The dying brain does appear to generate bursts of electrical activity linked to memory and conscious experience, though calling it a clean “playback” oversimplifies what researchers have found. In a small but striking 2023 study, two of four comatose patients showed a rapid surge of gamma brain waves after their ventilators were removed, the same type of oscillations the brain uses during memory recall and perception in healthy people. Animal studies have shown a similar phenomenon. Whether this activity produces the “life flashing before your eyes” that people describe remains one of the hardest questions in neuroscience, because the people whose brains do this almost never survive to tell us about it.
The Gamma Surge in the Dying Brain
The most direct evidence comes from electroencephalogram (EEG) recordings made during the final moments of life. In 2013, a research team at the University of Michigan documented what happens in rat brains immediately after cardiac arrest. Within the first 30 seconds, the rats’ brains produced a global surge of highly synchronized gamma oscillations, a fast frequency of brain activity associated with waking consciousness. These gamma waves were more coherent and more organized than anything seen during normal wakefulness, and they showed strong connectivity patterns running from the front to the back of the brain.1PubMed Central. Surge of neurophysiological coherence and connectivity in the dying brain
A decade later, the same research group published findings from human patients. They analyzed EEG and heart rhythm data from four comatose patients who were dying after withdrawal of life support. Two of the four showed a marked surge of gamma power after their ventilators were removed. The gamma waves coupled tightly with slower brain rhythms, and connectivity between the brain’s two hemispheres increased sharply in the gamma band.2PubMed Central. Surge of neurophysiological coupling and connectivity of gamma oscillations in the dying human brain The fact that only half the patients showed the surge is itself interesting. It suggests this is not an automatic shutdown process that happens in every dying brain, and researchers still do not know what distinguishes a brain that surges from one that does not.
Why Gamma Waves Are Relevant to Memory
The reason these dying-brain findings spark so much excitement is that gamma oscillations are not random electrical noise. In living, healthy brains, gamma rhythms play a specific role in binding together the components of a memory or a scene. They help the cortex stitch sensory features into a unified perception, and they help the hippocampus (the brain’s primary memory hub) combine contextual and sensory information into what we experience as a specific memory of a specific moment.3PubMed Central. Functional role of gamma and theta oscillations in episodic memory
Memory consolidation also depends on a related process involving fast electrical events called sharp-wave ripples, which replay compressed fragments of recent experience during sleep and quiet rest. These ripples coordinate with gamma activity to transfer hippocampal memories to long-term storage in the cortex, and disrupting them impairs memory formation.4PubMed Central. Hippocampal sharp wave-ripple: A cognitive biomarker for episodic memory and planning So when a dying brain produces an organized burst of gamma activity coupled to slower rhythms, it is producing the same electrical signature that the brain normally uses when it is actively processing or retrieving memories. That does not prove the dying person is reliving their past, but it does mean the hardware is running in a mode consistent with conscious experience.
What Survivors of Cardiac Arrest Actually Report
Most of what we know about the subjective side of dying comes from people who were clinically dead and came back. The AWARE (AWAreness during REsuscitation) studies are the largest systematic attempts to document what happens in people’s minds during cardiac arrest. The original AWARE study, published in 2014, was a four-year multicenter investigation that used quantitative and qualitative interviews to assess whether patients had awareness during the period when their hearts had stopped.5PubMed. AWARE-AWAreness during REsuscitation-a prospective study
The follow-up, AWARE-II, published in 2023, narrowed the focus and added EEG and brain oxygen monitoring. Of 567 in-hospital cardiac arrests, 53 patients survived and 28 completed interviews. Among those 28, about 40 percent reported memories or perceptions suggestive of consciousness during the arrest. Researchers identified four distinct categories of experience: some patients regained awareness while CPR was being performed, others emerged from coma only in the post-resuscitation period, some had dream-like experiences, and roughly a fifth described what the study authors called a “transcendent recalled experience of death,” the classic near-death experience involving feelings of separation from the body, moving through a tunnel, encountering deceased relatives, or reviewing one’s life.6PubMed. AWAreness during REsuscitation – II: A multi-center study of consciousness and awareness in cardiac arrest
That last category is where the idea of memory “playback” sits. Life reviews, as they are called in the near-death experience literature, involve a rapid and vivid re-experiencing of autobiographical memories, sometimes described as panoramic or nonlinear, as though years of life are compressed into seconds. Not every cardiac arrest survivor reports this. In the AWARE-II data, only about one in five interviewed survivors had the transcendent version. But those who do report it tend to describe it with extraordinary conviction.
Why These Memories Feel “Realer Than Real”
One of the stranger findings in this field is that people who have near-death experiences do not remember them the way they remember ordinary events. A study that administered standardized memory questionnaires to 122 survivors of a close brush with death found that the memories people had of their near-death experience scored higher on measures of richness, vividness, and detail than memories of real events from around the same time. Those real-event memories, in turn, scored higher than memories of things the person had merely imagined. In other words, when you ask people to rate the “realness” of their near-death memory alongside other memories, the near-death memory consistently wins.7ScienceDirect. Characteristics of memories for near-death experiences
This is difficult to explain with any simple model. If the dying brain were just producing garbled hallucinations from a failing system, you would expect those memories to be vague and dreamlike, not hyper-detailed. One possibility is that the gamma surge described earlier produces a state of unusually intense neural activation, encoding the experience with more sensory and emotional weight than everyday events receive. Another is that the experience is reconstructed after the fact, shaped by the emotional weight of having nearly died. Both explanations have supporters, and neither has been definitively tested.
How the Brain Shuts Down (and the Window That Creates)
To understand why there might be a brief window for memory-like activity in the dying brain, it helps to know the timeline of what happens when blood flow stops. Neurons do not die the instant the heart stops. Research on dying patients has shown that spontaneous electrical activity goes silent across the cortex within seconds to minutes as oxygen levels plummet, but the neurons themselves remain alive and structurally intact for several more minutes. A terminal wave of depolarization, a slow electrochemical collapse, begins to spread through the cortex roughly four minutes after the final drop in blood flow.8PubMed Central. Terminal spreading depolarization and electrical silence in death of human cerebral cortex
This spreading depolarization is what actually kills the neurons if blood flow is not restored. Before it arrives, there is a brief period in which the neurons have gone electrically quiet but are not yet dead. If energy supply is restored in time, the tissue can recover.9PubMed Central. Spreading depolarization: A wave that precedes and drives cerebral ischemic cell death—Target for neuroprotection The gamma surge observed in the dying brain likely occurs in the narrow window before spreading depolarization reaches each region of the cortex, and may even be triggered by the metabolic disruption itself. In rat studies, the final depolarization wave reached 50 percent of its maximum faster in the cortex than in deeper brain structures like the olfactory bulb, suggesting that different regions fail at different rates.10PubMed. Terminal anoxic depolarization proceeds more slowly in the olfactory bulb than in the cerebral cortex of rats The staggered shutdown may help explain why the experience, when it occurs, unfolds in stages rather than as a single coherent replay.
The DMT Theory and Its Problems
A popular idea in pop neuroscience holds that the dying brain floods itself with dimethyltryptamine, or DMT, a naturally occurring psychedelic compound, and that this chemical release is what produces the tunnel of light, life review, and mystical feelings reported during near-death experiences. The idea was popularized in a 2000 book by psychiatrist Rick Strassman, who proposed that the pineal gland releases a massive dose of DMT at death.
The scientific picture is more complicated and less dramatic. A 2018 review of the evidence found that while tiny amounts of DMT have been detected in the brain, the concentrations measured are far too low to produce psychedelic effects. The notion that the pineal gland serves as a DMT factory that floods the brain at death is not supported by the available data.11PubMed. N,N-dimethyltryptamine and the pineal gland: Separating fact from myth A 2019 study did find something genuinely interesting, though: in rats, the brain itself can synthesize DMT in multiple regions (not just the pineal gland), and the extracellular concentration of DMT in the cortex is comparable to that of serotonin under normal conditions. After cardiac arrest, DMT levels in the rat visual cortex increased significantly, and this increase happened regardless of whether the pineal gland was intact.12PubMed Central. Biosynthesis and Extracellular Concentrations of N,N-dimethyltryptamine (DMT) in Mammalian Brain
So there is a real DMT signal at death, at least in rats, but the popular narrative gets the mechanism wrong. The increase comes from the cortex, not the pineal gland, and nobody has yet shown that the concentrations reach a level that would produce a full psychedelic experience in humans. DMT may play some modulatory role in the dying brain’s electrical activity, but it is unlikely to be the sole or even the primary explanation for near-death experiences.
When Dreaming and Dying Overlap
There is a curious connection between near-death experiences and the sleep state associated with dreaming. A 2019 study surveyed over a thousand people and found that those who reported near-death experiences were far more likely to also experience REM sleep intrusion, a phenomenon in which features of dreaming (visual imagery, muscle paralysis, hallucinations upon waking or falling asleep) bleed into waking consciousness. Among people with near-death experiences, about 47 percent showed evidence of REM intrusion, compared with 14 percent of those who had never had such an experience. After adjusting for age, gender, and other factors, people with REM intrusion were nearly three times more likely to report a near-death experience.13PubMed Central. Prevalence of near-death experiences in people with and without REM sleep intrusion
This does not mean near-death experiences are “just dreams.” But it does suggest that the brain mechanisms involved in generating vivid internal experience during sleep may be the same ones activated during the dying process. The brainstem circuits that regulate REM sleep are among the last to fail during oxygen deprivation, which could mean that some dying brains slip into a REM-like state in which internally generated imagery, including memory fragments, becomes extremely vivid. Whether this pathway is separate from or overlapping with the gamma surge is an open question.
Electrical Stimulation and the “Life Review” Template
Another thread of evidence comes from decades of research on what happens when you electrically stimulate the brains of living patients, typically during surgery for epilepsy. A review covering 80 years of such cases found that stimulation can trigger vivid autobiographical memories, but the results depend heavily on where you stimulate. Almost all memory-triggering stimulation was in the temporal lobes, the brain region that houses the hippocampus and surrounding memory-related cortex. Hippocampal stimulation in particular was the only type that reliably produced full episodic memories, the kind with a specific time, place, and felt sense of re-experiencing. Stimulation of nearby temporal cortex tended to produce vaguer personal-semantic memories, like knowing a fact about your life without reliving the moment itself.14ScienceDirect. Memory scrutinized through electrical brain stimulation: A review of 80 years of experiential phenomena
This matters because the temporal lobes, particularly the hippocampus, are part of the network that the dying brain’s gamma surge activates. If spreading depolarization or oxygen deprivation produces a wave of excitation through this tissue before it shuts down, the result could be something like the electrically stimulated memory recall that neurosurgeons have been documenting for decades. The fact that most stimulation-induced memories are fragmentary and poorly detailed rather than panoramic does raise questions about whether the “whole life flashing before your eyes” narrative is an accurate description or a culturally shaped interpretation of briefer, less coherent memory fragments.
Cultural Filters on the Dying Experience
While certain elements of near-death experiences seem to appear across cultures (a sense of leaving the body, encountering light, meeting deceased individuals), the specific content varies considerably. A cross-cultural review published in Transcultural Psychiatry found that the differences are most likely shaped by language, religious background, education, and personal belief systems. A Hindu patient might encounter Yamaraj, the god of death, while a Christian patient might see Jesus or a tunnel of light. The underlying neurological process may be similar, but the interpretation is filtered through whatever framework the person has available.15PubMed. Phenomenology of near-death experiences: a cross-cultural perspective
This has implications for the memory-playback question. If the dying brain is genuinely replaying stored memories, you would expect the content to be deeply personal and culture-specific, which it is. But if the brain is producing a more generic altered state that each person interprets through their own experience, then the “life review” is less a playback and more a construction, a story the brain builds from available material under extreme physiological conditions. Both accounts are consistent with the EEG data. The gamma surge tells us something is happening; it does not tell us whether the person is passively watching a highlight reel or actively assembling meaning from fragments.
Could Evolution Have Built This In?
One speculative but testable idea is that near-death experiences have evolutionary roots in thanatosis, the “playing dead” defense strategy seen across the animal kingdom. A 2021 systematic investigation published in Brain Communications traced thanatosis from insects to mammals and argued that when a human is in a life-threatening situation, the brain may activate a deeply conserved circuit that produces immobility and altered consciousness. The researchers showed that humans under threat from predators (animal or human) can experience both thanatosis-like freezing and near-death experience phenomenology, and that the two overlap in their features and apparent survival function.16PubMed Central. The evolutionary origin of near-death experiences: a systematic investigation
If this is right, the vivid internal experience during dying may not be a pointless byproduct of a brain shutting down. Instead, it could be a remnant of a defensive program designed to keep the organism physiologically calm and internally occupied while “playing dead,” buying time for the threat to pass. The memory replay component could be part of this calming effect or could be incidental, a side effect of activating the same brain circuits that are involved in both immobility and memory consolidation. The hypothesis is still young and untested in any rigorous mechanistic way, but it offers a plausible reason why evolution would not have selected against such an energetically expensive final act.
Terminal Lucidity and Related Phenomena
A related but distinct puzzle is terminal lucidity, sometimes called paradoxical lucidity: the unexpected return of mental clarity in patients who have been cognitively impaired, sometimes severely, for months or years. Patients with advanced dementia who have not recognized family members in months may suddenly become alert, conversational, and coherent in the hours or days before death. Clinicians who work in end-of-life care widely recognize the phenomenon, but systematic scientific evidence remains scarce.17PubMed Central. Reports About Paradoxical Lucidity from Health Care Professionals: A Pilot Study
Terminal lucidity is not the same thing as a near-death experience or a life review. It happens in conscious patients, not during cardiac arrest or clinical death. But it raises the same uncomfortable question: how can a brain that appears to be irreversibly damaged suddenly produce organized, coherent cognitive function? If the dying process itself triggers a cascade of neurochemical and electrical changes that can temporarily restore network function in a severely degraded brain, then the idea that a relatively intact brain might produce vivid memory experiences during the same process becomes somewhat less far-fetched. The mechanism behind terminal lucidity is unknown, and studying it is logistically difficult because it is unpredictable and brief. But its existence suggests the dying brain is doing more than simply winding down.
Psychedelics, Autobiographical Memory, and the Dying Brain
Research on psychedelic drugs offers another window into how altered brain states interact with memory. Psilocybin, the active compound in psychedelic mushrooms, has been shown in small studies to increase the subjective intensity of positive autobiographical memories, making them feel more vivid, more visual, and more emotionally charged.18SpringerOpen. Psychedelics and autobiographical memory – six open questions This is notable because psychedelics produce many of the same subjective effects reported in near-death experiences: ego dissolution, a sense of cosmic unity, vivid visual imagery, and encounters with seemingly meaningful narratives. They also alter gamma oscillation patterns and serotonin signaling, two of the same systems implicated in the dying brain’s activity.
None of this proves the dying brain is operating on the same pharmacological principles as a psilocybin trip. But the overlap is suggestive enough that several research groups are now explicitly studying psychedelic experiences as a model for understanding near-death phenomenology. If both states share a final common pathway through gamma-coupled memory networks and serotonin-receptor activation, studying one under controlled conditions could shed light on the other, which is by definition impossible to study in a controlled setting.
Reading the Dying Brain With Technology
One question people often ask is whether we could ever “see” what a dying person is experiencing by reading their brain activity. The technology is not there yet, but the building blocks are emerging. Researchers have demonstrated that visual information can be decoded from EEG signals in living subjects using machine-learning models. In one approach, neural network systems were trained to match brain activity patterns with the images that produced them, achieving meaningful accuracy at reconstructing what a person was seeing from their EEG alone.19PubMed. Decoding Brain Representations by Multimodal Learning of Neural Activity and Visual Features
Applying this to the dying brain is a much harder problem. The gamma surge is brief and chaotic compared to the controlled visual stimulation used in decoding experiments. The brain’s metabolic state is collapsing, which degrades signal quality. And there are profound ethical barriers to instrumenting dying patients for the purpose of decoding their final experiences. Still, as brain-computer interface technology advances and EEG monitoring in critical care becomes more routine, it is possible that future recordings will capture enough data to attempt at least coarse reconstruction of dying-brain activity patterns. Whether that would tell us anything about the person’s subjective experience, or only about the brain’s electrical behavior, is a philosophical question that the technology alone cannot answer.