Premium
The use of guideline images to improve histological estimation of hepatic steatosis
Author(s) -
Hall Andrew R.,
Green Anna C.,
Luong TuVinh,
Burroughs Andrew K.,
Wyatt Judith,
Dhillon Amar P.
Publication year - 2014
Publication title -
liver international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.873
H-Index - 110
eISSN - 1478-3231
pISSN - 1478-3223
DOI - 10.1111/liv.12614
Subject(s) - steatosis , guideline , concordance , grading (engineering) , magnification , medicine , liver steatosis , fatty liver , radiology , pathology , artificial intelligence , computer science , disease , biology , ecology
Background & Aims Guideline images of specific fat proportionate area (FPA) percentages have recently been published to aid the histological assessment of liver steatosis as subjective estimates of FPA are usually overestimated. To assess, (i) the effect of guideline images on accuracy and concordance of estimated FPA ( eFPA ), (ii) experience of steatosis grading systems on eFPA , (iii) the effect of magnification on assessment of FPA (iv) and produce a range of guideline images at x4 objective magnification (OM). Methods Two circulations of sample images (C1 and C2) were circulated to UK liver external quality assessment histopathology scheme members who were asked to independently evaluate steatosis. Each circulation consisted of 15 images taken at both x20 and x4 OM representing the full range of steatosis. C1 was distributed first, then C2 with guideline images of FPA 6 weeks later. Results Participants overestimated FPA in C1. In C2, there was significant improvement in accuracy ( P < 0.001) of eFPA for sample images with mFPA >5%. Concordance of x4OM eFPA was substantial in both circulations (C1 K = 0.878, C2 K = 0.724). Conclusion The tendency to overestimate eFPA has been corroborated and can be largely corrected with the use of guideline images (without needing digital image analysis). There is a need to redefine steatosis grades that are clinically significant and validated using an accurate quantification of steatosis.