TY - GEN
T1 - A Distance-Based Loss for Smooth and Continuous Skin Layer Segmentation in Optoacoustic Images
AU - Gerl, Stefan
AU - Paetzold, Johannes C.
AU - He, Hailong
AU - Ezhov, Ivan
AU - Shit, Suprosanna
AU - Kofler, Florian
AU - Bayat, Amirhossein
AU - Tetteh, Giles
AU - Ntziachristos, Vasilis
AU - Menze, Bjoern
N1 - Publisher Copyright:
© 2020, Springer Nature Switzerland AG.
PY - 2020
Y1 - 2020
N2 - Raster-scan optoacoustic mesoscopy (RSOM) is a powerful, non-invasive optical imaging technique for functional, anatomical, and molecular skin and tissue analysis. However, both the manual and the automated analysis of such images are challenging, because the RSOM images have very low contrast, poor signal to noise ratio, and systematic overlaps between the absorption spectra of melanin and hemoglobin. Nonetheless, the segmentation of the epidermis layer is a crucial step for many downstream medical and diagnostic tasks, such as vessel segmentation or monitoring of cancer progression. We propose a novel, shape-specific loss function that overcomes discontinuous segmentations and achieves smooth segmentation surfaces while preserving the same volumetric Dice and IoU. Further, we validate our epidermis segmentation through the sensitivity of vessel segmentation. We found a 20% improvement in Dice for vessel segmentation tasks when the epidermis mask is provided as additional information to the vessel segmentation network.
AB - Raster-scan optoacoustic mesoscopy (RSOM) is a powerful, non-invasive optical imaging technique for functional, anatomical, and molecular skin and tissue analysis. However, both the manual and the automated analysis of such images are challenging, because the RSOM images have very low contrast, poor signal to noise ratio, and systematic overlaps between the absorption spectra of melanin and hemoglobin. Nonetheless, the segmentation of the epidermis layer is a crucial step for many downstream medical and diagnostic tasks, such as vessel segmentation or monitoring of cancer progression. We propose a novel, shape-specific loss function that overcomes discontinuous segmentations and achieves smooth segmentation surfaces while preserving the same volumetric Dice and IoU. Further, we validate our epidermis segmentation through the sensitivity of vessel segmentation. We found a 20% improvement in Dice for vessel segmentation tasks when the epidermis mask is provided as additional information to the vessel segmentation network.
UR - http://www.scopus.com/inward/record.url?scp=85092792329&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-59725-2_30
DO - 10.1007/978-3-030-59725-2_30
M3 - Conference contribution
AN - SCOPUS:85092792329
SN - 9783030597245
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 309
EP - 319
BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2020 - 23rd International Conference, Proceedings
A2 - Martel, Anne L.
A2 - Abolmaesumi, Purang
A2 - Stoyanov, Danail
A2 - Mateus, Diana
A2 - Zuluaga, Maria A.
A2 - Zhou, S. Kevin
A2 - Racoceanu, Daniel
A2 - Joskowicz, Leo
PB - Springer Science and Business Media Deutschland GmbH
T2 - 23rd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2020
Y2 - 4 October 2020 through 8 October 2020
ER -