University of Heidelberg
Faculty of Medicine Mannheim
University Hospital Mannheim
CKM receives DFG funding for Project "Prädiktion von Therapieansprechen und Outcome beim lokal fortgeschrittenen Rektum-Karzinom mittels Radiomics und Deep Learning: eine beispielhafte Anwendung für eine allgemein verwendbare, Deep Learning basierte Prozessierungs-Pipeline für die Bild-Klassifikation" read more.
Alena-Kathrin Schnurr presented on our research on AI in medical imaging @ Research Plus forum
read more
Dr. Mathias Davids wins the I. I. Rabi Award of the International Society for Magnetic Resonance in Medicine (ISMRM),
read more, article in local newspaper (Mannheimer Morgen)
Collaborate Research Projects

T2*- and Diffusion Measurements with 3He

Michael Bock


MR imaging of human lungs and airways is possible with hyperpolarized gases such as 3He [1]. Apart from the longitudinal relaxation time T1 two additional parameters determine the contrast of 3He images: the diffusion coefficient D and T2*. As D of 3He in the gas phase is 5 to 6 orders of magnitude larger than that of water, already the relatively weak imaging gradients of several mT/m create a significant diffusion weighting. In the human lung T2* is very short, because air-tissue boundaries induce rapid susceptibility changes.
To measure these two parameters simultaneously in an acquisition time, that does not exceed the life-time of hyperpolarization in the human lung of about 30 s, a sequence was developed, that employs two gradient-echo trains with different inter-echo spacings and readout gradient strengths.

Materials and Methods

The sequence for the simultaneous measurement of D and T2* was implemented on a clinical MR scanner (SIEMENS Magnetom 63/84 SP, Erlangen, Germany) using a home-built, double resonant 3He/1H Helmholtz coil (H.J. Zabel, DKFZ Heidelberg). Two gradient-echo trains with 8 alternating readout gradients each were sampled after a slice-selective rf excitation (Fig. 1). Readout gradient strength and bandwidth of the second echo train were half of that of the first to maintain spatial resolution (echo spacings: DTE1 = 4.56 ms, DTE2 = 7.12 ms, bandwidths: BW1 = 391 Hz/pixel, BW2 = 195 Hz/pixel).
Fig. 1: Timing of the 3He sequence. 2 series of 8 gradient echoes are acquired. Because echo spacing and readout gradient strength are different in the 2 series, the individual contributions of D and T2* to the signal decay can be separated.
The decay of the MR signal S during an echo train (normalized to the signal of the first echo S(0)) is described by



where n = 0..7 is the echo number and i = 1 or 2 is the series index. As the b-value bi(DTEi) is dependent on the details of the gradient switching, it is different for series 1 and 2. Values of D and T2* are calculated from the decay constants ai, which are determined by a fitting procedure.
Measurements with both hyperpolarized and non-prepolarized gas samples of 3He in glass cells were performed. Hyperpolarization of about 60% was created at the University of Mainz (Group: E. W. Otten) by laseroptical pumping of metastable 3He with subsequent collisional polarization transfer to the ground state [2]. MR imaging was performed using very low flip angles (a < 1°) to preserve hyperpolarization (Matrix / FOV / SL / NEX = 96-128 / 500 mm / 40 mm / 1). The non-prepolarized sample contained a 3He/O2 gas mixture in a sealed cell (T1 = 2.5 s). At TR = 100 ms flip angles around the Ernst angle aE = 16° were chosen. The other imaging parameters were: Matrix / FOV / SL / NEX = 642 / 300 mm / 40 mm / 1. For comparison, 3He diffusion coefficients were also measured with a spectroscopy sequence.

Results and Discussion

Fig. 2 shows the result of a measurement with hyperpolarized 3He. The measured value of D is in good agreement with the spectroscopy measurement (1.87× 10-4 m2/s) and the literature value (1.83×10-4 m2/s, T = 20°C, p = 1 bar). In general it was observed, that T2*-values showed larger uncertainties than values of D.
From the 2 image series maps of D and T2* can be calculated. Changes in the structure of lung tissue modify local diffusion coefficients and T2*. Parameter maps derived from 3He imaging data might help to identify regions in the human lung, that are affected by pathologic processes.
Fig. 2: Normalized 3He MR signal of the hyperpolarized sample as a function of echo spacings n TE1 (squares) and n TE2 (circles). From the fitted constants a1 = 7.5 ms and a2 = 10.1 ms and b-values b1(TE1) = 2.62 103 s/m2 and b2(TE2) = 2.76 103 s/m2 D and T2* were calculated. 


[1] Bachert P., Schad L.R., Bock M., Magn. Reson. Med. 36, 192-196 (1996)
[2] Eckert G., Heil W., Meyerhoff M., et. al.: Nucl. Instrum. Meth. A 346, 45-51 (1994)
Contact: Dr. Frank Zöllner last modified: 19.02.2008
to top of page