The sensorimotor cortex may play a role in the functional recovery after Spinal Cord Injury (SCI) through efference generated in the absence of the afference. OBJECTIVE: The purpose of this study was to evaluate the potential reorganization in the sensorimotor cortex in SCI patients after Body Weight Supported Treadmill Training (BWSTT) associated with conventional motor rehabilitation. METHOD: Seven SCI patients with ASIA C and D participated in this study. They were submitted to a motor-task functional magnetic resonance imaging study (fMRI) before and after the rehabilitation treatment.

Motor Rehabilitation The rehabilitation program consisted of five hours per day of the treatment, five times a week for two months (40 sessions). The treatment protocol included kinesiotherapy (passive and active mobilizations, muscle lengthening), BWSTT, bicycle, manual therapy (with and without assistance of a mechanical device) and daily life activities training. The duration of each event was approximately one hour.

The initial time of the treatment on the BWSTT was 20 minutes, and the modification of the time depended on the endurance and capability of each patient. Training charge began at 40% body weight supported and at treadmill speeds of at least 1.5 km per hour. The treadmill speed was increased progressively to 2.5 km per hour, and the level of weight supported was adjusted within sessions to achieve knee extension. Table 1 shows the initial and final training parameters for each patient.

To assess functional ability, each patient was assigned a Walking Index for Spinal Cord Injury (WISCI II) (Ditunno and Ditunno 2001) (Table 1), Manual Muscle Testing and Functional Independence Measure (FIM) before and after the treatment period (Table 2).

fMRI Protocol The patients were scanned using a 3T Siemens TIM TRIO system before and after the rehabilitation treatment protocol. The fMRI protocol consisted of a two motor-task fMRI series with the following technical parameters: TE=29 ms; TR=2000 ms, number of slices=36, matrix=128x128, FOV=240 mm, slice thickness= 3 mm; and FA=90º. The patients were given instructions for the motor task they were to perform prior to each scan. The patients were positioned with comfortable pads below the knee to allow the active movement of the ankle. During the functional imaging, each motor activation period was signalled with a short instruction to the subject through the headphones.

The motor task for the fMRI exploration of the foot was the ankle plantar flexion and toe flexion. The block design began with a 20 second rest period, followed by 30 second periods of movement of the right foot, 30 seconds of the left foot movement and 10 seconds of rest. This cycle was repeated three times. The movement performance was controlled visually by two examiners to monitor any movement or apparent change during the rest periods of the non-moving limbs. The patients were oriented to perform the motor task through a voluntary movement. Approximately 15 repetitions were made in each period.

Finally, a T1-weighted structural image of the whole brain was acquired to a perfect localization of activation areas. Acquisition parameters of the structural image were TE= 2.98 ms; TR= 2300 ms; T1= 900 ms; matrix= 256x256; FA= 9º; and T voxel=1x1x1 m3.

Inclusion Criteria:

• Patients with ASIA C or D (Table 1) and the ability to cooperate with instructions of the motor task required. The motor task was to perform ankle plantar flexion, toe flexion and fingers and thumbs movement of both sides; they also had to be able to fulfill the general conditions necessary for an MRI exploration.

Exclusion Criteria:

• Patients with uncontrollable spasticity-induced body movements.