Digital liver biopsy: Bio-imaging of fatty liver for translational and clinical research

Table 1 Intrahepatic fat measurement

Ref.	Imaging modality	Classification	ROC-derived parameters
			AUC	Sensitivity (%)	Specificity (%)
Mancini et al[60] 2009	US	1H-MRS fat content > 5%	0.996	100	95
Xia et al[61] 2012	US	1H-MRS fat content > 5.56%	NA	95.1	100
Edens et al[62] 2009	US	1H-MRS fat content > 5.56%	NA	66.7	100
Di Lascio et al[65] 2018	US	1H-MRS fat content > 5%	0.97	89	94
Sasso et al[108] 2012	CAP score (imaging derived)	Liver biopsy	S0vs S1S2S3: 0.80	S0vs S1S2S3: 76	S0vs S1S2S3: 71
		S0: < 10% of hepatocytes	S0S1vs S2S3: 0.86	S0S1vs S2S3: 87	S0S1vs S2S3: 74
		S1: 11%-33% of hepatocytes	S0S1S2vs S3: 0.88	S0S1S2vs S3: 78	S0S1S2vs S3: 93
		S2: 34%-66% of hepatocytes
		S3: 67%-100% of hepatocytes

Both histologic and digital liver biopsy provided reliable measures of intrahepatic fat that are significantly correlated, but categorically different. Liver biopsy describes the histologic characteristics of the pathologic lesions and accounts for the percentage of hepatocytes with intracellular fat-derived vacuoles using categorical grading systems that are not directly representative of the hepatic triglyceride concentration[19-21]. On the other hand

¹H-MRS measures protons in acyl groups of liver tissue triglycerides and provides continuous quantitative values expressed as mg/g of hepatic tissue[109]. Moreover, ¹H-MRS uses a much larger volume of liver tissue than biopsy reducing sampling error and representing the most accurate measure of the overall liver triglyceride content. US: Ultrasound; ROC: Receiving operator characteristic; CAP: Controlled attenuation parameter.