Stroboscopic light stimulation is a sensory practice that uses rhythmic flickering light to induce visual phenomena and altered states, with its contemporary form developed by Brion Gysin and Ian Sommerville in 1959. It is among the few non-pharmacological methods under clinical investigation for Alzheimer's disease. It alters consciousness primarily through photic driving — rhythmic visual input that synchronizes occipital cortex activity to the stimulus frequency.
The practitioner sits or lies with eyes closed at roughly arm's distance from a pulsing light source — historically a slotted cylinder rotating before an incandescent bulb, today an LED-and-halogen array driven by programmable frequency sequences. The instruction is simply to keep the eyes closed and let imagery arise. Sessions range from 10–20 minutes in introductory protocols to 40 minutes in extended commercial-device formats.
A typical session lasts 20 to 40 minutes. At introductory exposures, geometric phosphenes appear within the first few minutes — concentric rings, grids, spirals, and tunneling motion toward the visual center, often in saturated complementary colors. With sustained exposure, imagery deepens toward figural content (faces, landscapes, narrative fragments) and a hypnagogic, near-sleep phenomenology. Headache, eye strain, anxiety, and disorientation are common, particularly in migraine-prone individuals.
The primary risk is seizure in photosensitive individuals — photoparoxysmal responses occur in roughly 0.3–3% of the general population, with the 15–25 Hz band most provocative.[1] Photosensitive epilepsy, frequent migraine with aura, and active psychotic symptoms are contraindications. Outside the 40 Hz Alzheimer's research strand at MIT, no controlled trials have evaluated wellness applications of commercial stroboscopic devices; phenomenology is documented but mechanism beyond primary visual entrainment remains uncharacterized.[2]
History & Tradition
Origins
The systematic observation of flicker-induced visual phenomena predates modern neuroscience. The Czech physiologist Jan Evangelista Purkinje documented phosphene-like geometric hallucinations in the early 1800s when he closed his eyes and waved his hand rapidly before his face in sunlight.[3] This early observation eventually entered neurophysiology via the development of clinical electroencephalography in the mid-twentieth century.
W. Grey Walter, Director of the Burden Neurological Institute in Bristol, rediscovered and systematically mapped the phenomenon in the 1940s and early 1950s. Walter applied intermittent photic stimulation as a diagnostic tool in EEG research and noted that subjects reported vivid, unsolicited visual hallucinations at specific flicker frequencies. He observed a correspondence between stimulus frequency and the alpha oscillation band (roughly 8–12 Hz), and characterized geometric, color-saturated, and morphing imagery as the typical response at those frequencies.[4] His popular-science book The Living Brain (1953) introduced photic-driving research to a wide audience and seeded the cultural transmission of the practice.[2]
Transmission and secularization
On 21 December 1958, the artist and writer Brion Gysin was riding a bus through an avenue of trees in the south of France when afternoon sunlight flickered through the trunks at an alpha-band frequency, producing what he described as an overwhelming flood of patterns in supernatural colors. Gysin had been reading Walter's work and immediately recognized the experience as a replicable neurological phenomenon.[3] In collaboration with Cambridge mathematician Ian Sommerville he built the Dreamachine: a slotted cardboard cylinder mounted on a 78 rpm gramophone turntable with a 100-watt incandescent bulb suspended at its center, producing roughly 8–13 flashes per second when viewed with closed eyes at arm's distance.
The Dreamachine entered the Beat Generation through William S. Burroughs, Gysin's collaborator at the Beat Hotel in Paris, who adopted it as a model for disrupting habitual thought and language. Despite a granted patent and overtures to Philips and other manufacturers, commercial production did not proceed — manufacturer concerns about photosensitive-epilepsy liability blocked the consumer launch.[2]
Contemporary context
The practice remained primarily an avant-garde and counterculture phenomenon through the 1970s–90s. The engineering shift from analog turntable devices to digitally controlled stroboscopes in the early 2000s enabled precise, programmable, variable-frequency flicker and set the conditions for a second wave of device development. Current commercial devices include the PandoraStar (a ceiling-mounted 12-LED array), the Roxiva, and the Lucia N°03 — the latter developed in Austria by neurologist Dirk Proeckl and clinical psychologist Engelbert Winkler, and sold through a manufacturer-run "Light Attendant" training network.[5]
A separate scientific lineage runs through the MIT Picower Institute, where Li-Huei Tsai's group has developed the GENUS (Gamma ENtrainment Using Sensory Stimuli) protocol — 40 Hz audiovisual flicker as a candidate intervention for Alzheimer's disease. Preclinical work in mouse models published in 2016 reported substantial reductions in amyloid-beta load in the visual cortex after one hour of 40 Hz exposure,[6] and a Phase 3 randomized trial through the MIT spinoff Cognito Therapeutics is now underway. Proponents of wellness devices frequently invoke the GENUS work to legitimize their products, though the wellness applications use different parameters and have no comparable evidence base.[2]
The Practice
Technique
The practice has no ancient lineage. What follows describes the contemporary device-based form, which has converged on a common protocol despite differences in hardware.
Postural foundation
The practitioner sits in a chair or lies supine at roughly arm's distance from the light source. Devices vary: the historical Dreamachine is a slotted cylinder rotating before an incandescent bulb; the Lucia N°03 uses an array of eight LEDs surrounding a central halogen lamp; the PandoraStar is a ceiling-mounted 12-LED array; the Roxiva is a commercial stroboscopic lamp on a stand. In every case, the eyes are closed throughout the session. The light penetrates the eyelids, which act as a diffuse filter, producing a soft, pulsating red-orange luminance on the retina rather than discrete flashes.
Core method
The method has no attentional instruction in the meditative sense. The practitioner closes the eyes, faces the light, and allows imagery to arise. The device delivers a programmed sequence of flicker frequencies — typically beginning at lower frequencies (theta range, 4–7 Hz), passing through the alpha band (8–12 Hz), and in longer protocols ascending into beta and gamma ranges. Sequences are pre-programmed on the device; the practitioner does not control them in real time. The physiological response is involuntary: the retina and primary visual cortex synchronize to the stimulus frequency within seconds, producing steady-state visual evoked potentials (SSVEPs) and, at certain frequencies, broader entrainment of ongoing EEG oscillations.[2] The most common departure from the method is opening the eyes during peak imagery, which collapses the closed-eye visual field and ends the hypnagogic sequence — the imagery is generated by the pattern of light through the eyelids, not by direct visual perception of the device.
Progressive structure
There is no skill progression in the meditative sense. What changes with repeated sessions is habituation: first-time users typically spend the opening minutes adjusting to the novel sensory environment before imagery deepens, while experienced users report faster onset. The two intensity tiers reflect session length rather than developmental milestones.
| introductory | 10–20 minutes; geometric phosphenes; eyes-closed exposure |
| established | 20–40 minutes; sustained variable-frequency sequence; hypnagogic imagery |
| clinical (GENUS) | 60 minutes; fixed 40 Hz audiovisual; daily for AD research |
Variations and lineages
The Dreamachine is the original form — analog, fixed-frequency (roughly 8–13 Hz, determined by the cylinder geometry and turntable speed), and conceptually tied to its Beat Generation context. Lucia N°03, PandoraStar, and Roxiva are digitally controlled successors marketed in the wellness sector; they differ in lamp configuration and pre-programmed sequences but share the closed-eye protocol. The GENUS protocol is a separate clinical lineage: a fixed 40 Hz audiovisual stimulus targeting gamma-band entrainment, delivered for one hour daily over months in trial populations with mild Alzheimer's dementia.[7] The wellness and clinical lineages share a common neurophysiological substrate (photic driving) but differ substantially in parameters, supervision, and evidence base.
Setting & Preparation
A dim or darkened room, a chair or recliner, and the device itself constitute the full setup for most contemporary forms. Sessions are typically conducted with the practitioner alone or in the presence of a trained operator; commercial wellness sessions in the Lucia N°03 lineage take place at practitioner-run studios where a Light Attendant manages the device and screens for contraindications.[5] Group sessions exist (the 2022 public Dreamachine installation in the United Kingdom was designed for groups of up to 30 participants in a shared dome) but are uncommon. No specific clothing, posture, or breathing instruction is required.
Conduct the session in a quiet room with the practitioner seated or reclining at the manufacturer-specified distance from the device. A trained operator or a second person should be physically present for any first session, and for any session involving an unfamiliar device, frequency, or protocol — both to manage the device and to respond if the practitioner becomes distressed or shows seizure signs. Sessions should not be conducted alone for first-time users. The setting should permit immediate access to a quiet, dim recovery space; some users find the post-session minutes disorienting. Reduce ambient light during the session — bright ambient light competes with the stroboscopic signal and reduces both intended phenomenology and the subjective sense of immersion. Avoid scheduling sessions when sleep-deprived, recently intoxicated, or under significant acute stress, all of which may amplify both seizure susceptibility and adverse psychological responses.
Guidelines
- Screen for photosensitive epilepsy before any session — review personal and family history, and use a validated pre-participation tool such as the 4SQ where available.[8]
- Do not attempt if you have a personal history of seizure disorder, active migraine with aura, current psychosis, or recent head injury.[1]
- Limit a first session to 10–15 minutes; do not begin with a 40-minute protocol.
- Have a second person physically present for any first session or new protocol; do not conduct first sessions alone.
- Stop immediately at the first sign of involuntary muscle jerking, loss of awareness, severe headache, or rising panic — open your eyes and signal the operator.
- Avoid combining with psychoactive substances, particularly those that lower seizure threshold (alcohol withdrawal, stimulants, tramadol).
- Allow 15–30 minutes of quiet integration after the session before driving, operating equipment, or returning to demanding cognitive tasks.
Mechanism
Steady-state visual evoked potentials
The most robust and uncontested mechanism of stroboscopic stimulation is the steady-state visual evoked potential (SSVEP) — an oscillatory EEG response in occipital cortex that occurs at the frequency of the driving stimulus and its harmonics. SSVEPs are reliably elicited by any periodic visual input between roughly 1 and 100 Hz, have high signal-to-noise ratios, and are routinely used in brain-computer interface research. The SSVEP reflects sensory synchronization of primary visual cortex to the external stimulus; it is not, on its own, evidence of a systemic change in brain state.[2]
Photic driving and broader entrainment
Photic driving — the entrainment of ongoing EEG oscillations beyond the primary occipital SSVEP — is less uniformly observed. Coherence analysis has documented alpha-rhythm entrainment by the second harmonic of the stimulation frequency in approximately 75% of subjects, supporting the photic-driving mechanism but also indicating substantial individual variability.[9] The degree to which entrainment propagates beyond visual cortex is parameter-dependent: a systematic human study found that red light and 400 cd/m² white light flickering at 34–38 Hz produced gamma entrainment that spread beyond the occipital lobe, while standard 40 Hz parameters produced more localized responses.[10]
Frequency-specific phenomenology
Geometric visual hallucinations cluster in the alpha band (8–12 Hz), with rotational and zooming-toward-center motion the dominant reported phenomenology.[2] The mechanism for these complex hallucinations is not fully characterized — entrainment alone does not explain why specific geometric forms (concentric rings, spirals, tunnels) recur across individuals, and current hypotheses invoke the columnar architecture of primary visual cortex without a settled account.
Gamma entrainment and Alzheimer's research
A separate strand of mechanistic research targets 40 Hz gamma stimulation specifically. The preclinical foundation is the Iaccarino et al. demonstration that one hour of 40 Hz visual flicker reduced amyloid-beta in the visual cortex of transgenic AD mouse models by approximately 40–50%, with associated microglial activation.[6] The proposed mechanism is that gamma entrainment recruits microglia and adjusts inflammatory signaling in ways that promote amyloid clearance, though the precise pathway remains under investigation. Human extrapolation is preliminary: the Phase 2A GENUS pilot in 15 patients with mild probable AD found cognitive measures preserved relative to a matched registry cohort over three months of daily stimulation,[7] and an open-label extension reported preservation through two-year follow-up with sex-stratified heterogeneity.[11]
Wellness-context claims
Wellness device manufacturers (Lucia N°03, PandoraStar, Roxiva) routinely invoke "brainwave entrainment," "neuroplasticity," and "gamma coherence" as mechanisms of action. The underlying SSVEP is real and the alpha-frequency phenomenology is documented; broader claims about neuroplasticity, autonomic regulation, or therapeutic remodeling outside the AD gamma research strand are not empirically supported and should be attributed to proponents rather than asserted in the encyclopedia's voice. No fMRI study of wellness-oriented stroboscopic sessions has been published, and no peer-reviewed evidence characterizes systemic changes induced by Lucia N°03, Dreamachine, or PandoraStar protocols specifically.
Why does alpha-band stroboscopic exposure preferentially produce centrally-organized geometric imagery — concentric rings, spirals, tunneling motion — rather than arbitrary visual content? The columnar architecture of primary visual cortex is the leading candidate, but no model accounts for both the cross-individual consistency and the strong frequency dependence.[2]
The Open Questions disclosure renders within the Mechanism section above.
Subjective Effects
Stroboscopic exposure produces a characteristic sequence with closed eyes. The opening minutes are dominated by simple phosphenes — flickering luminance, complementary-color afterimages, and unstructured patterning at the rhythm of the stimulus. Within several minutes the imagery resolves into geometric forms: concentric rings, grids, lattices, and tunnels moving rotationally or zooming toward the visual center, frequently in saturated colors that the user did not consciously generate.[2] Practitioners commonly report that the imagery feels autonomous and self-organizing rather than imagined.
At longer exposures (20+ minutes) imagery in many users transitions toward figural content — fragments of faces, landscapes, and brief narrative scenes — accompanied by dissociation from body and environment and a near-sleep, hypnagogic quality. Individual variation is high. A meaningful fraction of users report little beyond mild phosphene patterning and find the experience underwhelming; others report intense altered states comparable in subjective dimensions (though not in pharmacology) to low-to-moderate doses of classical psychedelics.[2] Unsought effects — anxiety, derealization, headache, eye strain, and disorientation — are common, particularly in migraine-prone or anxiety-prone individuals.
▸▾Effect Grid
Eyes-closed exposure to rhythmic flicker; geometric imagery arises.
Safety
Adverse Effects
Photosensitive seizure risk
Photosensitivity — an abnormal EEG response (photoparoxysmal response, PPR) to intermittent photic stimulation — occurs in approximately 0.3–3% of the general population, with a substantially higher prevalence in syndromes such as Juvenile Myoclonic Epilepsy and Jeavon's syndrome (epilepsy with eyelid myoclonia).[1] The estimated incidence of clinically manifest photosensitive seizures is roughly 1 per 4,000 individuals in the 5–24 year age range. The provocative frequency range spans 3–65 Hz, with 15–25 Hz the most reliable seizure trigger; red light is more potent than white or blue, and binocular large-field stimulation increases risk relative to small-field or monocular exposure.[1] This was the documented liability that blocked the original commercial Dreamachine launch and remains the single most important medical risk of the practice.
Non-epileptic adverse events
A 2025 multi-laboratory safety survey integrating 24 studies and 1,070 participants documented 20 minor adverse events across laboratory stroboscopic exposures — roughly 18.7 per 1,000 participants — with no severe incidents requiring medical intervention.[8] (The survey is a preprint as of 2026-06-09 and has not yet completed peer review.) Documented non-epileptic responses include migraine and photophobia, anxiety, eye strain, headache, and discomfort. The same group developed the 4SQ (Stroboscopic Safety Screening Questionnaire), a brief pre-participation tool to identify established stroboscopic sensitivities.
Migraine
Flickering light is among the most commonly reported migraine triggers. [UNCITED — verify: up to 50% of migraineurs identify visual triggers including flicker, with 10–20 Hz most aversive]. 64% of migraineurs versus 19% of controls report adverse reactions to flickering stimuli including headache, eye fatigue, and disorientation [UNCITED — verify: TheraSpecs / migraine prevalence literature citation needed]. Wellness stroboscopic sessions should be considered contraindicated for individuals with active or frequent migraine with aura.
Psychological adverse effects
No published systematic survey of psychological adverse effects specific to wellness-context stroboscopic sessions exists. Practitioner-community reports describe anxiety, derealization, dissociation, and reactivation of difficult memories in some sessions. These domains overlap with the documented meditation-adverse-effects categories surveyed by adverse-experience programs such as Cheetah House, and are considered relevant here by analogy rather than direct evidence.[12]
Contraindications
Hard contraindications
- Personal history of epilepsy, seizure disorder, or known photosensitive seizure response.[1]
- Active or frequent migraine with aura, especially with documented flicker sensitivity.
- Current acute psychotic symptoms; active untreated severe mental illness.
- Recent significant head injury or post-concussive syndrome.
Soft contraindications and cautions
- Family history of epilepsy or photosensitive seizure response, particularly in syndromes such as Juvenile Myoclonic Epilepsy.[1]
- Pregnancy — listed as a caution by most device operators; no peer-reviewed evidence supports or excludes risk, so the conservative position is to defer until after pregnancy.
- History of dissociative disorders or significant trauma; the hypnagogic and dissociative phenomenology may surface difficult content.
- Acute or recent alcohol or stimulant intoxication; both can lower seizure threshold and compound photosensitive risk.
The Schwartzman et al. 4SQ screening questionnaire is the most concrete pre-participation tool currently available for laboratory and practitioner use.[8]
Substance Interactions
No controlled studies have examined interactions between stroboscopic light stimulation and psychoactive substances. The following are theoretical risks supported by mechanistic reasoning rather than empirical data.
- Classical psychedelics (psilocybin, LSD, DMT, mescaline): 5-HT2A activation enhances visual cortex excitability and may intensify both the geometric phenomenology and any underlying seizure susceptibility. Commercial device marketing frequently draws psychedelic analogies, but no human studies have characterized the combined exposure.
- Cannabis (THC-dominant): may intensify visual phenomenology and dissociation; theoretical only.
- Stimulants (amphetamines, cocaine, high-dose caffeine): elevate seizure threshold risk; sympathetic activation may compound anxiety responses to flicker.
- Alcohol, particularly in withdrawal: substantially lowers seizure threshold; combined exposure during acute withdrawal carries serious seizure risk.
- Tramadol and other seizure-threshold-lowering medications (some antihistamines, bupropion at high doses): increase the underlying photosensitive seizure risk.
- Dissociatives (ketamine, DXM): theoretical risk of compounding dissociative/derealization responses.
The absence of empirical interaction data is itself a relevant safety finding. Practitioners should not assume the combination is safe simply because no controlled harm has been documented.
Habituation
No evidence indicates that stroboscopic light stimulation produces tolerance, dependence, or compulsive use in the way pharmacological substances or some contemplative practices can. Practitioner-community reports describe enthusiasm and frequent sessions in regular users, but no published case series or survey documents loss-of-control patterns. The practice is bounded by session length and device availability rather than by an internal escalation mechanism. Habituation, where it occurs, manifests as diminished novelty rather than diminished response — experienced users continue to report SSVEP-driven phenomenology in subsequent sessions, but the experience may feel less remarkable over time.
If Distress Arises
If imagery becomes overwhelming, intrusive, or accompanied by rising anxiety, the response is to open the eyes — this collapses the closed-eye visual field and ends the hypnagogic sequence within seconds. There is no need to wait out the session; the practitioner has full agency to stop at any point. If you experience early seizure warning signs — involuntary muscle twitching, loss of orientation, a sudden "absent" feeling, or visual disturbance that persists with eyes open — the session must end immediately and a second person should be alerted. Mild symptoms during a session (headache, eye strain, queasiness, mild anxiety) generally resolve within minutes of stopping; severe symptoms (sustained involuntary movements, loss of consciousness, persistent disorientation) require emergency medical attention. The same rules apply whether the session is run by a practitioner-network operator or self-administered with a personal device.
Recovery & Support
Mild post-session effects — eye strain, mild headache, residual after-imagery, transient disorientation — typically resolve within 30 to 60 minutes of quiet rest in normal light conditions. Hydration, a low-stimulus environment, and avoidance of further bright flickering light (screens, fluorescent lighting) shorten recovery for most users.
Persistent symptoms warrant medical evaluation. Seek medical attention promptly for any post-session seizure, sustained visual disturbance lasting more than several hours, severe headache that does not respond to standard analgesics, or new neurological symptoms (weakness, numbness, speech changes). Sessions that surface significant emotional content or dissociative symptoms warrant follow-up with a trauma-informed mental health professional; while no stroboscopic-specific clinical referral network exists, organizations such as Cheetah House maintain a directory of practitioners experienced with contemplative- and altered-state-related adverse experiences.[12] Compulsive-use patterns are not documented for this practice; recovery and support content emphasizes acute symptom management rather than long-term cessation support.
Recovery resources, crisis lines, and support pathways → View all
Progression
No formal progression model has been published for wellness-oriented stroboscopic light practice. The two intensity tiers in current commercial protocols (10–20 minute introductory; 20–40 minute established) reflect session length and device-program selection rather than developmental milestones. Practitioners generally report faster onset of imagery and reduced novelty across repeated sessions, but no longitudinal study has tracked phenomenological or neural changes across a course of practice. Phenomenological progression within a single session has been informally characterized as: simple phosphenes in the opening minutes, geometric imagery (concentric rings, spirals, tunneling motion) over the following 5–15 minutes, and figural/narrative content emerging in extended sessions for some users.[2] Individual variation is substantial; a meaningful fraction of users do not progress beyond geometric imagery regardless of session length.
Evidence Summary
The evidence base for stroboscopic light stimulation divides cleanly into two non-overlapping bodies of work.
The 40 Hz gamma stimulation strand — anchored at the MIT Picower Institute as the GENUS protocol — has the strongest evidence by a wide margin. Preclinical demonstration in AD mouse models established that 40 Hz visual flicker reduces amyloid-beta in the visual cortex.[6] Human Phase 1 feasibility and Phase 2A pilot data in mild probable AD found the protocol well-tolerated, with several cognitive measures preserved relative to a matched registry cohort over three months of daily stimulation.[7] A 2025 open-label extension reported preservation at two-year follow-up, with sex-stratified heterogeneity in response.[11] A Phase 3 randomized trial through Cognito Therapeutics is now underway. A narrative review of gamma multisensory stimulation in AD concluded that evidence is promising but preliminary pending Phase 3 data.[13]
The wellness-device strand — Lucia N°03, PandoraStar, Roxiva, Dreamachine — has no published RCT evidence for any health, wellness, or psychological outcome. The available literature is limited to (1) the well-characterized SSVEP and photic-driving response,[2] which is a primary visual cortex phenomenon and does not establish broader effect, (2) a parametric human study of optimal gamma entrainment conditions,[10] and (3) a multi-laboratory safety preprint.[8] Hewitt et al. (2025) is the most comprehensive recent synthesis of phenomenology and historiography but is a narrative review, not a systematic clinical synthesis.[2]
Limitations
Wellness applications lack controlled trials
No randomized controlled trial has been conducted on any wellness device (Lucia N°03, PandoraStar, Roxiva, contemporary Dreamachine). No fMRI study has characterized neural responses to commercial wellness protocols specifically. No published study has measured neuroendocrine markers (cortisol, melatonin), heart-rate variability, or sleep outcomes following commercial-device sessions. Claims of neuroplasticity, autonomic regulation, mood improvement, or therapeutic effect in commercial marketing exceed what the literature supports and should be attributed to manufacturers rather than to the evidence base.
Safety data are thin
The most comprehensive safety dataset (Schwartzman et al., 2025) is a preprint integrating laboratory studies rather than wellness sessions; its findings should be characterized accordingly until peer review is complete.[8] No systematic survey has characterized psychological adverse-effect rates in wellness contexts. The photosensitive seizure literature is well-established but predates contemporary commercial devices and does not directly address their specific parameters.[1]
AD evidence is preliminary
GENUS human trials have been small (n=15 in Phase 2A), open-label or feasibility-stage, and conducted in a single research group at MIT. Replication outside this lineage is limited. The open-label extension showed sex-stratified heterogeneity that constrains interpretation: benefits were reported in three female participants and absent in two male participants with early-onset forms.[11] Pending Phase 3 randomized trial results, the GENUS evidence base should be characterized as promising but not established.
Manufacturer-to-evidence framing gap
Wellness device marketing frequently invokes the GENUS literature, "brainwave entrainment," and "neuroplasticity" as mechanism explanations. The SSVEP and primary alpha-band photic driving are well-established neural responses; broader claims that link these responses to therapeutic outcomes for stress, mood, sleep, cognition, or chronic disease in commercial wellness contexts are not empirically supported. This gap between scientific marketing language and the state of the evidence is the dominant limitation of the wellness-context literature.
references
- [1]
- [2]
- [3] ^abMudrik, L. and Zeki, S. (2009) From stroboscope to dream machine: a history of flicker-induced hallucinations — European Neurology PMID:19729929
- [4]
- [5]
- [6] ^abcIaccarino, H.F. and Singer, A.C. and Martorell, A.J. and Rudenko, A. and Gao, F. and Gillingham, T.Z. and Bhanner, H. and Mungenast, A.E. and Tsai, L.H. (2016) Gamma frequency entrainment attenuates amyloid load and modifies microglia — Nature doi:10.1038/nature20587