Multilayer perceptron applied to the clinical prediction of hepatic steatosis
Keywords:
fatty liver; clinical record; neural networks; computer-assisted decision makingAbstract
Introduction: Hepatic steatosis is the accumulation of hepatic fat; In general, without visible symptoms, and with the possibility of progressing to steatohepatitis and cirrhosis. An early diagnosis is crucial for its management.
Objective: To predict the presence or absence of hepatic steatosis using clinical data through a multilayer perceptron model.
Methods: Analytical and cross-sectional study of 912 adults from a secondary database of the Dryada repository. The dependent variable was hepatic steatosis. Eleven laboratory tests and body mass index were used as predictive variables. Multilayer perceptron neural networks were used, evaluated by means of a classification table of correct predictions and predictive capacity of the model.
Results: The area under the curve of the model was 0.917. In the training set, the percentage of correct prediction to identify hepatic steatosis was 71.10% and to rule it out, 92.70%. In the test phase, these values were 70.50% and 88.50%, respectively. When comparing the presence or absence of hepatic steatosis diagnosed by ultrasound with the classification obtained by multilayer perceptron model, a moderate association (Phi coefficient=0.612) and an acceptable concordance (Kappa=0.59) were observed. Patients classified as having the disease by the model presented it on ultrasound 18.48 times more frequently than those classified without it (OR=18.48; 95% CI: 11.4-29.5). Sensitivity was 71%, specificity 89%, and positive and negative predictive values were 71% and 88%, respectively.
Conclusions: The multilayer perceptron neural network demonstrated a high capacity to rule out hepatic steatosis from clinical laboratory data.
Downloads
References
1. Poniachik J, Roblero JP, Urzúa A, et al. A new definition for non-alcoholic fatty liver disease. J Hepatol. 2021;74(4):982-3. DOI: 10.1016/j.jhep.2020.09.002.
2. Ramai D, Facciorusso A, Vigandt E, et al. Progressive liver fibrosis in non-alcoholic fatty liver disease. Cells. 2021;10(12):3401. DOI: 10.3390/cells10123401.
3. Teng ML, Ng CH, Huang DQ, et al. Global incidence and prevalence of nonalcoholic fatty liver disease. Clin Mol Hepatol. 2023;29(Suppl):S32-42. DOI: 10.3350/cmh.2022.0365.
4. Lee CM, Yoon EL, Nakajima A, et al. A reappraisal of the diagnostic performance of B-mode ultrasonography for mild liver steatosis. Am J Gastroenterol. 2023;118(5):840-7. DOI: 10.14309/ajg.0000000000002020.
5. Kontrick AV, VanWagner LB, Yeh C, et al. Hepatic steatosis: An incidental finding that deserves attention. Acad Emerg Med. 2021;28(5):578-81. DOI: 10.1111/acem.14174.
6. Fukuda T, Hamaguchi M, Kojima T, et al. Association between serum γ-glutamyltranspeptidase and atherosclerosis: a population-based cross-sectional study. BMJ Open. 2014;4(10):e005413. DOI: 10.1136/bmjopen-2014-005413.
7. Majidzadeh Gorjani O, Byrtus R, Dohnal J, et al. Human activity classification using multilayer perceptron. Sensors (Basel). 2021;21(18):6207. DOI: 10.3390/s21186207.
8. Abdullah M. Artificial intelligence-based framework for early detection of heart disease using enhanced multilayer perceptron. Front Artif Intell. 2025;7:1539588. DOI: 10.3389/frai.2024.1539588.
9. Nahm FS. Receiver operating characteristic curve: overview and practical use for clinicians. Korean J Anesthesiol. 2022;75(1):25-36. DOI: 10.4097/kja.21209.
10. Zhao Q, Lan Y, Yin X, et al. Image-based AI diagnostic performance for fatty liver: a systematic review and meta-analysis. BMC Med Imaging. 2023;23(1):208. DOI: 10.1186/s12880-023-01172-6.
11. Ghandian S, Thapa R, Garikipati A, et al. Machine learning to predict progression of non-alcoholic fatty liver to non-alcoholic steatohepatitis or fibrosis. JGH Open. 2022;6(3):196-204. DOI: 10.1002/jgh3.12716.
12. Yao Y, Zhang Z, Peng B, et al. Bio-inspired network for diagnosing liver steatosis in ultrasound images. Bioengineering (Basel). 2023;10(7):768. DOI: 10.3390/bioengineering10070768.
13. Lizuka K, Takao K, Yabe D. ChREBP-mediated regulation of lipid metabolism: Involvement of the gut Microbiota, liver, and adipose tissue. Front Endocrinol (Lausanne). 2020;11:587189. DOI: 10.3389/fendo.2020.587189.
14. Lu S, Xie Q, Kuang M, et al. Lipid metabolism, BMI and the risk of nonalcoholic fatty liver disease in the general population: evidence from a mediation analysis. J Transl Med. 2023;21(1):191. DOI: 10.1186/s12967-023-04047-0.
15. Xuan Y, Wu D, Zhang Q, et al. Elevated ALT/AST ratio as a marker for NAFLD risk and severity: insights from a cross-sectional analysis in the United States. Front Endocrinol (Lausanne). 2024;15:1457598. DOI: 10.3389/fendo.2024.1457598.
16. Ramatchandirin B, Pearah A, He L. Regulation of liver glucose and lipid metabolism by transcriptional factors and coactivators. Life (Basel). 2023;13(2):515. DOI: 10.3390/life13020515.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Alberto Guevara-Tirado
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
All content published in this journal is Open Access, distributed under the terms of the CC BY-NC 4.0 License.
It allows:
- Copy and redistribute published material in any medium or format.
- Adapt the content.
This will be done under the following terms:
- Attribute the authors' credits and indicate whether changes were made, in which case it must be in a reasonable way.
- Non-commercial use.
- Recognize the journal where it is published.
The copyrights of each article are maintained, without restrictions.
