Tim symphony

The imaging of the singers’ breathing apparatus was performed using a clinical 1.5 T MRI system (Tim Symphony, Siemens, Erlangen, Germany) as reported in the pilot study (Traser et al., 2017b (link)). The subject positioning was supine (see limitation section for detailed discussion of the potential effects of gravity on the results). Dynamic imaging was done using a 2D trueFISP imaging sequence (repetition time/echo time = 3/1.5 ms, α = 6°, bandwidth (BW) = 977 Hz/px, slice thickness = 10 mm, acquisition matrix = 256, field of view (FOV) = 420 mm) with a temporal resolution of approximately three frames per second. First, localizer images at three different planes were acquired to define the imaging planes in sagittal and coronal orientations. For the sagittal trueFISP images, a slice through the right lung was chosen to avoid artefacts caused by heart motion, which would complicate the image analysis. The sagittal plane was selected so that the vertex of the DPH cupola and the apex of the lung could be identified (see Figure 1 for details). The coronal plane was chosen similarly, encompassing both vertices of the left and right DPH cupola and the apices of the left and right lung. During imaging the subjects wore headphones for hearing protection and communication.

MRI was performed on a 1.5 Tesla MRI system (TIM Symphony, Siemens, Germany). With a 5 mm slice thickness, standardized T2 (repetition time (TR) 4,000–6,000 ms, time to echo (TE) 102 ms), T1 (TR 573 ms, TE 12 ms), fluid attenuated inversion recovery (FLAIR, TR 6180 ms, TE 112), T2^* (TR 800–1,000 ms, TE 24 ms) and DWI (TR 4,000–6,000 ms, TE 100 ms) were acquired. ADC maps were obtained from DWI. All MRI data sets were obtained in the same orientation and slice positions. We recorded all findings in the first MRI performed after seizure activity, as well as on subsequent MRIs at our department and the time between seizure onset and first MRI. DWI abnormalities were only accepted if a high contrast to background was detectable, or a corresponding reduction of ADC was present and artifacts could be excluded. Furthermore, DWI alterations were only classified as seizure-related when they were cortical or localized in adjacent subcortical areas and not respecting vascular territories or when they were restricted to the hippocampal area or pulvinar thalami.

The first system was a clinical whole-body 1.5 T MRI system (Tim Symphony, Siemens, Erlangen, Germany) which was also used in a pilot study^{8 (link)}. The subject positioning and measurement were here only possible in supine body positioning (referred to as supine hMRI). In accordance with the pilot study, a 3D localizer data was first applied to position a sagittal slice in the right lung in which the vertex of the DPH cupola and the apex of the lung could be identified. Then, a dynamic 2D trueFISP imaging sequence (repetition time/ echo time (TR/TE) = 3/1.5 ms, α = 6°, bandwidth (BW) = 977 Hz/px, slice thickness (ST) = 10 mm, acquisition matrix = 256, field of view (FOV) = 420 mm) was applied with a temporal resolution of approximately 3 frames per second (fps) while the singer performed each singing task. The subjects wore headphones for hearing protection and communication during the session.