Alzheimer’s disease (AD) has very limited treatment options and therapies to prevent or reverse neurodegeneration remain elusive. Deep brain stimulation (DBS), whereby high-frequency pulses of electricity are delivered continuously to a specific part of the brain, has been trialled as an experimental treatment for AD. In AD patients, DBS has been targeted at a group of fibres called the fornix, which carry signals to and from the hippocampus; a crucial region for encoding and recalling memory. Fornix-DBS has been shown to be safe and modulate hippocampal markers of synaptic plasticity, but is not reliably effective when delivered using the standard, continuous high frequency pulses that are effective in other disorders. We propose that a different approach is required in AD, whereby closed-loop stimulation is used to restore key activities disrupted by the primary degenerative process. Closed-loop stimulation uses real-time tracking of specific aspects of neural signals to control when electrical pulses are delivered. DBS delivered in this way would act like a network prosthesis, compensating for circuit-level activities that the diseased memory circuits can no longer provide. To this end, we have developed and utilised methods of real-time tracking of hippocampal oscillations (theta/sharp-wave ripples) to control the timing of electrical stimulation of the fornix of freely moving rats, using lightweight algorithms that can readily be employed in human devices. Our results demonstrate that such approaches allow the precise modulation of the synchronisation of hippocampal neurons towards states aligned with the facilitation of memory performance, providing a rationale for a closed-loop approach to DBS for AD.