{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,22]],"date-time":"2026-01-22T02:23:21Z","timestamp":1769048601041,"version":"3.49.0"},"reference-count":52,"publisher":"MDPI AG","issue":"20","license":[{"start":{"date-parts":[[2023,10,14]],"date-time":"2023-10-14T00:00:00Z","timestamp":1697241600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"US Department of Energy, Office of Nuclear Energy, Nuclear Energy Enabling Technology (NEET) Advanced Methods of Manufacturing (AMM) program","award":["DE-AC02-06CH11357"],"award-info":[{"award-number":["DE-AC02-06CH11357"]}]},{"DOI":"10.13039\/100006147","name":"National Nuclear Security Administration (NNSA) office of Defense Nuclear Nonproliferation Research and Development","doi-asserted-by":"publisher","award":["DE-AC02-06CH11357"],"award-info":[{"award-number":["DE-AC02-06CH11357"]}],"id":[{"id":"10.13039\/100006147","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>One of the key challenges in laser powder bed fusion (LPBF) additive manufacturing of metals is the appearance of microscopic pores in 3D-printed metallic structures. Quality control in LPBF can be accomplished with non-destructive imaging of the actual 3D-printed structures. Thermal tomography (TT) is a promising non-contact, non-destructive imaging method, which allows for the visualization of subsurface defects in arbitrary-sized metallic structures. However, because imaging is based on heat diffusion, TT images suffer from blurring, which increases with depth. We have been investigating the enhancement of TT imaging capability using machine learning. In this work, we introduce a novel multi-task learning (MTL) approach, which simultaneously performs the classification of synthetic TT images, and segmentation of experimental scanning electron microscopy (SEM) images. Synthetic TT images are obtained from computer simulations of metallic structures with subsurface elliptical-shaped defects, while experimental SEM images are obtained from imaging of LPBF-printed stainless-steel coupons. MTL network is implemented as a shared U-net encoder between the classification and the segmentation tasks. Results of this study show that the MTL network performs better in both the classification of synthetic TT images and the segmentation of SEM images tasks, as compared to the conventional approach when the individual tasks are performed independently of each other.<\/jats:p>","DOI":"10.3390\/s23208462","type":"journal-article","created":{"date-parts":[[2023,10,14]],"date-time":"2023-10-14T14:59:59Z","timestamp":1697295599000},"page":"8462","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":13,"title":["Multi-Task Learning of Scanning Electron Microscopy and Synthetic Thermal Tomography Images for Detection of Defects in Additively Manufactured Metals"],"prefix":"10.3390","volume":"23","author":[{"ORCID":"https:\/\/orcid.org\/0009-0001-3534-8774","authenticated-orcid":false,"given":"Sarah","family":"Scott","sequence":"first","affiliation":[{"name":"Nuclear Science and Engineering Division, Argonne National Laboratory, Lemont, IL 60439, USA"},{"name":"Department of Civil and Environmental Engineering, Duke University, Durham, NC 27708, USA"}]},{"given":"Wei-Ying","family":"Chen","sequence":"additional","affiliation":[{"name":"Nuclear Science and Engineering Division, Argonne National Laboratory, Lemont, IL 60439, USA"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8891-9323","authenticated-orcid":false,"given":"Alexander","family":"Heifetz","sequence":"additional","affiliation":[{"name":"Nuclear Science and Engineering Division, Argonne National Laboratory, Lemont, IL 60439, USA"}]}],"member":"1968","published-online":{"date-parts":[[2023,10,14]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"2834","DOI":"10.1007\/s11837-019-03607-4","article-title":"Advanced Manufacturing for Nuclear Energy","volume":"71","author":"Lou","year":"2019","journal-title":"JOM"},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"152846","DOI":"10.1016\/j.jnucmat.2021.152846","article-title":"Qualification Pathways for Additively Manufactured Components for Nuclear Applications","volume":"548","author":"Hensley","year":"2021","journal-title":"J. Nucl. Mater."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"116041","DOI":"10.1016\/j.apenergy.2020.116041","article-title":"Additive Manufacturing for Energy: A Review","volume":"282","author":"Sun","year":"2021","journal-title":"Appl. Energy"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"36","DOI":"10.1016\/j.actamat.2016.02.014","article-title":"Laser Powder-Bed Fusion Additive Manufacturing: Physics of Complex Melt Flow and Formation Mechanisms of Pores, Spatter and Denudation Zones","volume":"108","author":"Khairallah","year":"2016","journal-title":"Acta Mater."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"4808","DOI":"10.1007\/s11665-021-05919-6","article-title":"Defects in Metal Additive Manufacturing Processes","volume":"30","author":"Brennan","year":"2021","journal-title":"J. Mater. Eng. Perform."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"7308","DOI":"10.1038\/s41598-017-06504-5","article-title":"The Influence of Porosity on Fatigue Crack Initiation in Addi-tively Manufactured Titanium Components","volume":"7","author":"Withers","year":"2017","journal-title":"Sci. Rep."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"465","DOI":"10.1007\/s11837-019-03910-0","article-title":"ICME Approach to Determining Critical Pore Size of IN718 Produced by Selective Laser Melting","volume":"72","author":"Sangid","year":"2019","journal-title":"JOM"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"9927","DOI":"10.1364\/OE.416659","article-title":"In-situ Digital Image Correlation and Thermal Monitoring in Directed Energy Deposition Additive Manufacturing","volume":"29","author":"Haley","year":"2021","journal-title":"Opt. Express"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"107338","DOI":"10.1016\/j.optlastec.2021.107338","article-title":"On the Application of Machine Learning for Defect Detection in L-PBF Additive Manufacturing","volume":"143","author":"Mohammadi","year":"2021","journal-title":"Opt. Laser Technol."},{"key":"ref_10","first-page":"273","article-title":"A Multi-Scale Convolutional Neural Network for Autonomous Anomaly Detection and Classification in a Laser Powder Bed Fusion Additive Manufacturing Process","volume":"24","author":"Scime","year":"2018","journal-title":"Addit. Manuf."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"375","DOI":"10.1007\/s10845-018-1451-6","article-title":"A Deep Neural Network for Classification of MeltPool Images in Metal Additive Manufacturing","volume":"31","author":"Kwon","year":"2020","journal-title":"J. Intell. Manuf."},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Cui, W., Zhang, Y., Zhang, X., Li, L., and Liou, F. (2020). Metal Additive Manufacturing Parts Inspection Using Convolutional Neural Network. Appl. Sci., 10.","DOI":"10.3390\/app10020545"},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"91","DOI":"10.1038\/s41524-023-01032-5","article-title":"Enabling rapid X-ray CT Characterisation for Additive manufacturing using CAD Models and Deep Learning-Based Reconstruction","volume":"9","author":"Ziabari","year":"2023","journal-title":"Npj Comput. Mater."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"022002","DOI":"10.1088\/2631-7990\/abe0d0","article-title":"Defect Inspection Technologies for Additive Manufacturing","volume":"3","author":"Chen","year":"2021","journal-title":"Int. J. Extrem. Manuf."},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Zheng, Z., Peng, L., and Wang, D. (2021). Defect Analysis of 316 L Stainless Steel Prepared by LPBF Additive Manufacturing Processes. Coatings, 11.","DOI":"10.3390\/coatings11121562"},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"21","DOI":"10.1017\/S1431927618015635","article-title":"High Throughput Quantitative Metallography for Complex Microstructures Using Deep Learning: A Case Study in Ultrahigh Carbon Steel","volume":"25","author":"DeCost","year":"2019","journal-title":"Microsc. Microanal."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"2128","DOI":"10.1038\/s41598-018-20037-5","article-title":"Advanced Steel Microstructural Classification by Deep Learning Methods","volume":"8","author":"Azimi","year":"2018","journal-title":"Sci. Rep."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"1559","DOI":"10.1093\/micmic\/ozad067.802","article-title":"Deep Learning for Automated Quantification of Irradiation Defects in TEM Data: Relating Pixel-Level Errors to Defect Properties","volume":"29","author":"Sainju","year":"2023","journal-title":"Microsc. Microanal."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"301","DOI":"10.1080\/17452759.2017.1357319","article-title":"Application of non-destructive testing techniques for post-process control of additively manufactured parts","volume":"12","author":"Lu","year":"2017","journal-title":"Virtual Phys. Prototyp."},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Hassani, S., and Dackermann, U.A. (2023). Systematic Review of Advanced Sensor Technologies for Non-Destructive Testing and Structural Health Monitoring. Sensors, 23.","DOI":"10.3390\/s23042204"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"9094","DOI":"10.1109\/ACCESS.2022.3141654","article-title":"Compression of Pulsed Infrared Thermography Data with Unsupervised Learning for Nondestructive Evaluation of Additively Manufactured Metals","volume":"10","author":"Zhang","year":"2021","journal-title":"IEEE Access"},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"4682","DOI":"10.1007\/s11837-020-04408-w","article-title":"Quality Control of Additively Manufactured Metallic Structures with Machine Learning of Thermography Images","volume":"72","author":"Zhang","year":"2020","journal-title":"JOM"},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Moradi, M., Ghorbani, R., Sfarra, S., Tax, D.M.J., and Zarouchas, D. (2022). A Spatiotemporal Deep Neural Network Useful for Defect Identification and Reconstruction of Artworks Using Infrared Thermography. Sensors, 22.","DOI":"10.3390\/s22239361"},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Venegas, P., Ivorra, E., Ortega, M., and S\u00e1ez de Oc\u00e1riz, I. (2022). Towards the Automation of Infrared Thermography Inspections for Industrial Maintenance Applications. Sensors, 22.","DOI":"10.3390\/s22020613"},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Alhammad, M., Avdelidis, N.P., Ibarra-Castanedo, C., Torbali, M.E., Genest, M., Zhang, H., Zolotas, A., and Maldgue, X.P.V. (2022). Automated Impact Damage Detection Technique for Composites Based on Thermographic Image Processing and Machine Learning Classification. Sensors, 22.","DOI":"10.3390\/s22239031"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Fang, Q., Ibarra-Castanedo, C., Garrido, I., Duan, Y., and Maldague, X. (2023). Automatic Detection and Identification of Defects by Deep Learning Algorithms from Pulsed Thermography Data. Sensors, 23.","DOI":"10.3390\/s23094444"},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Szymanik, B. (2022). An Evaluation of 3D-Printed Materials\u2019 Structural Properties Using Active Infrared Thermography and Deep Neural Networks Trained on the Numerical Data. Materials, 15.","DOI":"10.3390\/ma15103727"},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Szymanik, B., Psuj, G., Hashemi, M., and Lopato, P. (2021). Detection and Identification of Defects in 3D-Printed Dielectric Structures via Thermographic Inspection and Deep Neural Networks. Materials, 14.","DOI":"10.3390\/ma14154168"},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"105318","DOI":"10.1063\/5.0016222","article-title":"Thermal Tomography 3D Imaging of Additively Manufactured Metallic Structures","volume":"10","author":"Heifetz","year":"2020","journal-title":"AIP Adv."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"4244","DOI":"10.1007\/s11837-020-04428-6","article-title":"Detection of Defects in Additively Manufactured Stainless Steel 316L with Compact Infrared Camera and Machine Learning Algorithms","volume":"72","author":"Zhang","year":"2020","journal-title":"JOM"},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"244901","DOI":"10.1063\/5.0089072","article-title":"Classification of Computed Thermal Tomography Images with Deep Learning Convolutional Neural Network","volume":"131","author":"Ankel","year":"2022","journal-title":"J. Appl. Phys."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"41","DOI":"10.1023\/A:1007379606734","article-title":"Multitask Learning","volume":"28","author":"Caruana","year":"1997","journal-title":"Mach. Learn."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"1","DOI":"10.2478\/v10053-008-0082-4","article-title":"Associations between Music Education, Intelligence, and Spelling Ability in Ele-mentary School","volume":"7","author":"Hille","year":"2011","journal-title":"Adv. Cogn. Psychol."},{"key":"ref_34","first-page":"475","article-title":"Musical Training as an Alternative and Effective Method for Neuro-Education and Neuro-Rehabilitation","volume":"6","author":"Duarte","year":"2015","journal-title":"Front. Psychol."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"645","DOI":"10.3389\/fnhum.2016.00645","article-title":"Motor-Enriched Learning Activities Can Improve Mathematical Performance in Preadolescent Children","volume":"10","author":"Beck","year":"2016","journal-title":"Front. Hum. Neurosci."},{"key":"ref_36","doi-asserted-by":"crossref","unstructured":"Ronneberger, O., Fischer, P., and Brox, T. (2015). U-Net: Convolutional Networks for Biomedical Image Segmentation. arXiv.","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1177\/01617346221075769","article-title":"A Multi-Task Learning Framework for Automated Segmentation and Classification of Breast Tumors from Ultrasound Images","volume":"44","author":"Chowdary","year":"2022","journal-title":"Ultrason. Imaging"},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"690244","DOI":"10.3389\/fonc.2021.690244","article-title":"A Deep Multi-Task Learning Framework for Brain Tumor Segmentation","volume":"11","author":"Huang","year":"2021","journal-title":"Front. Oncol."},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"157","DOI":"10.1016\/j.cirp.2018.04.119","article-title":"Machine Learning in Tolerancing for Additive Manufacturing","volume":"67","author":"Zhu","year":"2018","journal-title":"CIRP Ann."},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"607","DOI":"10.1016\/j.jmsy.2021.04.012","article-title":"Hybrid Multi-Task Learning-Based Response Surface Modeling in Manufacturing","volume":"59","author":"Yang","year":"2021","journal-title":"J. Manuf. Syst."},{"key":"ref_41","first-page":"4034592","article-title":"Improving Machined Surface Shape Prediction by Integrating Multi-Task Learning With Cutting Force Variation Modeling","volume":"139","author":"Shao","year":"2016","journal-title":"J. Manuf. Sci. Eng."},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"107041","DOI":"10.1016\/j.ultras.2023.107041","article-title":"A Review of Synthetic and Augmented Training Data for Machine Learning in Ultrasonic Nondestructive Evaluation","volume":"134","author":"Sebastian","year":"2023","journal-title":"Ultrasonics"},{"key":"ref_43","doi-asserted-by":"crossref","unstructured":"Liu, K., Wang, F., He, Y., Liu, Y., Yang, J., and Yao, Y. (2023). Data-Augmented Manifold Learning Thermography for Defect Detection and Evaluation of Polymer Composites. Polymers, 15.","DOI":"10.3390\/polym15010173"},{"key":"ref_44","doi-asserted-by":"crossref","first-page":"045401","DOI":"10.1088\/1361-6501\/abc63f","article-title":"A Thermographic Data Augmentation and Signal Separation Method for Defect Detection","volume":"32","author":"Liu","year":"2021","journal-title":"Meas. Sci. Technol."},{"key":"ref_45","doi-asserted-by":"crossref","first-page":"064006","DOI":"10.1088\/1361-6501\/ac5280","article-title":"Nondestructive Detection and Analysis Based on Data Enhanced Thermography","volume":"33","author":"Li","year":"2022","journal-title":"Meas. Sci. Technol."},{"key":"ref_46","doi-asserted-by":"crossref","unstructured":"Wei, Z., Osman, A., Valeske, B., and Maldague, X. (2023). Pulsed Thermography Dataset for Training Deep Learning Models. Appl. Sci., 13.","DOI":"10.20944\/preprints202301.0483.v1"},{"key":"ref_47","doi-asserted-by":"crossref","unstructured":"Fang, Q., Ibarra-Castanedo, C., and Maldague, X. (2021). Automatic Defects Segmentation and Identification by Deep Learning Algorithm with Pulsed Thermography: Synthetic and Experimental Data. Big Data Cogn. Comput., 5.","DOI":"10.3390\/bdcc5010009"},{"key":"ref_48","doi-asserted-by":"crossref","unstructured":"Long, J., Shelhamer, E., and Darrell, T. (2015, January 7\u201312). Fully Convolutional Networks for Semantic Segmentation. Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, MA, USA.","DOI":"10.1109\/CVPR.2015.7298965"},{"key":"ref_49","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1115\/1.4053078","article-title":"Segmentation of Additive Manufacturing Defects Using U-NET","volume":"22","author":"Wong","year":"2021","journal-title":"J. Comput. Inf. Sci. Eng."},{"key":"ref_50","unstructured":"(2022, November 14). Open Source Data Labeling. Available online: https:\/\/labelstud.io\/."},{"key":"ref_51","unstructured":"Kingma, D.P., and Ba, J. (2017). Adam: A Method for Stochastic Optimization. arXiv."},{"key":"ref_52","doi-asserted-by":"crossref","unstructured":"Yang, L., Zhuo, W., Qi, L., Shi, Y., and Gao, Y. (2022). ST++: Make Self-Training Work Better for Semi-Supervised Semantic Segmentation. arXiv.","DOI":"10.1109\/CVPR52688.2022.00423"}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/23\/20\/8462\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T21:06:50Z","timestamp":1760130410000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/23\/20\/8462"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,10,14]]},"references-count":52,"journal-issue":{"issue":"20","published-online":{"date-parts":[[2023,10]]}},"alternative-id":["s23208462"],"URL":"https:\/\/doi.org\/10.3390\/s23208462","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,10,14]]}}}