How to calculate Intraclass Correlation Coefficient (ICC) for non-normal data?

In our study which includes 12 persons, each person does three repetitions of a shoulder flexion movement and we measured three peak angles for each person by a camera. Then, one week later same subjects perform the same procedure exactly.I would like to calculate find ICC (Intaclass Correlation Coefficient) , SEM (Standard Error of Measurement) and Minimum Detectable Change (MDC) to assess Test-Retest Reliability of measurements
The problem is : ICC assumes data to be disributed normally but my data does not fit normal disribution. Is there any method to calculate ICC for non-normal data?