{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T07:06:58Z","timestamp":1771830418695,"version":"3.50.1"},"reference-count":59,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T00:00:00Z","timestamp":1771804800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Accurate measurements of porosity of cementitious matrices are critical in pre- Q7 dicting mechanical behavior, durability, and transportation processes. Traditional methods based on SEM, such as manual thresholding, a simple binarization method, and end-to-end convolutional neural network (CNN) regressors, are, however, highly affected by image contrast variation, polishing quality, magnification, and operator bias. To address these limitations, the current article develops a hybrid deep-boosting framework for fully automatic porosity estimation directly from raw backscattered-electron SEM images of fly-ash cement mortar. The key novelty of the proposed approach lies in the adaptive stabilization of porosity labels and the hybrid fusion of deep semantic and handcrafted texture features, which together improve robustness to imaging artifacts, boundary ambiguity, and overfitting.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Annotation ground-truth porosity is optimized using an Adaptive Porosity Label Stabilizer (APLS) that successively improves Otsu threshold masks, first using entropy measures and morphological consistency measures to reduce label noise. Multiscale semantic representations are learned on a ResNet-18 backbone, which is trained with SimCLR on SEM data, while local statistical texture is captured using handcrafted gray-level cooccurrence map (GLCM) features. The resulting mismatched set of features is combined with a learnable Hybrid Feature Refinement Block (HFRB) together with a Feature-Interaction Attention (FIA) block, which explicitly characterizes inter-scale relationships among convolutional channels and texture regressors. The latent representation is then condensed and regressed using a weighted ensemble including CatBoost, XGBoost, and LightGBM learners.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results and discussion<\/jats:title>\n                    <jats:p>\n                      The proposed methodology achieves R\n                      <jats:sup>2<\/jats:sup>\n                      = 0.9816, RMSE = 0.0236, and MAE = 0.00875 on a rigorously held out test set, outperforming baseline methods that rely exclusively on CNN features, handcrafted descriptors, or na\u00efve hybrid combinations. The validity, stability, and physical plausibility of the model are ensured through a comprehensive assessment, including ablation studies, domain-shift experiments, uncertainty and stability calibration, and a hybrid explainability framework (Grad-CAM++, SHAP). The architecture does not require any manual segmentation, generalizes across magnifications and imaging conditions, and provides transparent, domain-consistent explanatory visualizations. Overall, the proposed framework represents an important step toward fast, reliable, and scalable SEMbased porosity estimation in cementitious systems.\n                    <\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/frai.2026.1766671","type":"journal-article","created":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T06:30:47Z","timestamp":1771828247000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["A hybrid deep boosting framework with adaptive label stabilization for SEM-based porosity estimation in fly-ash cement mortar"],"prefix":"10.3389","volume":"9","author":[{"family":"Abdullah","sequence":"first","affiliation":[{"name":"Center for Computing Research, Instituto Polit\u00e9cnico Nacional","place":["Mexico City, Mexico"]},{"name":"Department of Computer Sciences, Bahria University","place":["Lahore, Pakistan"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Muhammad Ateeb","family":"Ather","sequence":"additional","affiliation":[{"name":"Center for Computing Research, Instituto Polit\u00e9cnico Nacional","place":["Mexico City, Mexico"]},{"name":"Department of Computer Sciences, Bahria University","place":["Lahore, Pakistan"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jos\u00e9 Luis Oropeza","family":"Rodr\u00edguez","sequence":"additional","affiliation":[{"name":"Center for Computing Research, Instituto Polit\u00e9cnico Nacional","place":["Mexico City, Mexico"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Carlos Guzm\u00e1n","family":"S\u00e1nchez-Mejorada","sequence":"additional","affiliation":[{"name":"Center for Computing Research, Instituto Polit\u00e9cnico Nacional","place":["Mexico City, Mexico"]}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2026,2,23]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"3312","DOI":"10.3390\/ijms24043312","article-title":"Recent developments in polymer nanocomposites for bone regeneration","volume":"24","author":"Abbas","year":"2023","journal-title":"Int. J. Mol. Sci."},{"key":"ref2","doi-asserted-by":"publisher","first-page":"2311","DOI":"10.3390\/buildings15132311","article-title":"A state-of-the-art review on the influence of porosity on the compressive strength of porous concrete for infrastructure applications","volume":"15","author":"Abousnina","year":"2025","journal-title":"Buildings"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"608","DOI":"10.3390\/toxins15100608","article-title":"Ensemble machine learning of gradient boosting (XGBoost, LightGBM, CatBoost) and attention-based CNN-LSTM for harmful algal blooms forecasting","volume":"15","author":"Ahn","year":"2023","journal-title":"Toxins"},{"key":"ref4","first-page":"D021S009R001","article-title":"Machine learning for 3D image recognition to determine porosity and lithology of heterogeneous carbonate rock","volume-title":"SPE reservoir characterisation and simulation conference and exhibition","author":"Al-Farisi","year":"2019"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"12600","DOI":"10.3390\/app132312600","article-title":"Mineral characterization using scanning electron microscopy (SEM): a review of the fundamentals, advancements, and research directions","volume":"13","author":"Ali","year":"2023","journal-title":"Appl. Sci."},{"key":"ref6","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1016\/j.cemconcomp.2016.02.001","article-title":"A combined SEM\u2013calorimetric approach for assessing hydration and porosity development in GGBS concrete","volume":"68","author":"Attari","year":"2016","journal-title":"Cem. Concr. Compos."},{"key":"ref7","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1177\/16878132251347390","article-title":"Machine learning techniques in ultrasonics-based defect detection and material characterization: a comprehensive review","volume":"17","author":"Barashok","year":"2025","journal-title":"Adv. Mech. Eng."},{"key":"ref8","doi-asserted-by":"publisher","first-page":"137","DOI":"10.6028\/jres.107.013","article-title":"The visible cement data set","volume":"107","author":"Bentz","year":"2002","journal-title":"J. Res. Natl. Inst. Stand. Technol."},{"key":"ref9","volume-title":"Improving stability of tree-based models","author":"Bl\u00f8rstad","year":"2023"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1039\/d4ta00251b","article-title":"Leveraging machine learning in porous media","volume":"12","author":"Delpisheh","year":"2024","journal-title":"J. Mater. Chem. A"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"989","DOI":"10.1617\/s11527-006-9194-9","article-title":"Microstructural features of a mortar as seen by computed microtomography","volume":"40","author":"Diamond","year":"2007","journal-title":"Mater. Struct."},{"key":"ref12","doi-asserted-by":"publisher","first-page":"1495","DOI":"10.1007\/s11831-022-09836-2","article-title":"A review of image-based simulation applications in high-value manufacturing","volume":"30","author":"Evans","year":"2023","journal-title":"Arch. Comput. Methods Eng."},{"key":"ref13","doi-asserted-by":"publisher","first-page":"21488","DOI":"10.1038\/s41598-020-78415-x","article-title":"Predicting porosity, permeability, and tortuosity of porous media from images by deep learning","volume":"10","author":"Graczyk","year":"2020","journal-title":"Sci. Rep."},{"key":"ref14","doi-asserted-by":"publisher","first-page":"22164","DOI":"10.1038\/s41598-025-07211-2","article-title":"Comparative analysis of sandstone microtomographic image segmentation using advanced convolutional neural networks with pixelwise and physical accuracy evaluation","volume":"15","author":"Hayatdavoudi","year":"2025","journal-title":"Sci. Rep."},{"key":"ref15","doi-asserted-by":"publisher","first-page":"5985","DOI":"10.1007\/s11661-020-06008-4","article-title":"Overview: computer vision and machine learning for microstructural characterization and analysis","volume":"51","author":"Holm","year":"2020","journal-title":"Metall. Mater. Trans. A"},{"key":"ref16","doi-asserted-by":"crossref","first-page":"190","DOI":"10.1007\/978-3-031-30477-4","volume-title":"Tortuosity and microstructure effects in porous media: classical theories, empirical data, and modern methods","author":"Holzer","year":"2023"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"812","DOI":"10.3390\/ma16020812","article-title":"Traditional vs. automated computer image analysis\u2014a comparative assessment of use for analysis of digital SEM images of high-temperature ceramic material","volume":"16","author":"Jastrz\u0119bska","year":"2023","journal-title":"Materials"},{"key":"ref18","volume-title":"Utilizing literature to characterize materials from images","author":"Jiang","year":"2022"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"1739","DOI":"10.2118\/205347-pa","article-title":"Machine-learning-assisted segmentation of focused ion beam-scanning electron microscopy images with artifacts for improved void-space characterization of tight reservoir rocks","volume":"26","author":"Kazak","year":"2021","journal-title":"SPE J."},{"key":"ref20","doi-asserted-by":"publisher","first-page":"13381","DOI":"10.1007\/s13369-022-06820-7","article-title":"The properties of cement-mortar at different cement strength classes: experimental study and multi-objective modeling","volume":"47","author":"Kazemi","year":"2022","journal-title":"Arab. J. Sci. Eng."},{"key":"ref21","doi-asserted-by":"publisher","first-page":"30205","DOI":"10.1021\/acsomega.3c10131","article-title":"Machine learning assisted prediction of porosity and related properties using digital rock images","volume":"9","author":"Khan","year":"2024","journal-title":"ACS Omega"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"63","DOI":"10.3390\/jcs3030063","article-title":"Constituents phase reconstruction through applied machine learning in nanoindentation mapping data of the mortar surface","volume":"3","author":"Koumoulos","year":"2019","journal-title":"J. Compos. Sci."},{"key":"ref23","doi-asserted-by":"crossref","first-page":"35","DOI":"10.1007\/978-981-99-3393-8_2","article-title":"Image preprocessing, morphology, segmentation, colorimetry","volume-title":"Computer vision metrics: survey, taxonomy, and analysis of computer vision, visual neuroscience, and visual AI","author":"Krig","year":"2025"},{"key":"ref24","doi-asserted-by":"publisher","first-page":"131925","DOI":"10.1016\/j.conbuildmat.2023.131925","article-title":"Hydration and microstructure of cement paste mixed with seawater\u2013an advanced investigation by SEM-EDS method","volume":"392","author":"Li","year":"","journal-title":"Constr. Build. Mater."},{"key":"ref25","doi-asserted-by":"publisher","first-page":"113765","DOI":"10.1016\/j.measurement.2023.113765","article-title":"Quantifying the phase assemblage in the hydrated cement paste by an interpretable and automatic EDS image analysis framework","volume":"223","author":"Li","year":"","journal-title":"Measurement"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"e12425","DOI":"10.1002\/inf2.12425","article-title":"Methods, progresses, and opportunities of materials informatics","volume":"5","author":"Li","year":"2023","journal-title":"InfoMat"},{"key":"ref27","doi-asserted-by":"publisher","first-page":"2664","DOI":"10.3390\/ma15072664","article-title":"Fly ash application as supplementary cementitious material: a review","volume":"15","author":"Li","year":"2022","journal-title":"Materials"},{"key":"ref28","doi-asserted-by":"publisher","first-page":"982","DOI":"10.3390\/coatings15090982","article-title":"Review on durability deterioration and mitigation of concrete structures","volume":"15","author":"Ma","year":"2025","journal-title":"Coatings"},{"key":"ref29","doi-asserted-by":"publisher","first-page":"102698","DOI":"10.1016\/j.mtcomm.2021.102698","article-title":"SEM-EDS and microindentation-driven large-area high-resolution chemomechanical mapping and computational homogenization of cementitious materials","volume":"28","author":"Maragh","year":"2021","journal-title":"Mater Today Commun"},{"key":"ref30","article-title":"SEM supervised image analysis contribution on understanding sulfate attack on blended cement mortars","volume-title":"15th Euroseminar on microscopy applied to building materials, 2015","author":"Meulenyzera","year":"2015"},{"key":"ref31","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s11242-025-02167-3","article-title":"A data-driven approach for efficient prediction of permeability of porous rocks by combining multiscale imaging and machine learning","volume":"152","author":"Nabipour","year":"2025","journal-title":"Transp. Porous Media"},{"key":"ref32","doi-asserted-by":"publisher","first-page":"165","DOI":"10.1016\/j.cemconcomp.2021.104294","article-title":"Artificial neural network application on microstructure\u2013compressive strength relationship of cement mortar","volume":"41","author":"Onal","year":"2010","journal-title":"Adv. Eng. Softw."},{"key":"ref33","doi-asserted-by":"publisher","first-page":"104496","DOI":"10.1016\/j.cemconcomp.2022.104496","article-title":"Image-based microstructure classification of mortar and paste using convolutional neural networks and transfer learning","volume":"129","author":"Qian","year":"2022","journal-title":"Cem. Concr. Compos."},{"key":"ref34","doi-asserted-by":"publisher","first-page":"104819","DOI":"10.1016\/j.cemconcomp.2022.104819","article-title":"Segmentation and analysis of cement particles in cement paste with deep learning","volume":"136","author":"Qian","year":"2023","journal-title":"Cem. Concr. Compos."},{"key":"ref35","doi-asserted-by":"publisher","first-page":"e2020WR029472","DOI":"10.1029\/2020wr029472","article-title":"Review of data science trends and issues in porous media research with a focus on image-based techniques","volume":"57","author":"Rabbani","year":"2021","journal-title":"Water Resour. Res."},{"key":"ref36","doi-asserted-by":"publisher","first-page":"139033","DOI":"10.1016\/j.conbuildmat.2024.139033","article-title":"Cement vs aggregates and textures of aggregates in a mortar: comparative image analysis methods and analytical protocols","volume":"453","author":"Radica","year":"2024","journal-title":"Constr. Build. Mater."},{"key":"ref37","doi-asserted-by":"publisher","first-page":"101077","DOI":"10.1016\/j.rineng.2023.101077","article-title":"Machine learning algorithms in wood ash-cement-Nano TiO2-based mortar subjected to elevated temperatures","volume":"18","author":"Raheem","year":"2023","journal-title":"Results Eng."},{"key":"ref38","doi-asserted-by":"publisher","first-page":"2538978","DOI":"10.1080\/24705314.2025.2538978","article-title":"Experimental investigations on performance efficiency characteristics of ultra-high-performance concrete (UHPC) with enhanced sustainable mineral admixtures","volume":"10","author":"Regalla","year":"2025","journal-title":"J. Struct. Integr. Maint."},{"key":"ref39","doi-asserted-by":"publisher","first-page":"e210290","DOI":"10.1148\/ryai.210290","article-title":"Mitigating bias in radiology machine learning: 1. Data handling","volume":"4","author":"Rouzrokh","year":"2022","journal-title":"Radiol. Artif. Intell."},{"key":"ref40","doi-asserted-by":"publisher","first-page":"106524","DOI":"10.1016\/j.cemconres.2021.106524","article-title":"Quantifying dispersion of nanosilica in hardened cement matrix using a novel SEM-EDS and image analysis-based methodology","volume":"147","author":"Sargam","year":"2021","journal-title":"Cem. Concr. Res."},{"key":"ref41","doi-asserted-by":"publisher","first-page":"11","DOI":"10.1038\/s41526-024-00349-9","article-title":"Unlocking the potential: analyzing 3D microstructure of small-scale cement samples from space using deep learning","volume":"10","author":"Saseendran","year":"2024","journal-title":"NPJ Microgravity"},{"key":"ref42","doi-asserted-by":"crossref","DOI":"10.2514\/6.2023-2025","volume-title":"Micromechanics-based modeling of mechanical properties of Tricalcium silicate processed under microgravity","author":"Saseendran","year":"2023"},{"key":"ref43","doi-asserted-by":"publisher","first-page":"100779","DOI":"10.1016\/j.mtcomm.2019.100779","article-title":"Multiscale pore structure analysis of nano titanium dioxide cement mortar composite","volume":"22","author":"Shafaei","year":"2020","journal-title":"Mater Today Commun"},{"key":"ref44","volume-title":"The use of a UK alum water treatment sludge as a supplementary cementitious material\u2013characteristics, hydration and performance","author":"Shamaki","year":"2021"},{"key":"ref45","doi-asserted-by":"publisher","first-page":"117846","DOI":"10.1016\/j.eswa.2022.117846","article-title":"Segmentation of backscattered electron images of geopolymers using convolutional autoencoder network","volume":"206","author":"Sheiati","year":"2022","journal-title":"Expert Syst. Appl."},{"key":"ref46","doi-asserted-by":"publisher","first-page":"107231","DOI":"10.1016\/j.cemconres.2023.107231","article-title":"Cementitious phase quantification using deep learning","volume":"172","author":"Sheiati","year":"2023","journal-title":"Cem. Concr. Res."},{"key":"ref47","doi-asserted-by":"publisher","first-page":"134281","DOI":"10.1016\/j.conbuildmat.2023.134281","article-title":"Assessment of mechanical properties with machine learning modeling and durability, and microstructural characteristics of a biochar-cement mortar composite","volume":"411","author":"Sobuz","year":"2024","journal-title":"Constr. Build. Mater."},{"key":"ref48","doi-asserted-by":"crossref","first-page":"341","DOI":"10.1002\/9781394191208.ch17","article-title":"Scaling in concrete\u2013causes, prevention, and repair","volume-title":"Industrial scale inhibition: principles, design, and applications","author":"Subash","year":"2024"},{"key":"ref49","volume-title":"Applications of deep convolutional neural networks for energy-based materials characterization on digital images","author":"Tang","year":"2023"},{"key":"ref50","doi-asserted-by":"publisher","first-page":"57","DOI":"10.2118\/231405-pa","article-title":"A segmentation-independent workflow for digital rock property prediction using pore-structure-preserving generative adversarial networks and synthetic data: porosity case study","volume":"31","author":"Temani","year":"2025","journal-title":"SPE J."},{"key":"ref51","doi-asserted-by":"publisher","first-page":"531","DOI":"10.3390\/s25020531","article-title":"A comprehensive survey of deep learning approaches in image processing","volume":"25","author":"Trigka","year":"2025","journal-title":"Sensors"},{"key":"ref52","doi-asserted-by":"crossref","first-page":"104294","DOI":"10.1016\/j.cemconcomp.2021.104294","article-title":"Object status identification of X-ray CT images of microcapsule-based self-healing mortar","volume":"125","author":"Wang","year":"2022","journal-title":"Cem. Concr. Compos."},{"key":"ref53","doi-asserted-by":"publisher","first-page":"133966","DOI":"10.1016\/j.conbuildmat.2023.133966","article-title":"Identification of experienced temperature in mortar and concrete using microstructural image and deep learning","volume":"409","author":"Wang","year":"2023","journal-title":"Constr. Build. Mater."},{"key":"ref54","doi-asserted-by":"publisher","first-page":"2981","DOI":"10.3390\/ma15092981","article-title":"Recent progress in single and combined porosity-evaluation techniques for porous materials","volume":"15","author":"Wang","year":"2022","journal-title":"Materials"},{"key":"ref55","doi-asserted-by":"publisher","first-page":"04025547","DOI":"10.1061\/jmcee7.mteng-20664","article-title":"Exploring the correlation between microstructure and compressive strength in Portland cement-based materials: a deep learning-based comparison of \u03bc CT and SEM-BSE microstructure images","volume":"38","author":"Wu","year":"2026","journal-title":"J. Mater. Civ. Eng."},{"key":"ref56","doi-asserted-by":"publisher","first-page":"M31","DOI":"10.1190\/geo2024-0340.1","article-title":"Advanced machine learning for low-data porosity and permeability prediction in tight sandstones","volume":"90","author":"Wu","year":"2025","journal-title":"Geophysics"},{"key":"ref57","doi-asserted-by":"publisher","first-page":"1324881","DOI":"10.3389\/fpls.2023.1324881","article-title":"Principles, developments, and applications of spatially resolved spectroscopy in agriculture: a review","volume":"14","author":"Xia","year":"2024","journal-title":"Front. Plant Sci."},{"key":"ref58","doi-asserted-by":"publisher","first-page":"213962","DOI":"10.1016\/j.geoen.2025.213962","article-title":"Multi-phase segmentation methods for micro-tomographic images based on deep learning","volume":"252","author":"Yu","year":"2025","journal-title":"Geoenergy Sci. Eng."},{"key":"ref59","doi-asserted-by":"publisher","first-page":"304","DOI":"10.3390\/fractalfract8060304","article-title":"Fractal analysis of cement-based composite microstructure and its application in evaluation of macroscopic performance of cement-based composites: a review","volume":"8","author":"Zhang","year":"2024","journal-title":"Fractal Fract."}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1766671\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T06:30:50Z","timestamp":1771828250000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1766671\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,23]]},"references-count":59,"alternative-id":["10.3389\/frai.2026.1766671"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1766671","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,23]]},"article-number":"1766671"}}