Introduction

Cancer is the second leading cause of death in the United States1. MR studies of ex-vivo cancer tissues of breast, prostate and rat brain show that increased contrast (T₁ weighted) between normal and cancerous tissues can be obtained at ultra-low fields^2-4 as compared to high fields. Diffusion measurements using Pulsed Gradient Spin Echo (PGSE) sequence in polarised magnetic field has been widely used to characterize materials⁵.Polarizing field systems employ a polarizer pulse of 3 to 5 seconds long to increase the net magnetization. Variation of net magnetization is achieved by varying the strength of the polarization pulses. Hence, T₁ and T₂ measurements at Earth’s field have long acquisition times. Current work demonstrates a Fast, Robust and SimulTaneous T₁ and T₂ sequence called ‘FRoST’ using CPMG sequence on water and CuSO₄ phantoms.

Methods

All experiments were performed on an Earth’s field MRI system, Terranova (Magritek Gmbh, Germany). FRoST creates a variation of the net magnetization by varying the polarizing pulse (τ_p) through a series of CPMG acquisitions. CPMG was chosen for the implementation of FRoST due to its inherent acceleration capability by collecting multiple echoes in a single TR and hence requires reduced acquisition time for determination of T₂. All experiments were carried out on a 500ml water phantom and CuSO₄ phantoms of concentrations 0.25mM, 0.5mM and 1mM. One of the Fast schemes of FRoST termed as ‘FRoST-Fast 1’ employs variation of τp linearly from 1000ms to 4000ms with a step of 1000ms. Another fast scheme of FRoST termed as ‘FROST-Fast 2’ had τp chosen as 1000, 2000, 4000 and 5000ms to acquire exponentially increasing and saturation regions of the T₁ relaxation curve. For robust measurements, τ_p was varied from 500ms to 5000ms with a step of 500ms. Minimum possible echo time of 200ms for water and 80ms for CuSO₄ phantoms were for increased SNR and T weighting. The first signal intensity of every CPMG acquisition was curve fitted as a function of τp using the equation S=S₀ (1- e –τ_p/T₁) to obtain T₁ and; T₂ was obtained directly as a result of CPMG. The relaxometry measurements and their acquisition times were compared with standard T₁ and T₂ sequences like ‘T₁B_p’, ‘T₁B_e’, ‘T₂’ and ‘CPMG’. Diffusion mapping was also performed using PGSE sequence. The probe of the Terranova consists of a specific diffusion gradient coil with a field gradient per unit current of 2.38 (mT/m/A). The pulse width, δ, of the diffusion-weighting gradient was varied from 3 ms to 60 ms, in steps of 3. The polarisation period was 4000ms for a polarizing current of 6A. PGSE gradient coil produced a magnetic field gradient (G) of 4.76 (mT/m) for a maximum allowable current of 2 A. The distance between two gradient pulses, Δ, was 200 ms.

Results

Figures 1-4 show the T₁ and T₂ obtained using standard sequences (macros) for T₁ and T₂ measurements in Terranova and those obtained using FRoST for 1 and 4 averages for water and CuSO4 phantoms. It can be observed that FRoST provides fast and robust relaxometry measurements that are similar to the standard macros available in Terranova. Table 1 shows the acquisition times for T₁ and T₂ measurements from standard sequences and FRoST. It can be seen that FRoSt provides accelerated and simultaneous relaxometry measurements in much lesser time as compared to standard sequences. It also shows the value of diffusion coefficient for water and CuSO₄ phantoms at Earth’s field obtained by the PGSE experiment. Each diffusion coefficient is the mean value of 3 trials. The magnitude of the apparent diffusion coefficient tends to decrease with increase in the concentration of the CuSO₄ solution, as expected.

Discussion

The first echo of every TR over an entire scan is essentially T₁ weighted and the echoes obtained in each acquisition are T₂ weighted. The signal intensities from FRoST are however reduced due to effects of shorter τ_p. FRoST employs series of CPMG acquisitions which results in T₂ values to be inherently averaged 4 times in the case of fast acquisitions and 10 times in the case of robust acquisitions.

Conclusion

Increased averaging results in better T₁ measurements in a time much lesser compared to 1 average acquisition of T₁B_p and T₁B_e sequences. This would aid in rapid determination of T₁ and T₂ of dynamically changing biochemical reactions such as protein exchange reactions⁴. Variation of τp is analogous to variation of TR which enables application of FRoST to high field systems as well. Future work involves in application of FRoST and PGSE experiment to characterize tumours ex-vivo at Earth’s field

Acknowledgements

Department of Science and Technology (DST). DST/TSG/NTS/2013/100