The study is designed to explore the effectiveness of non-invasive cerebellar stimulation to enhance motor plasticity in the cortex after stroke. The investigators have shown that the human cerebellum is strongly activated during the act of swallowing and when stimulated with single TMS pulses can strongly facilitate the corticobulbar projection to the pharynx in humans. More recently the investigators have identified the most relevant frequency of stimulation of the cerebellum that can produce longer term excitation in the human swallowing motor system. The investigators therefore believe that the potential for cerebellar stimulation in improving swallowing is much greater than other methods for two reasons. Firstly, previous work has shown that unlike successful recovery of hand/arm function which relies on restoring activity in the stroke hemisphere, recovery of swallowing function relies on increased excitability in intact projections from the non-stroke hemisphere. The investigators believe that methods that can enhance these undamaged pathways have a greater chance of inducing recovery in the human swallowing system in unilateral stroke. Additionally cerebellar stimulation produces very high levels of corticobulbar excitation it may also have the advantage of improving dysphagia in posterior fossa strokes. Second, the human cerebellum is relatively easy to target and stimulate and has reduced risk of inducing unwanted effects (such as seizures) which as a consequence makes cerebellar stimulation a more pragmatic method for delivering therapeutic neurorehabilitation to dysphagic stroke patients compared to other more complex/riskier methods.

A final factor is that the investigators have developed a "virtual lesion" model of swallowing dysfunction in healthy volunteers which can be reversed quite successfully with other neuro-stimulation protocols. The investigators can therefore use this model to test the effectiveness of cerebellar stimulation protocols (ipsilateral and contralateral cerebellar sites) before choosing the most effective side to apply stimulation in a proof of principle trial/study in a small group of sub-acute dysphagic patients.

The hypotheses are that cerebellar TMS will:

i. Reverse the brain inhibition and behavioural dysfunction following a virtual lesion model of disrupted swallowing in healthy brain (phase 1); ii. Reduce the degree of aspiration in acute dysphagia after a stroke (phase 2).

Protocols:

Hypothesis 1:

Cortical excitability (in both dominant and non-dominant swallowing cortex) to TMS and swallowing behaviour assessed with a swallowing reaction time task will be measured at baseline. Thereafter, the virtual lesion paradigm will be applied to the dominant swallowing projection followed by the cerebellar intervention (10Hz, 250 pulses). Cerebellar stimulation will be applied (in a randomised fashion) on separate occasions to both the contralateral and ipsilateral cerebellar hemispheres, immediately after the virtual lesion. The investigators have found that placement using known reference landmarks are equivalent to using neuronavigation, so the latter will not be employed. Repeat measurements of cortical excitability and swallowing behaviour will then be performed and analysed against baseline data and a sham cerebellar paradigm. Differences in cortical excitability and swallowing responses will thus be an indication of which cerebellar region (contralesional/ipsilesional) can influence excitability and reverse any behavioural changes most effectively.

Hypothesis 2:

I. Dysphagic stroke patients (n=24) recruited over a 9 month period within 14 days of stroke ictus will have their swallowing assessed by videofluoroscopy before and after receiving either the real or sham cerebellar stimulation (10Hz, 250 pulses) as determined from phase 1. As before, placement of the coil for cerebellar stimulation will be performed using landmarks already established and validated in the work from question 1. Patients will be intubated with the pharyngeal EMG catheter for pre and post recordings of pharyngeal motor evoked potentials (MEPs) to cortical TMS. Power calculations, based on a similar study of pharyngeal stimulation indicated that the investigators would need 12 patients per group to achieve a statistical power of 80% at a 5% significance level to detect changes in the primary outcome measure of aspiration. In this study, the investigators will examine if the chosen site of cerebellar stimulation from phase 1 can induce short-term changes in brain function (pharyngeal MEPs) and swallowing function (videofluoroscopy) compared to sham interventions and baseline measurement, up to 1 hour after the intervention.

II. Following stage I, once it is established that cerebellar stimulation can alter brain and swallowing functions in stroke in the short-term, it will need to be appraised for clinical feasibility, dose response, and longer term efficacy. Dysphagic stroke patients (n=48, 16 patients per group) admitted to the stroke unit and identified by videofluoroscopy will be recruited over the next 21 months. Since it is not known how the cerebellar stimulation method should be delivered to patients, a dose ranging treatment trial will be utilised, randomising patients to one of three groups (A-C), low level stimulation, high level stimulation and sham stimulation. From our previous work with pharyngeal stimulation, the investigators propose that group A will receive stimulation once per day for 3 days. Group B will receive stimulation twice per day for 5 days. Group C will receive sham stimulation (delivered as in protocol I) twice a day for 5 days. Groups A and B will receive stimulation at the optimal site (10Hz, 250 pulses) found in question 1, but all groups will also receive standard speech and language therapy.

Assessments:

Swallowing before and after stimulation will be evaluated using videofluoroscopy, at 1 hour for protocol I. For protocol II the investigators propose both videofluoroscopy and functional scores (Functional oral ingestion scale (FOIS), the dysphagia severity rating scale (DSRS) including feeding status and modified rankin scale (mRS)) at baseline and at 2 weeks.

Inclusion Criteria:

• Patients aged 18 years and over
• All patients with an acute anterior or posterior cerebral circulation stroke within 6 weeks of symptom onset.

Exclusion Criteria:

• Advanced dementia
• Previous history of dysphagia
• Patients judged to be clinically unstable
• Presence of implanted cardiac pacemaker or defibrillator
• Any severe chronic medical condition that compromises cardiac or respiratory status
• Patients with acute lower respiratory tract infections requiring antibiotic treatment.

No identifiable individual participant information will be shared outside the immediate research team. However, identifiable individual participant information will be made available to the Medical research council and the University of Manchester if required for the purposes of auditing the study.