Optimising T2 relaxation measurements on MS patients utilising a multi-component tissue mimicking phantom and different fitting algorithms in T2 calculations

Purpose: To evaluate the optimal regression fitting algorithm for T2 relaxation time measurements on relapsing remitting multiple sclerosis (RRMS) patients and Healthy Subjects (HS) with the aid of a multi-component tissue mimicking phantom. Material and Methods: Twenty eight glass test tubes were filled mainly with (a) standard EUROSPIN test objects, (b) Gd-DTPA hydatic solutions, (c) milk creams with various fat contents and (d) raw eggs with various relative concentrations of egg-white and egg-yellow parts. Two patients with relapsing remitting multiple sclerosis (RRMS) and a healthy volunteer were examined. A multi-echo spin echo sequence (32 echoes) was used for all the phantom and human subjects T2 measurements. T2 relaxation parametric maps for the phantoms and the human subjects were calculated utilising a Conventional-Linear (CL), a Weighted-Linear (WL), a Non-Linear (NL) and a Double-Exponential-Non-Linear-Fit (DENLF) regression fitting methods. Results: A single T2 relaxation behaviour was observed for EUROSPIN and Gd-DTPA solution test tubes. A double T2 relaxation behaviour, revealed only by DENLF, was observed for milk creams, raw eggs, Normal-White-Matter (NWM) of a healthy subject and Normal-Appearing-White-Matter (NAWM) and focal lesions of RRMS patients. Conclusions: WL, NL and DENLF algorithms proved to be an excellent means for optimised measurements of T2 values on tissue mimicking phantoms, NWM, NAWM and demyelinating lesions of RRMS patients. DENLF provides additional information related to differentiation of molecular environments in either phantoms or human subjects.