Gandon et al used Magnetic Resonance Imaging (MRI) to estimate hepatic iron content. The ratio of liver to muscle signal intensity under various imaging conditions can identify significant increases in hepatic iron stores. The authors are from Hopital Pontchaillou in Rennes, France.


Liver iron content is normally < 36 µmol per gram liver dry weight.


Field strength for the MRI was 1.5 T. GRE (gradient recalled echo) sequences ranged from T1 to T2++.


ratio liver to muscle signal intensity =

= (signal intensity in liver) / (signal intensity in muscle)


Using highly weight GRE sequences (T2++ images, Figure 1):

• Threshold for ratio was 0.88. A value < 0.88 correlated with hepatic iron stores > 60 µmol per gram liver dry weight.

• A ratio of 0.2 was reached at about 86 µmol per gram liver dry weight.


Using T1 images (Figure 2), the data was a curve.

• The red line at a ratio of 0.13 corresponded with hepatic iron stores > 355 µmol per gram liver dry weight.


The authors developed an algorithm to estimate hepatic iron concentrations from the most sensitive sequence (the one with L/M ratio greater than its own saturation threshold). The algorithm is available at



• For the T2++ images, the sensitivity was 89% and specificity 80% for the threshold of 0.88. A threshold of 1 had a sensitivity of 95% and specificity of 88%.

• MRI has several advantages over liver biopsy for determining hepatic iron content. It was more acceptable to patients. About 3% of patients failed the MRI, while 5% of liver biopsies may too small for doing iron measurement studies.


To read more or access our algorithms and calculators, please log in or register.