Measurement parameters were as follows: TR was set to 8 ms; TE was 2.5 ms; number of averages was 4; slice thickness 2 mm; spacing between slices was 0.4 mm; matrix size 320 × 320 pixels; low flip angle excitation pulse was automatically set to 4 degrees; the high flip angle excitation pulse was automatically set to 21 degrees (setting of excitation pulse flip angles was based on a

prior estimate of expected T1 values); pixel resolution was 0.625 × 0.625 mm; total measurement time was 2 min 34 sec × 2; FOV was 200 × 200 mm; pixel bandwidth was 161 Hz/pix; and number of slices was 16. For the Ganetespib concentration 2D inversion recovery sequences, one region of interest (ROI) was drawn on the shortest inversion time image, covering the entire TMJ disc, and subsequently 3 ROIs were drawn separately on the anterior, middle and posterior parts of the TMJ disc, using the Syngo Siemens built-in standard evaluation software. The ROIs were copied and pasted onto the rest of the inversion time images. Signal intensities in each ROI were recorded and T1 maps were calculated offline, using an IDL fitting routine based on the curvefit IDL code (by Craig B. Markwardt, NASA/GSFC Code 662, Greenbelt, MD 20770, [email protected]).

For the 3D-GRE dual flip angle technique, T1 maps were calculated online using the built-in Syngo Siemens software. All cases Ku-0059436 chemical structure were analyzed, and ROIs were drawn by one observer (E.P., a dentist who has specialized in orthodontics for four years and TMDs for three years). ROIs were manually defined on the disc of the right and left TMJ, and three ROIs (anterior, central, posterior) each were manually defined for different parts of the disc (Fig. 3).

The observer attempted to include as many pixels as possible into the ROIs, to diminish non-systematic errors. The range of ROI sizes was between 0.25 cm2 (92 pixels) and 0.13 cm2 (48 pixels). Astemizole In order to compare T1 values measured by IR (30 and 60 minute intervals) with 3D GRE (8÷10 minute intervals), interpolation of the 3D GRE data was performed. The 3D GRE data measured in 8–10 minute intervals were interpolated into one minute intervals. Subsequently to match the IR time points (30 and 60 minute intervals), the GRE time points at 30 or 60 minutes were selected from interpolated data. Interpolation was performed using IDL software (RSI, Boulder, CO), using built-in SPL_INTERP” routine, which provides spline interpolation over the measured dataset in selected time points. SPL_INTERP is based on the routine “spline” described in section 3.3 of Numerical Recipes in C: The Art of Scientific Computing (Second Edition), published by Cambridge University Press, and is used in IDL by permission. All statistical evaluations were performed using IBM SPSS Statistics Version 19.0. Metric data, such as T1 values, are presented using mean +/− SD. Mean values and standard deviation for each ROI were recorded and statistically analyzed using a two-way ANOVA for repeated measures.