Skin3D: Detection and longitudinal tracking of pigmented skin lesions in 3D total-body textured meshes |
| |
| Institution: | 1. Department of Statistics, Florida State University, Tallahassee, 32304, USA;2. Department of Computer Science, University of North Carolina at Chapel Hill, Chapel Hill, 27599, USA;3. Department of Biostatistics, University of North Carolina at Chapel Hill, Chapel Hill, 27599, USA;4. Department of Radiology, University of North Carolina at Chapel Hill, Chapel Hill, 27599, USA;5. Thurston Arthritis Research Center, University of North Carolina at Chapel Hill, Chapel Hill, 27599, USA;6. Department of Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, 27599, USA;7. Biomedical Research Imaging Center, University of North Carolina at Chapel Hill, Chapel Hill, 27599, USA;1. Department of Computer Science, Technical University of Munich, Garching, Germany;2. GE Healthcare, Munich, Germany;3. IRCCS Fondazione Stella Maris, Pisa, Italy;4. Fondazione Imago7, Pisa, Italy;5. Pisa Division, National Institute for Nuclear Physics (INFN), Pisa, Italy;6. Department of Quantitative Biomedicine, University of Zurich, Zurich, Switzerland;7. Department of Neuroradiology, Klinikum rechts der Isar, Munich, Germany;8. Department of Computer Science, University of Pisa, Pisa, Italy;9. Azienda Ospedaliero-Universitaria Pisana, Pisa Italy;10. Department of Neuroscience, Rehabilitation, Ophtalmology, Genetics, Maternal and Child Sciences (DINOGMI), University of Genova, Genova, Italy;11. AImotion Bavaria, Faculty of Electrical Engineering and Information Technology, Technische Hochschule Ingolstadt, Ingolstadt, Germany;12. Department of Physics, Technical University of Munich, Garching, Germany;1. Google Health;2. Google Research, Brain Team;3. DeepMind;1. Wellcome/EPSRC Centre for Interventional and Surgical Sciences (WEISS), University College London (UCL), Charles Bell House, 43-45 Foley Street, London W1W 7TS, UK;2. Division of Translational Surgical Oncology, National Center for Tumor Diseases (NCT) Dresden, Dresden, 01307, Germany;3. Division of Medical and Biological Informatics, German Cancer Research Center (DKFZ), Heidelberg, Germany;1. Department of Computer Science and Engineering, University of Central Arkansas, 201 Donaghey Ave., Conway, AR 72035, USA;2. Institute for Systems and Robotics, University of Lisbon, Av. Rovisco Pais 1 Torre Norte, Lisbon 1049-001, Portugal;3. Dermatology Service, Memorial Sloan Kettering Cancer Center, 1275 York Ave., New York City, NY 10065, USA;4. Department of Dermatology, Medical University of Vienna, Spitalgasse 23, Vienna 1090, Austria;5. Hospital Clinic of Barcelona, C. de Villarroel 170, Barcelona 08036, Spain;6. Google Health, 3400 Hillview Ave., Palo Alto, CA 94304, USA;1. Department of Radiology & Nuclear Medicine, Erasmus MC, University Medical Center Rotterdam, The Netherlands;2. Department of Radiology, Haaglanden Medical Center, The Hague, The Netherlands;3. Clinical Neurophysiology, MIRA Institute for Biomedical Technology and Technical Medicine, University of Twente, Enschede, The Netherlands;4. Department of Neurology, Rijnstate Hospital, Arnhem, The Netherlands;5. Department of Radiology and Nuclear Medicine, Amsterdam University Medical Centers, location AMC, Amsterdam, The Netherlands;6. Department of Radiology, Leiden UMC, Leiden, The Netherlands;7. Philips Healthcare, Best, The Netherlands;8. Faculty of Applied Sciences, Delft University of Technology, The Netherlands |
| |
| Abstract: | We present an automated approach to detect and longitudinally track skin lesions on 3D total-body skin surface scans. The acquired 3D mesh of the subject is unwrapped to a 2D texture image, where a trained objected detection model, Faster R-CNN, localizes the lesions within the 2D domain. These detected skin lesions are mapped back to the 3D surface of the subject and, for subjects imaged multiple times, we construct a graph-based matching procedure to longitudinally track lesions that considers the anatomical correspondences among pairs of meshes and the geodesic proximity of corresponding lesions and the inter-lesion geodesic distances.We evaluated the proposed approach using 3DBodyTex, a publicly available dataset composed of 3D scans imaging the coloured skin (textured meshes) of 200 human subjects. We manually annotated locations that appeared to the human eye to contain a pigmented skin lesion as well as tracked a subset of lesions occurring on the same subject imaged in different poses. Our results, when compared to three human annotators, suggest that the trained Faster R-CNN detects lesions at a similar performance level as the human annotators. Our lesion tracking algorithm achieves an average matching accuracy of 88% on a set of detected corresponding pairs of prominent lesions of subjects imaged in different poses, and an average longitudinal accuracy of 71% when encompassing additional errors due to lesion detection. As there currently is no other large-scale publicly available dataset of 3D total-body skin lesions, we publicly release over 25,000 3DBodyTex manual annotations, which we hope will further research on total-body skin lesion analysis. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
