Frontiers in Artificial Intelligence最新文献

{"title":"ChatGPT in medical education: a cross-sectional analysis of usage, attitudes, perceptions, and practices among Saudi medical students.","authors":"Ali Qasem Mohammed AlAlwan, Ayoob Lone","doi":"10.3389/frai.2026.1751367","DOIUrl":"https://doi.org/10.3389/frai.2026.1751367","url":null,"abstract":"<p><strong>Background: </strong>Generative artificial intelligence (AI) platforms such as ChatGPT are rapidly reshaping higher education, particularly in medical learning environments. Although several researches have examined students' perceptions and attitudes toward ChatGPT in Saudi Arabia, context-specific evidence focusing on medical students and their practical engagement with ChatGPT remained limited.</p><p><strong>Objectives: </strong>This study aimed to describe patterns of ChatGPT usage, students' perceptions, attitudes, and practices among medical students and to explore their associations with demographic and academic characteristics.</p><p><strong>Methods: </strong>A cross-sectional study was carried out between August to October 2025 involving 328 undergraduate medical students in King Faisal University, Saudi Arabia. Data were obtained using a validated 44-item instrument measuring students' perceptions, attitudes, and practices toward ChatGPT. Descriptive and inferential statistics were performed using Statistical Package for Social Science (SPSS, Version, 27), with statistical significant set at <i>p</i> < 0.05. Parametric assumptions were tested, and multiple regression analysis was conducted to identify independent predictors.</p><p><strong>Results: </strong>An exceptional high proportion of students (97.6%) reported prior use of ChatGPT, indicating widespread familiarity with the tool. In terms of usage pattern, the majority (60.1%) indicated frequent use, followed by 26.8% who used it always, 11.3% sometimes, and only 1.08% who had never used it. Overall perception (<i>M</i> = 3.90 ± 0.69), attitude (<i>M</i> = 3.71 ± 0.62), and practice (<i>M</i> = 3.68 ± 0.73) scores reflected generally positive views. A large majority (81.7%) found ChatGPT helpful or extremely helpful in understanding medical concepts, and 73.8% believed it useful for summarizing research articles. Nonetheless, concerns persisted, with 65.2% expressed apprehension about over-reliance on ChatGPT potentially affecting originality and critical thinking. Multiple regression analyses indicated that gender, age, academic year, family status, family income, type of stay, and GPA were significant predictors of perception, attitude, and practice scores (<i>R</i> <sup>2</sup> = 0.11-0.15, <i>p</i> < 0.05), with academic year, age, GPA, and gender being the most consistent predictors. Younger students (<21 years), early-year students, those with higher GPAs (>4.5), and students from higher-income families exhibited more favorable responses.</p><p><strong>Conclusion: </strong>The findings highlight strong awareness, frequent use, and positive perception toward ChatGPT among medical students at King Faisal University, viewed as an accessible and effective academic aid. Nevertheless, apprehension remain regarding reliability, ethical use, and risk of overdependence. Institutional frameworks incorporating AI literacy, critical appraisal training, and ethical guidelines a","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1751367"},"PeriodicalIF":4.7,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13144081/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147843674","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Correction: The realism of behavioral theory-based vs. non-theory-based AI agents during a simulated infant formula shortage.","authors":"Linda Desens, Brandon Walling, Rhys O'Neill, Vanessa Howard, Mary Giammarino, Denise Scannell, Anya Kemble, Taylor Wilkerson, Nyalok Nhial, Sara Beth Elson, Maureen Leahy, Scott Rosen","doi":"10.3389/frai.2026.1812447","DOIUrl":"https://doi.org/10.3389/frai.2026.1812447","url":null,"abstract":"<p><p>[This corrects the article DOI: 10.3389/frai.2026.1719703.].</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1812447"},"PeriodicalIF":4.7,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13147506/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147843719","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Innovative technologies and workplace collaborations in the energy sector based in the United Arab Emirates.","authors":"Jack Charles Boath, Sayed Abdul Majid Gilani, Tamaralaiyefa Harold Tiemo, Ansarullah Tantry","doi":"10.3389/frai.2026.1798647","DOIUrl":"https://doi.org/10.3389/frai.2026.1798647","url":null,"abstract":"<p><strong>Introduction: </strong>The UAE energy sector is navigating digital transformation mandates such as the UAE AI Strategy 2031 and Net Zero commitments, with technologies like AI, IoT and cloud computing creating new avenues for real-time coordination, data-driven decision-making and cross-functional collaboration. These oppor tunities are tempered by challenges of organisational readiness, cultural iner tia and technological integration. Yet, research on innovative practices in the UAE energy context remains limited. Therefore, this study investigates the role of AI, IoT and cloud computing in shaping workplace collaboration in the UAE energy sector.</p><p><strong>Methods: </strong>An explanatory sequential mixed-methods design was adopted which involved Phase 1 (15 October, 2024-31 January, 2025) interviews with 15 professionals in operations, IT and leadership roles from major energy companies, analysed via thematic analysis. Phase 2 (15 February, 2025-15 May, 2025) distributed a survey to a broader sample, yielding 115 valid responses, which were analysed quan titatively. The study is primarily grounded in the Unified Theory of Acceptance and Use of Technology (UTAUT), with the Technology Acceptance Model (TAM), Resource-Based View (RBV) and Actor-Network Theory (ANT) serving as supporting interpretive lenses.</p><p><strong>Results: </strong>Findings show that AI, IoT, and cloud platforms enhance collaboration, especially in remote coordination and predictive decision sup port, but adoption is hindered by resistance to change, fragmented systems and uneven digital literacy.</p><p><strong>Discussion: </strong>Practical implications include modular rollouts, digital maturity audits and AI onboarding programs. Policy recommendations include national collaboration standards, KPI integration and incentives for joint innova tion projects.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1798647"},"PeriodicalIF":4.7,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13144080/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147843747","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The use of artificial intelligence based modelling techniques in One Health-related infectious disease studies in Sub-Saharan Africa: a review.","authors":"Bruno Enagnon Lokonon, Sèton Calmette Ariane Houetohossou, Bruno Amèdjiko Tchede, Richard B Yapi, Aurélie Cailleau, Daniel T Haydon, Bassirou Bonfoh","doi":"10.3389/frai.2026.1778800","DOIUrl":"https://doi.org/10.3389/frai.2026.1778800","url":null,"abstract":"<p><strong>Background: </strong>Sub-Saharan Africa continues to face a substantial burden of infectious diseases, many of which are zoonotic and shaped by complex interactions across human, animal, and environmental systems. Artificial Intelligence (AI), encompassing machine learning (ML) and deep-learning (DL) techniques, has emerged as a powerful tool for enhancing disease prediction, surveillance, diagnosis, and decision-making within a One Health (OH) framework.</p><p><strong>Method: </strong>This systematic review synthesizes evidence from 62 peer-reviewed studies to assess how AI-based modelling techniques have been applied to infectious disease research across Sub-Saharan Africa.</p><p><strong>Results: </strong>Results show that AI adoption has grown rapidly since 2019, with a pronounced surge in publications between 2021 and 2024. However, research leadership and implementation capacity remain geographically uneven, with South Africa, Ethiopia, Kenya, and Tanzania dominating the landscape. Across studies, AI tools were used primarily for classification and prediction tasks, with ensemble models and deep-learning architectures showing the strongest performance (with median accuracy close to 100% for Convolutional Neural Network model). Malaria (24%), HIV (12%), COVID-19 (12%), and Tuberculosis (6.7%) were the most frequently targeted diseases, while zoonotic and environmentally linked infections were comparatively underrepresented. Most studies relied exclusively on human data, revealing a persistent gap in the integration of animal and environmental components critical to the OH paradigm.</p><p><strong>Conclusion: </strong>Despite promising applications, including image-based parasite detection, IoT-enabled surveillance, ecological risk modelling, and smartphone-assisted diagnostics, AI deployment remains constrained by limited computational infrastructure, inadequate digital connectivity, data-governance weaknesses, and shortages of AI-trained specialists. Conversely, expanding mobile connectivity, cloud-based analytics, and advancements in multilingual AI tools could create new opportunities to strengthen surveillance systems, empower health workers, and improve community engagement.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1778800"},"PeriodicalIF":4.7,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13144102/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147843677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The intelligent neonatal healthcare: a systematic review of machine learning architectures integrating the internet of medical things and blockchain.","authors":"Sarlinraj Madhalaimuthu, R Sujatha","doi":"10.3389/frai.2026.1802559","DOIUrl":"https://doi.org/10.3389/frai.2026.1802559","url":null,"abstract":"<p><strong>Background and motivation: </strong>Neonatal healthcare involves managing extreme physiological vulnerability, rapid disease progression, and time-critical decision-making within Neonatal Intensive Care Units (NICUs). Recent developments in blockchain, the Internet of Medical Things (IoMT), and Machine Learning (ML) have provided new opportunities of enhanced health-data governance, non-stop physiological tracking, and prevention of risks. Although this has been achieved, the past studies have primarily evaluated these technologies separately or with regard to adult health care conditions with little consideration of their combined relevance to neonatal care.</p><p><strong>Methods: </strong>In this study, a PRISMA-guided systematic review was conducted to examine intelligent neonatal healthcare systems that integrate ML, IoMT, and blockchain technologies. A systematic search of major scientific databases identified 122 records, of which 76 studies satisfied predefined inclusion criteria and were included for qualitative synthesis. The selected studies were discussed according to clinical areas of application, system architecture, evaluation practices, and implementation limitations peculiar to neonatal contexts.</p><p><strong>Synthesis of current evidence and identified research gaps: </strong>The review indicates that ML-based approaches are the most mature, particularly for early disease detection, mortality risk prediction, and clinical decision support. Continuous and remote physiological monitoring is the primary use of IoMT-based systems, but blockchain-based solutions are still mainly conceptual or prototype-based systems with primary concerns on data integrity, access control and trust. There are still fewer fully integrated ML-IoMT-Blockchain systems specifically for neonates, and there are persistent problems with interoperability, scalability, clinical validation, and AI lifecycle governance. Taking everything considered, this review combines disparate information, identifies important research gaps, and describes future research routes in line with Sustainable Development Goal 3 toward secure, trustworthy, and therapeutically useful intelligent infant healthcare systems.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1802559"},"PeriodicalIF":4.7,"publicationDate":"2026-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13139162/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147843691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Brief report: Artificial intelligence meets small cell lung cancer-integrating clinicopathological and wholeslide image data for prognostic prediction in SCLC.","authors":"Pedro Rocha, Joan Gibert, Silvía Menendez, Raúl Del Rey-Vergara, Albert Iñañez, Laura Masfarré, Nil Navarro, Alejandro Ríos-Hoyo, Sandra Perez, Álvaro Taus, Mario Giner, Ana Rovira, Luis León-Mateos, Dolores Isla, Luis Paz-Ares, Jon Zugazagoitia, Cristina Martí Blanco, Rosario García-Campelo, Alberto Moreno-Vega, Ángel Callejo, Federico Rojo, Ignacio Sanchéz, Edurne Arriola","doi":"10.3389/frai.2026.1766576","DOIUrl":"https://doi.org/10.3389/frai.2026.1766576","url":null,"abstract":"<p><strong>Introduction: </strong>Small-cell lung cancer (SCLC) represents a unique clinical challenge characterized by its aggressive nature, poor prognosis, and limited therapeutic options. Upfront prediction of survival outcomes in this disease could impact patient care by refining risk stratification and thus, personalizing treatment strategies. Here, we investigate the utility of a deep learning (DL) model using digital pathology to predict outcomes of patients diagnosed with SCLC.</p><p><strong>Methods: </strong>We built a random forest (RF) model using clinical data and a DL based model using whole-slide image (WSI) as inputs from a total of 307 patients diagnosed with SCLC, including a training set of 263 patients, and a validation set comprising 44 patients who participated in the CANTABRICO phase IIIB clinical trial. Model performance was assessed using the area under the receiver operating characteristic curve (AUC) with 5-fold crossvalidation to minimize bias and variance of the performance. We report the mean and 95% confidence interval of the AUC values across the folds.</p><p><strong>Results: </strong>In the training set, the RF model achieved an AUC of 0.728 (95% CI: 0.662-0.792) for long-term overall survival (LT_OS) prediction, while the combined RF and DL model achieved an AUC of 0.744 (95% CI: 0.680-0.807). For long-term progression-free survival (LT_PFS) prediction, the RF model achieved an AUC of 0.689 (95% CI: 0.625-0.753), whereas the combined model achieved an AUC of 0.704 (95% CI: 0.640-0.767). Application of the combined RF and DL model to the validation cohort yielded an AUC for LT_OS of 0.604 (95% CI: 0.582-0.626) and an AUC for LT_PFS 0.690 (95% CI: 0.643-0.738), indicating potential clinical applicability.</p><p><strong>Conclusion: </strong>Our results showcase the feasibility of integrating clinicopathological data with WSI through a deep learning model to predict outcomes in patients with SCLC. This approach holds promise in helping physicians to personalize treatment strategies that better suit individual patient needs.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1766576"},"PeriodicalIF":4.7,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13136249/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147843684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Commentary: Artificial intelligence and precision medicine: a pilot study predicting optimal ceftaroline dosage for pediatric patients.","authors":"Hassan Nawaz Tahir, Ehtisham Haider, Shahnila Javed, Mursala Tahir, Yousaf Ali","doi":"10.3389/frai.2026.1808575","DOIUrl":"https://doi.org/10.3389/frai.2026.1808575","url":null,"abstract":"","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1808575"},"PeriodicalIF":4.7,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13136143/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147843739","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Urban tourist volume forecasting using internet search trends and deep learning methods.","authors":"Chenglin Song, Zhiming Wang","doi":"10.3389/frai.2026.1767465","DOIUrl":"https://doi.org/10.3389/frai.2026.1767465","url":null,"abstract":"<p><p>Accurate forecasting of tourist arrivals in major urban destinations is critical for optimizing tourism resource allocation and formulating data-driven marketing strategies. To address this need, this study presents a novel prediction framework that integrates deep learning methodologies with online search behavior data. Specifically, we propose the DTN (Dynamic Tourism Network) model, which combines Disentangled Shape and Time series Normalization (Dish-TS) with Temporal Convolutional Networks (TCN), and utilizes Baidu Index data as a key indicator of online search trends to predict tourist arrivals in Sanya, China. Empirical validation across multiple evaluation metrics demonstrates that the DTN model consistently surpasses conventional deep learning approaches, achieving statistically significant improvements in predictive accuracy for tourist volume estimation. This advancement provides a robust analytical foundation for real‑time tourism demand forecasting in destination management systems. Notably, the proposed method has been evaluated only on a popular urban tourist destination with pronounced seasonality and available Baidu Index data; its applicability to other destination types or regions where different search engines dominate therefore requires further validation.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1767465"},"PeriodicalIF":4.7,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13136134/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147843703","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Experience using artificial intelligence in the digital transformation of education: benefits and challenges.","authors":"Arturs Medveckis, Tamara Pigozne, Rita Birzina, Ivita Pelnena","doi":"10.3389/frai.2026.1756665","DOIUrl":"https://doi.org/10.3389/frai.2026.1756665","url":null,"abstract":"<p><strong>Aim: </strong>The aim of the research is to analyse teachers and adult educators' experiences of the application of artificial intelligence in the digital transformation of education.</p><p><strong>Methods: </strong>In the study the quantitative data collection method-a questionnaire-\"International Survey on Artificial Intelligence in Higher Education, Training and Adult Learning\" and data processing methods for secondary data collection-descriptive statistics (Mean, Median, Mode, Standard Deviation), and the Mann-Whitney test to determine the statistical significance of differences between two independent target groups (teachers and adult educators) have been applied.</p><p><strong>Results: </strong>The research sample consisted of representatives of educational institutions of Latvia-34 teachers and 83 adult educators. The descriptive statistics and Mann-Whitney test results show that both teachers and adult educators similarly assess the impact of AI application on performance and work efficiency and productivity, decision-making, problem-solving, awareness formation and interdisciplinary concept application skills, mental health, learning outcomes and challenges related to AI use (<i>p</i> ≥ 0.05). Teachers, compared to adult educators, have a higher opinion of the application of artificial intelligence for work purposes. Adult educators have a higher opinion of the impact of AI on the development of learners' awareness formation skills, learners' employment and their work performance, and physical and social health (<i>p</i> ≤ 0.05).</p><p><strong>Conclusion: </strong>AI is a new global reality that opens up new paths for knowledge acquisition, whereas the social environment is also facing new challenges. AI tools can be used by a wide range of users who have prior knowledge and skills in constantly changing IT application. Despite the inertia of the education system and the length of bureaucratized decision-making, proactive action is needed that would balance the technological development with the acquisition of new knowledge and skills based on high moral standards at all levels of education, involving high-tech implementers and cooperating with educational staff, scientists and other social partners. A descriptive cross-sectional study design has been chosen for the study and the instrument applied is the survey \"International Survey on Artificial Intelligence in Higher Education, Training and Adult Learning\" developed by the Singapore Institute of Adult Education within the framework of the 3rd network \"Professionalization of Adult Educators in ASEM Countries\" of the Asia-Europe Lifelong Learning and Education Research (ASEM LLL Hub).</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1756665"},"PeriodicalIF":4.7,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13136121/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147843686","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A novel diffuse liver nodule detector via integrating semantic edge features and probabilistic uncertainty modeling.","authors":"Lei Tian, Xiang Liu, Yunyu Shi, Yu Ji, Shuohong Wang","doi":"10.3389/frai.2026.1801342","DOIUrl":"https://doi.org/10.3389/frai.2026.1801342","url":null,"abstract":"<p><strong>Introduction: </strong>Ultrasound image segmentation of diffuse liver fibrosis nodules confronts three critical challenges: boundary ambiguity caused by gradual tissue transitions, texture heterogeneity arising from fibrotic variations, and inadequate uncertainty quantification, which manifests as overconfident misclassifications at fibrotic nodules.</p><p><strong>Methods: </strong>To address these challenges, this paper proposes the Edge-Semantics Probabilistic Dirichlet Network (ESPD-Net), which integrates Dirichlet evidential theory into diffuse lesion segmentation for nodule detection. ESPD-Net employs three synergistic components: (1) The Semantic-Probabilistic Dual Path Fusion (SPDF) bottleneck constructs parallel semantic and probabilistic pathways to capture local morphological and global distribution features. (2) The Dirichlet Evidential Guided Decoder (DEGD) reformulates segmentation as second-order probabilistic modeling under evidential theory, guiding the adaptive feature decoding process by outputting calibrated uncertainty distributions. (3) Guided by these localized high-uncertainty regions, the Dirichlet Boundary Aware Refinement (DBAR) module executes targeted geometric corrections to precisely align ambiguous lesion boundaries.</p><p><strong>Results: </strong>Evaluations on murine and clinical datasets demonstrate that ESPD-Net significantly outperforms state-of-the-art methods. Specifically, it achieves a Dice of 0.855 (+0.043) and an IoU of 0.747 (+0.057). Furthermore, the model effectively minimizes calibration and boundary errors, reducing ECE to 3.85% and HD95 to 3.25.</p><p><strong>Discussion: </strong>These findings demonstrate that the proposed ESPD-Net effectively addresses the challenges of diffuse lesion segmentation, thereby objectively confirming its clinical potential for computer-aided diagnosis.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1801342"},"PeriodicalIF":4.7,"publicationDate":"2026-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13136238/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147843672","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}