Validation of a mobile NMR sensor for the determination of skin layers and the local estimation of the T2 relaxation time distribution in the lower arm

Abstract

Purpose/Introduction: The Nuclear Magnetic Resonance Mobile Universal Surface Explorer (NMR-MOUSE) was evaluated for its capability in studying the morphology and the distribution of T2 relaxation times of the human skin. Subjects and methods: In 10 male subjects the thicknesses of the whole skin and its major dermal layers at the inner side of the left lower arm were measured for 5 times in the native state and after a venous occlusion manoeuvre. Skin structure was analyzed by means of a Profile NMR-MOUSE® (1H, 20 MHz). The native skin was also measured by means of an ultrasound scanner. Furthermore in 5 subjects the distribution of the T2 relaxation times was analyzed in two slices placed in the dermis and in the subcutis, respectively. T2 analysis was performed in the native state and after application of a rheumatic salve that visibly increases skin perfusion. Results: The dermis including the epidermis showed a rather homogeneous NMR signal intensity and the subcutis was characterized by more variable and higher signal intensity. Due to a large variation in repetitively recorded profiles, only in 1 of 10 subjects the 5 repeated venous occlusion experiments resulted in a significantly detectible increase in measured thicknesses values. Mean thickness values of the whole skin measured with ultrasound imaging were not different from the NMR values. The distribution of T2 relaxation times calculated with an inverse Laplace transformation 3 peaks occur at 5 ms, 10 ms, and 100 ms, respectively. In the dermis the 10 ms peak was dominant. In the subcutis the 10 ms and the 100 ms peaks were almost equally dominant. Due to large variations in the T2 distribution the salve only caused a significant attenuation of the 100 ms peak in the dermis layer. Moreover the T2 echo trains were analysed with a mono- and bi-exponential fit. In contrary to the mono fit the bi-exponential fit is a sufficient approximation of the T2 signal. For the mono-exponential fit no statistical changes occurred but for the bi-exponential fit a significant (p < 0.05) reduction in signal intensity in the dermis layer could be found for both the short and long T2 time due to the action time of the rheumatic salve. Dicussion/Conclusion: Reliable examinations of skin structure by means of a NMR-MOUSE were confronted by a poor reproducibility and long acquisition times. However, the increased skin perfusion by the rheumatic salve caused significant effects on T2 in the dermal layer.