SWIFT Cast: efficacy and neural-biomechanical correlates of response

Walking speed: during walking forwards in a straight line at participant-selected speed. Each participant will be asked to undertake 4 walks. Walking speed will be measured in the middle of the walkway using 2 inexpensive infra-red light beams placed 3 metres apart on shoulder-high stands and connected to an electronic timer. Subjects who cannot walk without support, i.e. a FAC score of 2 or less (support of 1 person), will be deemed to have a walking speed of zero.

Ability to walk independently, functional mobility, ability to walk with a normal gait pattern and structural brain imaging:
1. Functional Ambulation Category (FAC) ranges from unable to walk (score 0) to able to walk independently on level/non-level surfaces (score 5). This measure is clinically relevant to walking function and has been found to have strong inter-rater and test-retest reliability.
2. Modified Rivermead Mobility Index: This reliable and clinically relevant scale measures functional mobility including turning over in bed, standing up, walking indoors and ascending stairs. This measure is used widely in stroke rehabilitation research.
3. Efficiency of gait measured by:
3.1. Peak angular velocity of the knee during walking as a sensitive objective measure of gait performance
3.2. The ratio of step times on the paretic and non-paretic lower limbs as a measure of temporal symmetry
3.3. The ratio of step lengths on the paretic and non-paretic lower limbs as a measure of spatial symmetry
3.4. The ratio of sagittal angular velocity of the knee of the paretic and contra-lateral lower limbs as a measure of joint symmetry
3.5. The angle of the tibia with respect to the vertical at initial contact, foot flat, mid stance and terminal contact as a measure of smooth forward progression of the lower leg. These measures are made simultaneously with walking speed using the clinical gait analysis equipment (see primary outcome). They can be undertaken easily in a clinical environment using a technique developed by Wall and applied by Rowe and others in which the participant walks across a 10 metre long mat on which a high contrast grid has been printed while being video recorded from the side view. The resultant video can be played back in slow motion and timed using a multi-lap stopwatch to determine the step times. It can also be viewed frame by frame and using the grid the spatial location of the feet during the walk can be determined to give step lengths. The average angular velocity of the knee during stance can be estimated for the gait cycle in which the participant is perpendicular to the camera by measuring the angle of the maximum and minimum knee angles using a computer generated goniometer, subtracting one from the other and then dividing them by the time between the two occurrences. Finally the angle of the tibia with respect to vertical can be determined by using the freeze frame mode and the computer generated goniometer.
4. MR scan (baseline only): in order to have an accurate delineation of the cerebral lesion, the structural neuroimaging will be done 3 - 8 weeks after onset. To achieve accurate delineation of the lesion will mean that, for some participants, the structural neuroimaging will not be undertaken at exactly the same time point as other baseline measures. Standard FLAIR and T1-weighted SPGR high-resolution "volume" data sets will be obtained prospectively from all subjects at baseline using a standardised acquisition sequence including field inhomogeneity correction and identical voxel size; data sets will be collected and processed in a single laboratory under J-CB's supervision. Following harmonisation of image characteristics for different scanners lesions will be automatically segmented on the FLAIR data set using appropriate seeding and then the lesion contours will be projected onto the T1-SPGR data set following reslicing. All T1-SPGR data sets will then be spatially normalised to the MNI template (including lesion masking if necessary) using SPM2.