The split echo train method acquires the same number of echoes in the same total acquisition time as the full echo train. However, in this method each echo train is halved; the echoes in the first half are stored in the k-space for TEeff1 and the echoes in the second half contribute to the k-space data for TEeff2. (See Figure 1.22)

Figure 1.22 Diagram representing the Split Echo Train technique. In this method, the first half of the ETL is used to fill k-space for TEeff1, and the second half of the ETL is used to fill k-space for TEeff2.

For example, with a minimum echo space of 17 ms, ETL = 8, and a 256x256 acquisition matrix, the echo train would be divided as follows: The first four echoes (17, 34, 51, 68 ms) would contribute to TEeff1 and the last four echoes (85, 102, 119, 136 ms) would contribute to TEeff2. The central phase encodes would occur at TEeff1 for the first half of the echo train and at TEeff2 for the second half of the echo train. Because the split echo train sequence combines data from fewer echoes in an echo train to form an image, the contrast generated with a split echo train sequence will be more like the contrast in a conventional SE sequence than will the full echo train sequence. With all scan parameters the same, the TEeff1 should exhibit less blurring and increased relative proton density weighting in the split echo train method. Because the echo spacing is constant throughout the echo train, the selection of TEeff1 and TEeff2 is not as flexible with the split train method as it is with full echo train. In the above split echo train example, TEeff1 would have to occur at 17 , 34, 51, or 68 ms. TEeff2 would have to occur at 85, 102 , 119, or 136 ms. In other words, TEeff1 must occur at an echo in the first half of the echo train and the TEeff2 during the second half.