al. Generative artificial intelligence in anatomic pathology. Archives of pathology & laboratory medicine . 2025;149(4):298-318. Cifci D, Veldhuizen GP, Foersch S, Kather JN. AI in computational pathology of cancer: improving diagnostic workflows and clinical outcomes? Annual Review of Cancer Biology . 2023;7(1):57-71. Song AH, Jaume G, Williamson DF, et al. Artificial intelligence for digital and computational pathology. Nature Reviews Bioengineering . 2023;1(12):930-949. Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK. Agentic artificial intelligence: the power to change medicine and our world. Journal of the American College of Radiology . 2025; Satturwar S, Singh R, Ullah E, Poombal F. Identifying and Addressing Errors in Anatomical Pathology: A Survey Analysis. American Journal of Clinical Pathology . 2024;162:S98-S99. Angelilli S, Feathers L, McGonigal M. Decreasing Surgical Specimen-Handling Errors: A Network Quality Improvement Journey. PATIENT SAFETY . 2025;7(2) Ianni A, Virgili R, Piredda M, DE MARINIS MG, Petitti T. Healthcare safety management: The model of FMEA/FMECA in anatomic pathology service. MADRIDGE JOURNAL OF CLINICAL RESEARCH . 2018;2(1):47-50. Samimi A. Application of failure mode and effects analysis (FMEA) for risk management in hospital laboratories. Medicinal, Psychological, and Health Research Journal (mphrj) . 2025;1(5):160-168. Lv S, Sun Y, Zhang J, Jin T, Hu X. Application of the FMEA method in improving the quality management of emergency complete blood count testing. Laboratory Medicine . 2023;54(6):574-581. Van Hoof V, Bench S, Soto AB, et al. Failure Mode and Effects Analysis (FMEA) at the preanalytical phase for POCT blood gas analysis: proposal for a shared proactive risk analysis model. Clinical Chemistry and Laboratory Medicine (CCLM) . 2022;60(8):1186-1201. Carey RB, Bhattacharyya S, Kehl SC, et al. Practical guidance for clinical microbiology laboratories: implementing a quality management system in the medical microbiology laboratory. Clinical microbiology reviews . 2018;31(3):10.1128/cmr. 00062-17. Claxton K, Campbell-Allen NM. Failure modes effects analysis (FMEA) for review of a diagnostic genetic laboratory process. International journal of quality & reliability management . 2017;34(2):265-277. Kanwal N, Pérez-Bueno F, Schmidt A, Engan K, Molina R. The devil is in the details: Whole slide image acquisition and processing for artifacts detection, color variation, and data augmentation: A review. Ieee Access . 2022;10:58821-58844. Shakhawat H, Hossain S, Kabir A, Mahmud SH, Islam MM, Tariq F. Review of artifact detection methods for automated analysis and diagnosis in digital pathology. Artificial Intelligence for Disease Diagnosis and Prognosis in Smart Healthcare . CRC Press; 2023:177-202. Janowczyk A, Zuo R, Gilmore H, Feldman M, Madabhushi A. HistoQC: an open-source quality control tool for digital pathology slides. JCO clinical cancer informatics . 2019;3:1-7. Ko YS, Choi YM, Kim M, et al. Improving quality control in the routine practice for histopathological interpretation of gastrointestinal endoscopic biopsies using artificial intelligence. PLoS One . 2022;17(12):e0278542. Chukwujindu E, Faiz H, Sara A-D, Faiz K, De Sequeira A. Role of artificial intelligence in brain tumour imaging. European journal of radiology . 2024;176:111509. Cruz-Roa A, Gilmore H, Basavanhally A, et al. Accurate and reproducible invasive breast cancer detection in whole-slide images: A Deep Learning approach for quantifying tumor extent. Scientific reports . 2017;7(1):46450. Tolkach Y, Wolgast LM, Damanakis A, et al. Artificial intelligence for tumour tissue detection and histological regression grading in oesophageal adenocarcinomas: a retrospective algorithm development and validation study. The Lancet Digital Health . 2023;5(5):e265-e275. Zhang Z, Chen P, McGough M, et al. Pathologist-level interpretable whole-slide cancer diagnosis with deep learning. Nature Machine Intelligence . 2019;1(5):236-245. Chen C-L, Chen C-C, Yu W-H, et al. An annotation-free whole-slide training approach to pathological classification of lung cancer types using deep learning. Nature communications . 2021;12(1):1193. Korbar B, Olofson AM, Miraflor AP, et al. Deep learning for classification of colorectal polyps on whole-slide images. Journal of pathology informatics . 2017;8(1):30. Li H, Qin J, Li Z, et al. Systematic review and meta-analysis of deep learning for MSI-H in colorectal cancer whole slide images. npj Digital Medicine . 2025;8(1):456. van Eekelen L, Spronck J, Looijen-Salamon M, et al. Comparing deep learning and pathologist quantification of cell-level PD-L1 expression in non-small cell lung cancer whole-slide images. Scientific reports . 2024;14(1):7136. Jiao Y, Li J, Qian C, Fei S. Deep learning-based tumor microenvironment analysis in colon adenocarcinoma histopathological whole-slide images. Computer Methods and Programs in Biomedicine . 2021;204:106047. Zhang X, Yu X, Liang W, et al. Deep learning‐based accurate diagnosis and quantitative evaluation of microvascular invasion in hepatocellular carcinoma on whole‐slide histopathology images. Cancer Medicine . 2024;13(5):e7104. López-Úbeda P, Martín-Noguerol T, Aneiros-Fernández J, Luna A. Natural language processing in pathology: current trends and future insights. The American Journal of Pathology . 2022;192(11):1486-1495. Lebeau A, Turzynski A. Pathology: diagnostics, reporting and artificial intelligence. Senologie-Zeitschrift für Mammadiagnostik und-therapie . 2025;22(01):31-42. Naugler C, Church DL. Automation and artificial intelligence in the clinical laboratory. Critical reviews in clinical laboratory sciences . 2019;56(2):98-110. Acharya DB, Kuppan K, Divya B. Agentic AI: Autonomous intelligence for complex goals—A comprehensive survey. IEEe Access . 2025;13:18912-18936. Murugesan S. The rise of agentic AI: implications, concerns, and the path forward. IEEE Intelligent Systems . 2025;40(2):8-14. Karunanayake N. Next-generation agentic AI for transforming healthcare. Informatics and Health . 2025;2(2):73-83. Shaktah LA, Carrero ZI, Hewitt KJ, et al. Application of artificial intelligence and digital tools in cancer pathology. The Lancet Digital Health . 2025;7(10) Bellahsen-Harrar Y, Lubrano M, Lépine C, et al. Exploring the risks of over-reliance on AI in diagnostic pathology. What lessons can be learned to support the training of young pathologists? PLoS One . 2025;20(8):e0323270. Sunday O. Behavioral Impacts of AI Reliance in Diagnostics: Balancing Automation with Skill Retention. Epidemiol Health Data Insights . 2025;1:ehdi014. Mourya A, Jobanputra B, Pai R. AI-powered clinical trials and the imperative for regulatory transparency and accountability. Health and Technology . 2024;14(6):1071-1081. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 01 May, 2026 Reviews received at journal 28 Apr, 2026 Reviewers agreed at journal 22 Apr, 2026 Reviewers invited by journal 21 Apr, 2026 Editor assigned by journal 15 Apr, 2026 Submission checks completed at journal 15 Apr, 2026 First submitted to journal 09 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9369303","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":631225423,"identity":"b6f52728-df30-4042-8fed-333b8c4a7043","order_by":0,"name":"FNU Poombal","email":"data:image/png;base64,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","orcid":"","institution":"UMass Chan Medical School-Baystate Medical Center","correspondingAuthor":true,"prefix":"","firstName":"FNU","middleName":"","lastName":"Poombal","suffix":""},{"id":631225424,"identity":"fea08bce-cee3-4d12-9be5-88009a9aa2d9","order_by":1,"name":"Raees Lail","email":"","orcid":"","institution":"Sahiwal Medical College","correspondingAuthor":false,"prefix":"","firstName":"Raees","middleName":"","lastName":"Lail","suffix":""},{"id":631225425,"identity":"928a20a9-1001-48c8-b756-bbfa8a0d1669","order_by":2,"name":"Swati Satturwar","email":"","orcid":"","institution":"The Ohio State University","correspondingAuthor":false,"prefix":"","firstName":"Swati","middleName":"","lastName":"Satturwar","suffix":""},{"id":631225426,"identity":"40ca6317-3ecf-49db-a2fd-7bc648c6d869","order_by":3,"name":"Rajendra Singh","email":"","orcid":"","institution":"University of Pennsylvania","correspondingAuthor":false,"prefix":"","firstName":"Rajendra","middleName":"","lastName":"Singh","suffix":""},{"id":631225427,"identity":"7adaaab4-e514-4f4b-bb70-981e08e66779","order_by":4,"name":"Ehsan Ullah","email":"","orcid":"","institution":"The Ohio State University","correspondingAuthor":false,"prefix":"","firstName":"Ehsan","middleName":"","lastName":"Ullah","suffix":""},{"id":631225428,"identity":"c702fdfe-92d3-4590-8c3a-9f53453cd391","order_by":5,"name":"Anil Parwani","email":"","orcid":"","institution":"The Ohio State University","correspondingAuthor":false,"prefix":"","firstName":"Anil","middleName":"","lastName":"Parwani","suffix":""}],"badges":[],"createdAt":"2026-04-09 13:38:07","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9369303/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9369303/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108388851,"identity":"f059bea4-9a02-429b-846d-76cf62da5f9c","added_by":"auto","created_at":"2026-05-04 06:44:10","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":435333,"visible":true,"origin":"","legend":"\u003cp\u003eProportional distribution of errors across the total testing process in pathology, highlighting representative failure modes, illustrating that while errors occur at every stage, they are disproportionately concentrated in the pre-analytical phase.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9369303/v1/43e5d0ec78d62138343568f8.png"},{"id":108492319,"identity":"09667a68-7ec5-4dbb-b706-4c27edfd68bb","added_by":"auto","created_at":"2026-05-05 09:57:26","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":431879,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eClosed-Loop Agentic AI–FMEA Safety Architecture.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9369303/v1/3b228e60414a5da22bf9035b.png"},{"id":108494996,"identity":"cdb2c7f6-3f6d-45c5-ab9a-9a112aa63dc3","added_by":"auto","created_at":"2026-05-05 10:08:27","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1168472,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9369303/v1/df212926-dbfa-4095-884e-81ff98800f77.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Artificial Intelligence and Failure Mode and Effect Analysis Framework to Reduce Diagnostic Errors in Pathology","fulltext":[{"header":"Introduction","content":"\u003cp\u003ePathology services underpin most of the modern medical decision-making, with laboratory results influencing an estimated 60\u0026ndash;70% of clinical management decisions.\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e Despite advances in laboratory automation, quality management systems, and accreditation standards, diagnostic errors remain a persistent challenge across both anatomical pathology (AP)\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e and clinical pathology (CP)\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eErrors in pathology occur across the entire spectrum of testing process, encompassing pre-analytical, analytical, and post-analytical phases. The pre-analytical phase represents the most frequent source of laboratory errors.\u003csup\u003e\u003cspan additionalcitationids=\"CR4 CR5\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e Specimen misidentification, incorrect labelling, incomplete clinical information, and suboptimal specimen preservation are consistently identified as high-risk failure modes. Multiple studies have demonstrated that pre-analytical errors account for approximately 60\u0026ndash;70% of total laboratory testing errors.\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e Within anatomical pathology, specimen identification failures remain among the most consequential error categories due to their potential to produce catastrophic diagnostic and treatment consequences.\u003csup\u003e\u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e Transport delays, fixation variability, and incomplete requisition data further contribute to diagnostic uncertainty and may adversely affect downstream molecular and immunohistochemical analyses.\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e Analytical errors in anatomical pathology include tissue processing artefacts, slide preparation defects, immunohistochemical interpretation variability, sampling bias during gross examination, and diagnostic interpretation discrepancies.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e Clinical pathology analytical errors frequently involve instrument calibration failures, reagent instability, and automated flagging limitations.\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e Post-analytical failures include reporting delays, transcription errors, ambiguous report language, and miscommunication of critical results.\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e Increasing complexity of integrated molecular and multimodal diagnostic reporting has further expanded post-analytical vulnerability.\u003c/p\u003e \u003cp\u003eTraditional safety interventions have relied heavily on retrospective root cause analysis and incident reporting.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e However, increasing diagnostic complexity, expanding testing modalities, workforce pressures, and growing data volumes necessitate proactive, systems-based approaches to quality and patient safety. FMEA has emerged as a widely adopted prospective risk-assessment methodology capable of identifying and prioritising workflow vulnerabilities in high risk industries such as airline and healthcare, in latter it can avert harm by minimizing errors.\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e Simultaneously, digital transformation has enabled the rapid development of artificial intelligence technologies capable of automating image analysis, detecting workflow anomalies, and supporting diagnostic interpretation.\u003csup\u003e\u003cspan additionalcitationids=\"CR19 CR20\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e More recently, the emergence of agentic AI, defined as autonomous, goal-directed AI capable of multi-step reasoning and workflow orchestration in similar fields, presents new opportunities for integrated safety management across anatomical and clinical pathology.\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eThis paper presents results from our survey on the diagnostic errors, findings from the FMEA analysis on diagnostic errors; proposes a conceptual model for an agentic AI\u0026ndash;FMEA integrated safety architecture for error reduction in anatomical pathology; and outlines key implementation considerations.\u003c/p\u003e"},{"header":"Methods and Findings","content":"\u003cp\u003eFirstly, we surveyed 43 pathologists each reporting up to 10 errors in November 2023 to February 2024.\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e We received 266 counts of errors (62%) as the number of errors reported varied among respondents. Out of the total, we tracked back over two-third of errors (183, 69%) within pre-analytical phase, 59 (22%) errors within analytical phase and 22 (9%) errors within post-analytical phase. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e summarizes the common error types reported at each phase.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eFailure Mode and Effects Analysis\u003c/h2\u003e \u003cp\u003eAfter analysing the survey, we performed FMEA as a team of authors with author EU acting as FMEA process lead and other authors acted as pathologist domain experts. The FMEA was undertaken over three virtual meetings during May to October 2025 to identify the potential failure modes within pathology workflow. Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e summarizes the results of the most common failure modes identified by our FMEA process within pre-analytical, analytical and post-analytical phases of pathology workflow along with their severity, occurring and detectability ratings, all of which are multiplied to calculate the RPN for each failure mode. In the last column, we present the most likely AI approach that can be used as a mitigation strategy for each failure mode.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eFMEA Failure Modes Across Pathology Workflow\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhase\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFailure Mode\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSeverity (1\u0026ndash;10)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOccurrence (1\u0026ndash;10)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eDetectability (1\u0026ndash;10)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRPN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAI Mitigation Strategy\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePre-Analytical\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSpecimen mislabeling\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e480\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eComputer vision for barcode verification and patient identification matching at collection.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFixation artifacts\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e210\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003ePredictive analytics to monitor fixation time and temperature conditions.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAnalytical\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTissue processing defect\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e160\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAutomated image analysis to detect folds, bubbles, and microtome artifacts.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDiagnostic interpretation error\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e210\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAI-based diagnostic decision support and second-read systems.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePost-Analytical\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTranscription error in report\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e144\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eNLP for automated report proofreading and consistency checking against structured data.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eAI Applications as Error Mitigation Strategies\u003c/h3\u003e\n\u003cp\u003eAfter completing FMEA on the pathology workflow and identifying the most imperative failure modes, we identified the key domains of artificial intelligence and their practical applications in pathology workflows and the mechanisms through which these technologies can mitigate the risk and thus avert the failure modes to manifest themselves. Each application contributes to improved system reliability by either reducing the likelihood of errors (occurrence) or increasing the probability that errors are detected before affecting patient care (detectability) within the Failure Mode and Effects Analysis (FMEA) framework, as summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e below.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAI Technologies and Corresponding Risk Reduction Mechanisms\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAI Technology Domain\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSpecific Application\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRisk Reduction Mechanism\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTarget FMEA Domain\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003eComputer Vision\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAutomated Slide Quality Control\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDetects pre-analytical and analytical artifacts before pathologist review.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eDetectability\u003c/b\u003e (Increases likelihood of catching errors)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDiagnostic Decision Support (e.g., tumor detection)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHighlights areas of interest, reducing search errors and inter-observer variability.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eOccurrence\u003c/b\u003e (Reduces likelihood of diagnostic misses)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003eNatural Language Processing\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAutomated Report Validation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIdentifies inconsistencies, missing information, and critical value omissions in reports.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eDetectability\u003c/b\u003e (Ensures errors in reporting are caught)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRequisition Data Extraction\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAutomates data entry from clinical forms, reducing transcription errors.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eOccurrence\u003c/b\u003e (Reduces likelihood of data entry errors)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003ePredictive Analytics\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eInstrument Failure Prediction\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMonitoring equipment performance to predict and prevent breakdowns.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eOccurrence\u003c/b\u003e (Reduces instrument-related failures)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWorkflow Bottleneck Analysis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIdentifies and predicts delays, allowing for proactive resource allocation.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eOccurrence\u003c/b\u003e (Reduces delays and associated risks)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eFMEA is a structured, team-based methodology designed to identify potential failure modes within complex processes and prioritise mitigation strategies.\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e Each failure mode is evaluated across three domains i.e., severity of potential harm; probability of occurrence; and detectability prior to patient impact. These parameters generate Risk Priority Numbers (RPNs) used to prioritise quality interventions. Application of FMEA within histopathology workflows has successfully identified high-risk process steps including specimen handling, accessioning, gross examination, tissue processing, slide preparation, and reporting.\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e Studies demonstrate that FMEA-driven interventions, including barcode-based specimen tracking and standardised fixation protocols, significantly reduce specimen identification errors and improve workflow reliability.\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e FMEA has been extensively applied in clinical laboratory testing environments and is also supported by regulators in addressing risks for medical devices.\u003csup\u003e\u003cspan additionalcitationids=\"CR26 CR27\" citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e Implementation within emergency hematology workflows has demonstrated significant reductions in specimen rejection rates, turnaround times, and instrument failure events.\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e Similar applications have been reported in transfusion medicine, molecular diagnostics, and microbiology laboratories.\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eIn Anatomical Pathology, AI has found myriads of useful applications since the advent of digital scanners. From a perspective of using AI to reduce errors, these applications can be broadly categorized into three themes. In image quality assessment, deep learning algorithms have demonstrated high performance in detecting focus errors, tissue folds, staining artefacts, and scanning failures.\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e Automated quality control systems improve slide adequacy prior to diagnostic interpretation.\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e In terms of clinical decision support, AI models have demonstrated effectiveness in tumor detection\u003csup\u003e\u003cspan additionalcitationids=\"CR36 CR37\" citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, histological classification\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e,\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e, and biomarker quantification and prediction.\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e AI models also provide quantitative and spatially resolved outputs that reduce interobserver variability and support diagnostic consistency.\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e,\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e In addition, NLP tools can detect report inconsistencies, missing diagnostic elements, and contradictory statements.\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e AI systems can also validate clinical metadata completeness and identify diagnostic outliers.\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e Outside of anatomical pathology, AI has been extensively applied with other specialities of clinical laboratory medicine, such as AI tools have been applied to automated haematology interpretation, microbiological image analysis, predictive instrument maintenance, and anomaly detection within laboratory information systems.\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e These tools enhance diagnostic accuracy and operational efficiency.\u003c/p\u003e \u003cp\u003eAs outlined above, FMEA provides a structured framework for identifying high-risk process steps\u003csup\u003e\u003cspan additionalcitationids=\"CR25\" citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e, while AI enhances failure detection and process monitoring capabilities.\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e Agentic AI systems represent an evolution beyond static decision-support tools by incorporating autonomous multi-agent workflows capable of dynamic process monitoring and adaptive intervention across industries\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e,\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e including healthcare.\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e,\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e We identify the application of agentic AI functional domains within pathology safety ecosystems below and then discuss how a combined FMEA-Agentic AI framework could play a pivotal role in reducing the errors in pathology practice to near-elimination scale.\u003c/p\u003e \u003cp\u003eWe propose a robust engineering framework whereby carefully selected AI applications are strategically deployed to address the high-RPN failure modes to improve detectability through automated quality monitoring; to reduce failure occurrence through predictive analytics; and to provide real-time decision support as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. Continuous performance data generated by AI systems can then inform iterative FMEA recalibration. A \u003cem\u003eSpecimen Integrity Agent\u003c/em\u003e can be deployed to monitor specimen identification, tracking, and transport conditions. An \u003cem\u003eAnalytical Quality Agent\u003c/em\u003e can autonomously perform slide quality assessment, staining validation, and instrument performance monitoring. A \u003cem\u003eDiagnostic Consistency Agent\u003c/em\u003e can be positioned to identify discrepancies between gross, microscopic, molecular, and clinical findings, ideally prior to pathologist reviewing the case. A \u003cem\u003eReporting Validation Agent\u003c/em\u003e can perform NLP-based report verification and ensure compliance with structured reporting standards i.e., synoptic reporting. Once these four types of Agentic AI systems are deployed, this model can autonomously generate continuous operational data that can update FMEA risk scoring in real-time. This closed-loop safety model enables transition from static risk assessment to real-time adaptive and proactive safety governance, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis diagram illustrates how Agentic AI and FMEA can be integrated to create a continuously learning, predictive safety ecosystem.\u003c/p\u003e \u003cp\u003eAuthors would also like to outline key considerations for a meaningful implementation of this model. Automation bias and over-reliance on AI outputs may introduce new safety risks\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e,\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. Therefore, maintaining human oversight and multidisciplinary governance is essential. Secondly, validation standards, algorithm transparency, and auditability remain key regulatory priorities with every AI deployment.\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e A successful deployment of this model also requires interoperability between digital pathology platforms, laboratory information systems, and enterprise health data ecosystems.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIntegration of artificial intelligence with FMEA provides a powerful framework for comprehensive diagnostic error reduction across pathology and is supported by regulators. The addition of agentic AI workflows offers the potential to create continuously learning, predictive safety ecosystems capable of transforming pathology quality governance. We encourage prospective evaluation of this proposed integrated agentic AI\u0026ndash;FMEA safety model, development of standardised performance metrics, and assessment of clinical outcome improvements.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eArtificial Intelligence\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eAnatomical Pathology\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eClinical Pathology\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eFMEA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eFailure Mode and Effect Analysis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eNLP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eNatural Language Processing\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRPN\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRisk Priority Number\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization: EU, RS, Writing – Original Draft: P, RL, Writing – Review \u0026amp; Editing: SS, EU, RL; Supervision: RS, AP\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors received no specific funding for this work.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable. This article is an editorial and does not involve human participants, animal subjects, or identifiable patient data.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for Publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo datasets were generated or analyzed during the preparation of this editorial.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eSox HC, Higgins MC, Owens DK, Schmidler GS. \u003cem\u003eMedical decision making\u003c/em\u003e. John Wiley \u0026amp; Sons; 2024.\u003c/li\u003e\n \u003cli\u003eNakhleh RE, Nos\u0026eacute; V, Colasacco C, et al. Interpretive diagnostic error reduction in surgical pathology and cytology: guideline from the College of American Pathologists Pathology and Laboratory Quality Center and the Association of Directors of Anatomic and Surgical Pathology. \u003cem\u003eArchives of Pathology \u0026amp; Laboratory Medicine\u003c/em\u003e. 2016;140(1):29-40.\u003c/li\u003e\n \u003cli\u003eNordin N, Ab Rahim SN, Omar WFAW, et al. Preanalytical errors in clinical laboratory testing at a glance: source and control measures. \u003cem\u003eCureus\u003c/em\u003e. 2024;16(3)\u003c/li\u003e\n \u003cli\u003eLima-Oliveira G, Volanski W, Lippi G, Picheth G, Guidi GC. Pre-analytical phase management: a review of the procedures from patient preparation to laboratory analysis. \u003cem\u003eScandinavian journal of clinical and laboratory investigation\u003c/em\u003e. 2017;77(3):153-163.\u003c/li\u003e\n \u003cli\u003eMorias C, Palmer G, Santhakumar A. Pre-analytical errors and patient outcomes in the total testing process. \u003cem\u003eAustralian Journal of Medical Science\u003c/em\u003e. 2018;39(1/2):14-27.\u003c/li\u003e\n \u003cli\u003ePlebani M, Sciacovelli L, Aita A, Padoan A, Chiozza ML. Quality indicators to detect pre-analytical errors in laboratory testing. \u003cem\u003eClinica chimica acta\u003c/em\u003e. 2014;432:44-48.\u003c/li\u003e\n \u003cli\u003eBruny\u0026eacute; TT, Balla A, Drew T, et al. From image to diagnosis: characterizing sources of error in histopathologic interpretation. \u003cem\u003eModern Pathology\u003c/em\u003e. 2023;36(7):100162.\u003c/li\u003e\n \u003cli\u003eRaab SS, Grzybicki DM, Janosky JE, et al. Clinical impact and frequency of anatomic pathology errors in cancer diagnoses. \u003cem\u003eCancer: Interdisciplinary International Journal of the American Cancer Society\u003c/em\u003e. 2005;104(10):2205-2213.\u003c/li\u003e\n \u003cli\u003eZarbo RJ, Meier FA, Raab SS. Error detection in anatomic pathology. \u003cem\u003eArchives of pathology \u0026amp; laboratory medicine\u003c/em\u003e. 2005;129(10):1237-1245.\u003c/li\u003e\n \u003cli\u003eBauer DR, Otter M, Chafin DR. A new paradigm for tissue diagnostics: tools and techniques to standardize tissue collection, transport, and fixation. \u003cem\u003eCurrent pathobiology reports\u003c/em\u003e. 2018;6(2):135-143.\u003c/li\u003e\n \u003cli\u003eCorreia HMV, Bernardo S, Esteves F, Garcia C. Application of strategies to minimize the error in Pathological Anatomy. \u003cem\u003eMillenium-Journal of Education, Technologies, and Health\u003c/em\u003e. 2017:95-106.\u003c/li\u003e\n \u003cli\u003eMurphy CE, Knierim RH, Benjamin Pardue J. Anatomic and Clinical Pathology. \u003cem\u003eUnderstanding The Principles and Practice of Legal Oncology\u003c/em\u003e. 2022:145.\u003c/li\u003e\n \u003cli\u003eSchultze A, Irizarry A. Recognizing and reducing analytical errors and sources of variation in clinical pathology data in safety assessment studies. \u003cem\u003eToxicologic pathology\u003c/em\u003e. 2017;45(2):281-287.\u003c/li\u003e\n \u003cli\u003ePlebani M. Harmonizing the post-analytical phase: focus on the laboratory report. \u003cem\u003eClinical Chemistry and Laboratory Medicine (CCLM)\u003c/em\u003e. 2024;62(6):1053-1062.\u003c/li\u003e\n \u003cli\u003eWalz SE, Darcy TP. Patient safety \u0026amp; post-analytical error. \u003cem\u003eClinics in laboratory medicine\u003c/em\u003e. 2013;33(1):183-194.\u003c/li\u003e\n \u003cli\u003eFaiella G, Parand A, Franklin BD, et al. Expanding healthcare failure mode and effect analysis: A composite proactive risk analysis approach. \u003cem\u003eReliability Engineering \u0026amp; System Safety\u003c/em\u003e. 2018;169:117-126.\u003c/li\u003e\n \u003cli\u003eUllah E, Baig MM, GholamHosseini H, Lu J. Failure mode and effect analysis (FMEA) to identify and mitigate failures in a hospital rapid response system (RRS). \u003cem\u003eHeliyon\u003c/em\u003e. 2022;8(2)\u003c/li\u003e\n \u003cli\u003eBaxi V, Edwards R, Montalto M, Saha S. Digital pathology and artificial intelligence in translational medicine and clinical practice. \u003cem\u003eModern Pathology\u003c/em\u003e. 2022;35(1):23-32.\u003c/li\u003e\n \u003cli\u003eBrodsky V, Ullah E, Bychkov A, et al. Generative artificial intelligence in anatomic pathology. \u003cem\u003eArchives of pathology \u0026amp; laboratory medicine\u003c/em\u003e. 2025;149(4):298-318.\u003c/li\u003e\n \u003cli\u003eCifci D, Veldhuizen GP, Foersch S, Kather JN. AI in computational pathology of cancer: improving diagnostic workflows and clinical outcomes? \u003cem\u003eAnnual Review of Cancer Biology\u003c/em\u003e. 2023;7(1):57-71.\u003c/li\u003e\n \u003cli\u003eSong AH, Jaume G, Williamson DF, et al. Artificial intelligence for digital and computational pathology. \u003cem\u003eNature Reviews Bioengineering\u003c/em\u003e. 2023;1(12):930-949.\u003c/li\u003e\n \u003cli\u003ePowell K, Fishman EK, Chu LC, Rowe SP, Crawford CK. Agentic artificial intelligence: the power to change medicine and our world. \u003cem\u003eJournal of the American College of Radiology\u003c/em\u003e. 2025;\u003c/li\u003e\n \u003cli\u003eSatturwar S, Singh R, Ullah E, Poombal F. Identifying and Addressing Errors in Anatomical Pathology: A Survey Analysis. \u003cem\u003eAmerican Journal of Clinical Pathology\u003c/em\u003e. 2024;162:S98-S99.\u003c/li\u003e\n \u003cli\u003eAngelilli S, Feathers L, McGonigal M. Decreasing Surgical Specimen-Handling Errors: A Network Quality Improvement Journey. \u003cem\u003ePATIENT SAFETY\u003c/em\u003e. 2025;7(2)\u003c/li\u003e\n \u003cli\u003eIanni A, Virgili R, Piredda M, DE MARINIS MG, Petitti T. Healthcare safety management: The model of FMEA/FMECA in anatomic pathology service. \u003cem\u003eMADRIDGE JOURNAL OF CLINICAL RESEARCH\u003c/em\u003e. 2018;2(1):47-50.\u003c/li\u003e\n \u003cli\u003eSamimi A. Application of failure mode and effects analysis (FMEA) for risk management in hospital laboratories. \u003cem\u003eMedicinal, Psychological, and Health Research Journal (mphrj)\u003c/em\u003e. 2025;1(5):160-168.\u003c/li\u003e\n \u003cli\u003eLv S, Sun Y, Zhang J, Jin T, Hu X. Application of the FMEA method in improving the quality management of emergency complete blood count testing. \u003cem\u003eLaboratory Medicine\u003c/em\u003e. 2023;54(6):574-581.\u003c/li\u003e\n \u003cli\u003eVan Hoof V, Bench S, Soto AB, et al. Failure Mode and Effects Analysis (FMEA) at the preanalytical phase for POCT blood gas analysis: proposal for a shared proactive risk analysis model. \u003cem\u003eClinical Chemistry and Laboratory Medicine (CCLM)\u003c/em\u003e. 2022;60(8):1186-1201.\u003c/li\u003e\n \u003cli\u003eCarey RB, Bhattacharyya S, Kehl SC, et al. Practical guidance for clinical microbiology laboratories: implementing a quality management system in the medical microbiology laboratory. \u003cem\u003eClinical microbiology reviews\u003c/em\u003e. 2018;31(3):10.1128/cmr. 00062-17.\u003c/li\u003e\n \u003cli\u003eClaxton K, Campbell-Allen NM. Failure modes effects analysis (FMEA) for review of a diagnostic genetic laboratory process. \u003cem\u003eInternational journal of quality \u0026amp; reliability management\u003c/em\u003e. 2017;34(2):265-277.\u003c/li\u003e\n \u003cli\u003eKanwal N, P\u0026eacute;rez-Bueno F, Schmidt A, Engan K, Molina R. The devil is in the details: Whole slide image acquisition and processing for artifacts detection, color variation, and data augmentation: A review. \u003cem\u003eIeee Access\u003c/em\u003e. 2022;10:58821-58844.\u003c/li\u003e\n \u003cli\u003eShakhawat H, Hossain S, Kabir A, Mahmud SH, Islam MM, Tariq F. Review of artifact detection methods for automated analysis and diagnosis in digital pathology. \u003cem\u003eArtificial Intelligence for Disease Diagnosis and Prognosis in Smart Healthcare\u003c/em\u003e. CRC Press; 2023:177-202.\u003c/li\u003e\n \u003cli\u003eJanowczyk A, Zuo R, Gilmore H, Feldman M, Madabhushi A. HistoQC: an open-source quality control tool for digital pathology slides. \u003cem\u003eJCO clinical cancer informatics\u003c/em\u003e. 2019;3:1-7.\u003c/li\u003e\n \u003cli\u003eKo YS, Choi YM, Kim M, et al. Improving quality control in the routine practice for histopathological interpretation of gastrointestinal endoscopic biopsies using artificial intelligence. \u003cem\u003ePLoS One\u003c/em\u003e. 2022;17(12):e0278542.\u003c/li\u003e\n \u003cli\u003eChukwujindu E, Faiz H, Sara A-D, Faiz K, De Sequeira A. Role of artificial intelligence in brain tumour imaging. \u003cem\u003eEuropean journal of radiology\u003c/em\u003e. 2024;176:111509.\u003c/li\u003e\n \u003cli\u003eCruz-Roa A, Gilmore H, Basavanhally A, et al. Accurate and reproducible invasive breast cancer detection in whole-slide images: A Deep Learning approach for quantifying tumor extent. \u003cem\u003eScientific reports\u003c/em\u003e. 2017;7(1):46450.\u003c/li\u003e\n \u003cli\u003eTolkach Y, Wolgast LM, Damanakis A, et al. Artificial intelligence for tumour tissue detection and histological regression grading in oesophageal adenocarcinomas: a retrospective algorithm development and validation study. \u003cem\u003eThe Lancet Digital Health\u003c/em\u003e. 2023;5(5):e265-e275.\u003c/li\u003e\n \u003cli\u003eZhang Z, Chen P, McGough M, et al. Pathologist-level interpretable whole-slide cancer diagnosis with deep learning. \u003cem\u003eNature Machine Intelligence\u003c/em\u003e. 2019;1(5):236-245.\u003c/li\u003e\n \u003cli\u003eChen C-L, Chen C-C, Yu W-H, et al. An annotation-free whole-slide training approach to pathological classification of lung cancer types using deep learning. \u003cem\u003eNature communications\u003c/em\u003e. 2021;12(1):1193.\u003c/li\u003e\n \u003cli\u003eKorbar B, Olofson AM, Miraflor AP, et al. Deep learning for classification of colorectal polyps on whole-slide images. \u003cem\u003eJournal of pathology informatics\u003c/em\u003e. 2017;8(1):30.\u003c/li\u003e\n \u003cli\u003eLi H, Qin J, Li Z, et al. Systematic review and meta-analysis of deep learning for MSI-H in colorectal cancer whole slide images. \u003cem\u003enpj Digital Medicine\u003c/em\u003e. 2025;8(1):456.\u003c/li\u003e\n \u003cli\u003evan Eekelen L, Spronck J, Looijen-Salamon M, et al. Comparing deep learning and pathologist quantification of cell-level PD-L1 expression in non-small cell lung cancer whole-slide images. \u003cem\u003eScientific reports\u003c/em\u003e. 2024;14(1):7136.\u003c/li\u003e\n \u003cli\u003eJiao Y, Li J, Qian C, Fei S. Deep learning-based tumor microenvironment analysis in colon adenocarcinoma histopathological whole-slide images. \u003cem\u003eComputer Methods and Programs in Biomedicine\u003c/em\u003e. 2021;204:106047.\u003c/li\u003e\n \u003cli\u003eZhang X, Yu X, Liang W, et al. Deep learning‐based accurate diagnosis and quantitative evaluation of microvascular invasion in hepatocellular carcinoma on whole‐slide histopathology images. \u003cem\u003eCancer Medicine\u003c/em\u003e. 2024;13(5):e7104.\u003c/li\u003e\n \u003cli\u003eL\u0026oacute;pez-\u0026Uacute;beda P, Mart\u0026iacute;n-Noguerol T, Aneiros-Fern\u0026aacute;ndez J, Luna A. Natural language processing in pathology: current trends and future insights. \u003cem\u003eThe American Journal of Pathology\u003c/em\u003e. 2022;192(11):1486-1495.\u003c/li\u003e\n \u003cli\u003eLebeau A, Turzynski A. Pathology: diagnostics, reporting and artificial intelligence. \u003cem\u003eSenologie-Zeitschrift f\u0026uuml;r Mammadiagnostik und-therapie\u003c/em\u003e. 2025;22(01):31-42.\u003c/li\u003e\n \u003cli\u003eNaugler C, Church DL. Automation and artificial intelligence in the clinical laboratory. \u003cem\u003eCritical reviews in clinical laboratory sciences\u003c/em\u003e. 2019;56(2):98-110.\u003c/li\u003e\n \u003cli\u003eAcharya DB, Kuppan K, Divya B. Agentic AI: Autonomous intelligence for complex goals\u0026mdash;A comprehensive survey. \u003cem\u003eIEEe Access\u003c/em\u003e. 2025;13:18912-18936.\u003c/li\u003e\n \u003cli\u003eMurugesan S. The rise of agentic AI: implications, concerns, and the path forward. \u003cem\u003eIEEE Intelligent Systems\u003c/em\u003e. 2025;40(2):8-14.\u003c/li\u003e\n \u003cli\u003eKarunanayake N. Next-generation agentic AI for transforming healthcare. \u003cem\u003eInformatics and Health\u003c/em\u003e. 2025;2(2):73-83.\u003c/li\u003e\n \u003cli\u003eShaktah LA, Carrero ZI, Hewitt KJ, et al. Application of artificial intelligence and digital tools in cancer pathology. \u003cem\u003eThe Lancet Digital Health\u003c/em\u003e. 2025;7(10)\u003c/li\u003e\n \u003cli\u003eBellahsen-Harrar Y, Lubrano M, L\u0026eacute;pine C, et al. Exploring the risks of over-reliance on AI in diagnostic pathology. What lessons can be learned to support the training of young pathologists? \u003cem\u003ePLoS One\u003c/em\u003e. 2025;20(8):e0323270.\u003c/li\u003e\n \u003cli\u003eSunday O. Behavioral Impacts of AI Reliance in Diagnostics: Balancing Automation with Skill Retention. \u003cem\u003eEpidemiol Health Data Insights\u003c/em\u003e. 2025;1:ehdi014.\u003c/li\u003e\n \u003cli\u003eMourya A, Jobanputra B, Pai R. AI-powered clinical trials and the imperative for regulatory transparency and accountability. \u003cem\u003eHealth and Technology\u003c/em\u003e. 2024;14(6):1071-1081.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"diagnostic-pathology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"dpat","sideBox":"Learn more about [Diagnostic Pathology](http://diagnosticpathology.biomedcentral.com)","snPcode":"13000","submissionUrl":"https://submission.nature.com/new-submission/13000/3","title":"Diagnostic Pathology","twitterHandle":"@OncoBioMed","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Diagnostic errors, pathology, failure mode and effect analysis, agentic artificial intelligence","lastPublishedDoi":"10.21203/rs.3.rs-9369303/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9369303/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDiagnostic errors in pathology remain an important contributor to patient morbidity and healthcare inefficiency despite increasing laboratory automation and quality management initiatives. Failure Mode and Effects Analysis (FMEA) offers a structured framework for prospective identification and prioritization of workflow risks across the total testing process. Concurrently, artificial intelligence (AI) has demonstrated substantial capability in image analysis, data validation, and workflow optimisation in both anatomical and clinical pathology. We present our findings from a survey among 43 pathologists who identified 266 errors i.e., (183, 69%) within pre-analytical phase, 59 (22%) errors within analytical phase and 22 (9%) errors within post-analytical phase. Then we present the results of the FMEA along with the most suitable AI approaches to mitigate the errors identified. In the end, we present a conceptual framework to integrate AI particularly agentic AI with FMEA-driven quality governance. Agentic AI enables dynamic monitoring, predictive error detection, and autonomous workflow orchestration across pre-analytical, analytical, and post-analytical phases. Integrating agentic AI with FMEA may enable continuous recalibration of risk models, transition pathology quality programs from reactive to predictive paradigms, and support scalable diagnostic safety ecosystems. Implementation considerations including algorithmic transparency, human\u0026ndash;AI interaction, and regulatory governance are discussed. This integrated approach represents a potential next-generation model for improving diagnostic accuracy, operational efficiency, and patient safety across pathology services.\u003c/p\u003e","manuscriptTitle":"Artificial Intelligence and Failure Mode and Effect Analysis Framework to Reduce Diagnostic Errors in Pathology","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-04 06:44:06","doi":"10.21203/rs.3.rs-9369303/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-05-01T13:04:55+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-28T16:14:58+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"69817556891177822499536024907883843859","date":"2026-04-22T09:17:36+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-21T12:05:54+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-15T08:46:43+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-15T08:46:32+00:00","index":"","fulltext":""},{"type":"submitted","content":"Diagnostic Pathology","date":"2026-04-09T13:21:11+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"diagnostic-pathology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"dpat","sideBox":"Learn more about [Diagnostic Pathology](http://diagnosticpathology.biomedcentral.com)","snPcode":"13000","submissionUrl":"https://submission.nature.com/new-submission/13000/3","title":"Diagnostic Pathology","twitterHandle":"@OncoBioMed","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"12f2dd0c-9776-408e-811b-6ac59fdf8567","owner":[],"postedDate":"May 4th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-01T13:04:55+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-18T01:53:34+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-04 06:44:06","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9369303","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9369303","identity":"rs-9369303","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}