Digital image analysis in liver fibrosis: basic requirements and clinical implementation

Yegin E. G. , YEĞİN K., ÖZDOĞAN O. C.

BIOTECHNOLOGY & BIOTECHNOLOGICAL EQUIPMENT, vol.30, no.4, pp.653-660, 2016 (Journal Indexed in SCI) identifier

  • Publication Type: Article / Review
  • Volume: 30 Issue: 4
  • Publication Date: 2016
  • Doi Number: 10.1080/13102818.2016.1181989
  • Page Numbers: pp.653-660


Accurate assessment of liver fibrosis is a critical aspect of diagnosis, prognosis prediction, surveillance strategies, therapeutic planning and monitoring, and also for validation of non-invasive surrogates of fibrosis. Traditional histopathological stagings depend on subjective visual interpretation process of architectural changes of fibrosis without providing quantification as continuous numerical data, but rather in the form of discrete staging. This makes high level reproducibility practically impossible in its application, which should be minimized in scientific research. In the light of increasing demand for an objective method, digital image analysis (DIA) technology has been increasingly implemented for liver fibrosis assessment. Potential advantages and applications of reproducible quantitative fibrosis ratio measurements with DIA include performing broader scale of statistical analysis and comparison between studies, monitoring minor but potentially important quantity changes during fibrosis regression or progression (especially in the context of therapeutic trials), and to be a better histological reference standard for validity and accuracy of surrogates of fibrosis. DIA may also have a potential role within the new perspective of redefining and sub-classifying cirrhosis. Since DIA algorithm covers multiple domains of hepatopathology and engineering, it may seem to be complicated to a researcher. This review provides an understanding of all basic steps, techniques, clinical applications of computerized image analysis for the particular purpose of liver fibrosis aiming its better implementation in hepatology research. Further work is required for standardization of all stages of pre-imaging, digital image acquisition and digital image processing steps for generation of reproducible outputs.