Transfer learning with deep convolutional neural network for liver steatosis assessment in ultrasound images - PubMed
Transfer learning with deep convolutional neural network for liver steatosis assessment in ultrasound images
Michał Byra et al. Int J Comput Assist Radiol Surg. 2018 Dec.
Abstract
Purpose: The nonalcoholic fatty liver disease is the most common liver abnormality. Up to date, liver biopsy is the reference standard for direct liver steatosis quantification in hepatic tissue samples. In this paper we propose a neural network-based approach for nonalcoholic fatty liver disease assessment in ultrasound.
Methods: We used the Inception-ResNet-v2 deep convolutional neural network pre-trained on the ImageNet dataset to extract high-level features in liver B-mode ultrasound image sequences. The steatosis level of each liver was graded by wedge biopsy. The proposed approach was compared with the hepatorenal index technique and the gray-level co-occurrence matrix algorithm. After the feature extraction, we applied the support vector machine algorithm to classify images containing fatty liver. Based on liver biopsy, the fatty liver was defined to have more than 5% of hepatocytes with steatosis. Next, we used the features and the Lasso regression method to assess the steatosis level.
Results: The area under the receiver operating characteristics curve obtained using the proposed approach was equal to 0.977, being higher than the one obtained with the hepatorenal index method, 0.959, and much higher than in the case of the gray-level co-occurrence matrix algorithm, 0.893. For regression the Spearman correlation coefficients between the steatosis level and the proposed approach, the hepatorenal index and the gray-level co-occurrence matrix algorithm were equal to 0.78, 0.80 and 0.39, respectively.
Conclusions: The proposed approach may help the sonographers automatically diagnose the amount of fat in the liver. The presented approach is efficient and in comparison with other methods does not require the sonographers to select the region of interest.
Keywords: Convolutional neural networks; Deep learning; Hepatorenal index; Nonalcoholic fatty liver disease; Transfer learning; Ultrasound imaging.
Conflict of interest statement
Conflict of interest
The authors do not have any conflict of interest.
Ethical statements
All procedures performed in studies involving human participants were in accordance with the Ethical Standards of the Medical University of Warsaw and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
Figures

Histogram of steatosis level across the population of patients in the study group

Liver B-mode images and the ROIs selected for HI calculation, a steatosis level of 3% and b 25%, respectively

Illustration of feature extraction using the Inception-ResNet-v2 model [28]

The validation pipeline

The ROC curves for the HI method (AUC = 0.959), the GLCM algorithm (AUC = 0.893) and the classifier developed using CNN features (AUC = 0.977)

The usefulness of a the HI parameter (SCC = 0.80), b GLCM-based features (SCC = 0.39) and c the CNN-based features (SCC = 0.78) in steatosis level assessment

a The relation between the HI parameter and the Lasso regression (SCC = 0.78) and b the corresponding Bland–Altman plot
Similar articles
-
Constantinescu EC, Udriștoiu AL, Udriștoiu ȘC, Iacob AV, Gruionu LG, Gruionu G, Săndulescu L, Săftoiu A. Constantinescu EC, et al. Med Ultrason. 2021 May 20;23(2):135-139. doi: 10.11152/mu-2746. Epub 2020 Dec 29. Med Ultrason. 2021. PMID: 33626114
-
Sonographic hepatorenal ratio: a noninvasive method to diagnose nonalcoholic steatosis.
Borges VF, Diniz AL, Cotrim HP, Rocha HL, Andrade NB. Borges VF, et al. J Clin Ultrasound. 2013 Jan;41(1):18-25. doi: 10.1002/jcu.21994. Epub 2012 Sep 20. J Clin Ultrasound. 2013. PMID: 22997020
-
Computer aided diagnosis of fatty liver ultrasonic images based on support vector machine.
Li G, Luo Y, Deng W, Xu X, Liu A, Song E. Li G, et al. Annu Int Conf IEEE Eng Med Biol Soc. 2008;2008:4768-71. doi: 10.1109/IEMBS.2008.4650279. Annu Int Conf IEEE Eng Med Biol Soc. 2008. PMID: 19163782
-
Duncan K, Vealé BL. Duncan K, et al. Ultrasound Q. 2024 Jun 18;40(3):e00681. doi: 10.1097/RUQ.0000000000000681. eCollection 2024 Sep 1. Ultrasound Q. 2024. PMID: 38889400 Review.
-
Artificial intelligence in medical imaging of the liver.
Zhou LQ, Wang JY, Yu SY, Wu GG, Wei Q, Deng YB, Wu XL, Cui XW, Dietrich CF. Zhou LQ, et al. World J Gastroenterol. 2019 Feb 14;25(6):672-682. doi: 10.3748/wjg.v25.i6.672. World J Gastroenterol. 2019. PMID: 30783371 Free PMC article. Review.
Cited by
-
Shalbaf A, Bagherzadeh S, Maghsoudi A. Shalbaf A, et al. Phys Eng Sci Med. 2020 Dec;43(4):1229-1239. doi: 10.1007/s13246-020-00925-9. Epub 2020 Sep 14. Phys Eng Sci Med. 2020. PMID: 32926393
-
Lupsor-Platon M, Serban T, Silion AI, Tirpe GR, Tirpe A, Florea M. Lupsor-Platon M, et al. Cancers (Basel). 2021 Feb 14;13(4):790. doi: 10.3390/cancers13040790. Cancers (Basel). 2021. PMID: 33672827 Free PMC article. Review.
-
Nduma BN, Al-Ajlouni YA, Njei B. Nduma BN, et al. Cureus. 2023 Dec 15;15(12):e50601. doi: 10.7759/cureus.50601. eCollection 2023 Dec. Cureus. 2023. PMID: 38222117 Free PMC article. Review.
-
Arntfield R, Wu D, Tschirhart J, VanBerlo B, Ford A, Ho J, McCauley J, Wu B, Deglint J, Chaudhary R, Dave C, VanBerlo B, Basmaji J, Millington S. Arntfield R, et al. Diagnostics (Basel). 2021 Nov 4;11(11):2049. doi: 10.3390/diagnostics11112049. Diagnostics (Basel). 2021. PMID: 34829396 Free PMC article.
-
Stage-independent biomarkers for Alzheimer's disease from the living retina: an animal study.
Ferreira H, Serranho P, Guimarães P, Trindade R, Martins J, Moreira PI, Ambrósio AF, Castelo-Branco M, Bernardes R. Ferreira H, et al. Sci Rep. 2022 Aug 11;12(1):13667. doi: 10.1038/s41598-022-18113-y. Sci Rep. 2022. PMID: 35953633 Free PMC article.
References
-
- Beeman SC, Garbow JR. Imaging and metabolism. New York: Springer; 2018.
MeSH terms
LinkOut - more resources
Full Text Sources
Other Literature Sources