Display options
Cite
Guo X, Wang F, Teodoro G, et al. LIVER STEATOSIS SEGMENTATION WITH DEEP LEARNING METHODS. Proc IEEE Int Symp Biomed Imaging. 2019;2019:24-27doi: 10.1109/isbi.2019.8759600.
Guo, X., Wang, F., Teodoro, G., Farris, A. B., & Kong, J. (2019). LIVER STEATOSIS SEGMENTATION WITH DEEP LEARNING METHODS. Proceedings. IEEE International Symposium on Biomedical Imaging, 201924-27. https://doi.org/10.1109/isbi.2019.8759600
Guo, Xiaoyuan, et al. "LIVER STEATOSIS SEGMENTATION WITH DEEP LEARNING METHODS." Proceedings. IEEE International Symposium on Biomedical Imaging vol. 2019 (2019): 24-27. doi: https://doi.org/10.1109/isbi.2019.8759600
Guo X, Wang F, Teodoro G, Farris AB, Kong J. LIVER STEATOSIS SEGMENTATION WITH DEEP LEARNING METHODS. Proc IEEE Int Symp Biomed Imaging. 2019 Apr;2019:24-27. doi: 10.1109/isbi.2019.8759600. Epub 2019 Jul 11. PMID: 32670464; PMCID: PMC7363395.
Copy Download .nbib Format: NLM
Share it on

Proc IEEE Int Symp Biomed Imaging. 2019 Apr;2019:24-27. doi: 10.1109/isbi.2019.8759600. Epub 2019 Jul 11.

LIVER STEATOSIS SEGMENTATION WITH DEEP LEARNING METHODS.

Proceedings. IEEE International Symposium on Biomedical Imaging

Xiaoyuan Guo, Fusheng Wang, George Teodoro, Alton B Farris, Jun Kong

Affiliations

  1. Department of Computer Science, Emory University, Atlanta, GA, 30322, USA.
  2. Department of Computer Science, Stony Brook University, Stony Brook, NY, 11794, USA.
  3. Department of Computer Science, University of Brasília, Brasília, DF, Brazil.
  4. Department of Mathematics and Statistics, Georgia State University, Atlanta, GA, 30303, USA.

PMID: 32670464 PMCID: PMC7363395 DOI: 10.1109/isbi.2019.8759600

Abstract

Liver steatosis is known as the abnormal accumulation of lipids within cells. An accurate quantification of steatosis area within the liver histopathological microscopy images plays an important role in liver disease diagnosis and transplantation assessment. Such a quantification analysis often requires a precise steatosis segmentation that is challenging due to abundant presence of highly overlapped steatosis droplets. In this paper, a deep learning model Mask-RCNN is used to segment the steatosis droplets in clumps. Extended from Faster R-CNN, Mask-RCNN can predict object masks in addition to bounding box detection. With transfer learning, the resulting model is able to segment overlapped steatosis regions at 75.87% by Average Precision, 60.66% by Recall,65.88% by F1-score, and 76.97% by Jaccard index, promising to support liver disease diagnosis and allograft rejection prediction in future clinical practice.

Keywords: Liver steatosis segmentation; deep learning; whole-slide microscopy image

References

  1. Diagn Pathol. 2017 Nov 13;12(1):80 - PubMed
  2. Hum Pathol. 2004 Apr;35(4):430-5 - PubMed
  3. Conf Proc IEEE Eng Med Biol Soc. 2018 Jul;2018:3410-3413 - PubMed
  4. Am J Surg Pathol. 2017 Mar;41(3):365-373 - PubMed
  5. Gastroenterology. 2012 Jun;142(7):1592-609 - PubMed
  6. Conf Proc IEEE Eng Med Biol Soc. 2018 Jul;2018:810-813 - PubMed
  7. Histol Histopathol. 2009 Jan;24(1):49-60 - PubMed
  8. Clin Transplant. 2004 Dec;18(6):700-6 - PubMed
  9. Pathol Res Pract. 2013 Jun;209(6):371-9 - PubMed
  10. J Clin Pathol. 2001 Jun;54(6):461-5 - PubMed

Publication Types

Grant support