Objective
To investigate the clinical usefulness of the motor cortex mapping using transcranial magnetic stimulation (TMS) in stroke patients.

Method
Five stroke patients were studied. A piece of cloth which marked at 1 cm interval was fixed on the patient's head. Motor cortex mapping for abductor pollicis brevis muscles (APB) was performed with a butterfly coil or with a round coil if motor cortex mapping was impossible.

Results
Ipsilateral motor pathways were discovered from the unaffected motor cortex to the affected APB in patient 1. This patient showed delayed latency and low amplitude of ipsilateral motor evoked potentials (MEP) that seems to be evoked from the descending motor pathway rather than the corticospinal tract. In patient 2 and 3, contralateral motor pathways traveled from the affected hemisphere to the affected APB. The short latency and high amplitude of MEPs seems to be attributed to the corticospinal tract. In patient 4, no MEP was evoked by any hemisphere or magnetic stimulator. We believe that the affected APB had no motor pathway, and it correlated well with the poor motor function of her hand. In patient 5, contralateral pathways from the affected hemisphere to the affected APB were present. In this patient, the parameters of the motor cortex map such as the amplitude of MEP, the number of MEP evoked site, and the excitatory threshold were improved after 2 months, which correlated well with clinical improvement.

Conclusion
Motor cortex mapping using TMS is clinically useful for the evaluation of the characteristics of motor pathways and the change of motor cortex excitability in stroke patients.