A deep learning approach to gold nanoparticle quantification in computed tomography |
| |
| Institution: | 1. Medical Physics Program, Department of Physics and Applied Physics, University of Massachusetts Lowell, Lowell, MA 01854, United States;2. Landauer Medical Physics, 2 Science Road, Glenwood, IL 60425, United States;3. Department of Medical Physics and Radiation Safety, Rhode Island Hospital, Providence, RI 02903, United States;4. Department of Electrical and Computer Engineering, University of Massachusetts Lowell, Lowell, MA 01854, United States;1. Fondazione Policlinico Universitario A. Gemelli IRCCS, Roma, Italy;2. Medical Physics, San Raffaele Scientific Institute, Milano, Italy;3. Prostate Cancer Program, Fondazione IRCCS Istituto Nazionale dei Tumori, Milano, Italy;4. Health Physics Unit, ASST Santi Paolo e Carlo, Milan, Italy;5. Department of Experimental and Clinical Biomedical Sciences “Mario Serio”, University of Florence, Florence, Italy;6. Università di Napoli Federico II, Dipartimento di Fisica “Ettore Pancini”, Napoli, Italy;7. Medical Physics Unit, Azienda Ospedaliera Universitaria Integrata – Verona, Italy;8. Unit of Radiation Research, IEO European Institute of Oncology, IRCCS, Milano, Italy;1. Department of Medical Physics, Institute of Radiation Protection and Dosimetry (IRD), Av. Salvador Allende, 3773, Barra da Tijuca, Rio de Janeiro, RJ CEP 22783-127, Brazil;2. Nuclear Engineering Department (DNC), Federal University of Rio de Janeiro (UFRJ), sala 206, Centro de Tecnologia, Av. Horácio Macedo, 2030, Bloco G, Fundão, Rio de Janeiro, RJ CEP 21941-941, Brazil;3. Department of Cell Biology, University of the State of Rio de Janeiro (UERJ), Pavilhão Haroldo Lisboa da Cunha, LabAngio, Rua São Francisco Xavier, 524, Maracanã, Rio de Janeiro, RJ CEP 20550-900, Brazil;4. CGMI/DRS, Brazilian National Nuclear Energy Comission (CNEN), Rua General Severiano, 90, Botafogo, Rio de Janeiro, RJ CEP 22290-901, Brazil;5. Department of Radiological Sciences (LCR), State University of Rio de Janeiro (UERJ), Pavilhão Haroldo Lisboa da Cunha, Rua São Francisco Xavier, 524, Maracanã, Rio de Janeiro, RJ CEP 20550-900, Brazil;6. Department of General Biology, Federal Fluminense University, Niterói, RJ CEP 21045-900, Brazil;1. Princess Margaret Cancer Centre, University Health Network, Department of Radiation Oncology, University of Toronto, 610 University Avenue, Toronto, Ontario M5G 2M9, Canada;2. Modus QA, London, Ontario N6H 5L6, Canada;3. Cross Cancer Institute, Alberta Health Services, Department of Radiation Oncology, University of Alberta, 11560 University Avenue, Edmonton, AB T6G 1Z2, Canada;1. Instituto de F??sica, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil;2. Instituto de Radioproteção e Dosimetria, IRD/CNEN, Rio de Janeiro, Brazil;3. Oncologia D’Or São Cristóvão, Rede D’Or São Luiz, Rio de Janeiro, Brazil;1. Department of Radiation Oncology, The University of Michigan, Ann Arbor, MI, USA;2. London Health Sciences Centre, Western University, London, ON, Canada;3. Department of Radiation Oncology, UCSF Medical Center at Mission Bay, San Francisco, CA, USA |
| |
| Abstract: | IntroductionDeep learning (DL) is used to classify, detect, and quantify gold nanoparticles (AuNPs) in a human-sized phantom with a clinical MDCT scanner.MethodsAuNPs were imaged at concentrations between 0.0274 and 200 mgAu/mL in a 33 cm phantom. 1 mm-thick CT image slices were acquired at 120 kVp with a CTDIvol of 23.6 mGy. A convolutional neural network (CNN) was trained on 544 images to classify 17 different tissue types and AuNP concentrations. A second set of 544 images was then used for testing.ResultsAuNPs were classified with 95% accuracy at 0.1095 mgAu/mL and 97% accuracy at 0.2189 mgAu/mL. Both these concentrations are lower than what humans can visually perceive (0.3–1.4 mgAu/mL). AuNP concentrations were also classified with 95% accuracy at 150 and 200 mgAu/mL. These high concentrations result in CT numbers that are at or above the 12-bit limit for CT’s dynamic range where extended Hounsfield scales are otherwise required for measuring differences in contrast.ConclusionsWe have shown that DL can be used to detect AuNPs at concentrations lower than what humans can visually perceive and can also quantify very high AuNP concentrations that exceed the typical 12-bit dynamic range of clinical MDCT scanners. This second finding is possible due to inhomogeneous AuNP distributions and characteristic streak artifacts. It may even be possible to extend this approach beyond AuNP imaging in CT for quantifying high density objects without extended Hounsfield scales. |
| |
| Keywords: | Gold Nanoparticles Machine learning Deep learning AuNPs |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
