Picture the standard move in consciousness research: recruit some volunteers, show them a flickering image, slide them into an fMRI tube, and watch which regions light up. Rinse, repeat, publish. For decades that loop has produced a mountain of neural correlates, regions that seem to track awareness, but correlation is not causation, and the mountain keeps growing without anyone being able to say which rocks are actually holding it up. A visual stimulus may evoke neural activity that correlates with conscious perception, but that does not imply the activity constitutes the resulting conscious experience; teasing out which neural structures can directly elicit conscious perception versus those that merely react remains a central challenge of the field. A roadmap paper published in early 2026 by researchers at MIT, the University of Florida, Brigham and Women's Hospital, and Harvard Medical School argues that the deadlock is partly a tool problem, and that a quietly maturing technology called transcranial focused ultrasound (tFUS) is the tool that could break it.
The paper, "Transcranial focused ultrasound for identifying the neural substrate of conscious perception," appears in Neuroscience and Biobehavioral Reviews. Its authors are Daniel Freeman, a technical staff member at MIT Lincoln Laboratory; Brian Odegaard, an assistant professor of psychology at the University of Florida; Seung-Schik Yoo, an associate professor of radiology at Brigham and Women's Hospital and Harvard Medical School; and Matthias Michel, an associate professor in MIT's Department of Philosophy and Linguistics. That last affiliation is worth pausing on: a philosopher as co-author on a hardware roadmap. It signals that the team is not merely proposing a new scanner but a new epistemological strategy, a way to stop asking "where does the brain light up?" and start asking "which circuits are load-bearing for awareness?"
What makes tFUS different from the brain stimulation tools that already exist, things like transcranial magnetic stimulation (TMS) or alternating current stimulation, is a combination of depth and precision that neither of those technologies can match. TMS and tACS have been used in consciousness research before, but their spatial resolution is relatively coarse and they have limited use in targeting deep brain structures, which puts real constraints on the ability to selectively probe small targets such as a region of visual cortex corresponding to a small patch of the visual field, or millimeter-scale nuclei in the basal forebrain. tFUS bypasses those constraints by a wide margin. This noninvasive technology reaches deeper into the brain, with greater resolution, than techniques such as EEG and MRI; it works by sending acoustic waves through the skull to focus on an area of a few millimeters, allowing specific brain structures to be stimulated so the effects can be studied. Freeman put it plainly: "It truly is the first time in history that one can modulate activity deep in the brain, centimeters from the scalp, examining subcortical structures with high spatial resolution." That reach matters enormously because many leading theories of consciousness assign starring roles to deep structures, especially the thalamus, that have been essentially untouchable without a scalpel.
The authors are not only proposing ideas for others; Freeman and Michel are planning their own experiments, starting with visual cortex stimulation and later moving to frontal regions, with the explicit goal of linking neural manipulation directly to subjective experience. The logic of those experiments is elegant and pointed. As Freeman puts it: "It's one thing to say these neurons responded electrically. It's another thing to say a person saw light." In practice, if researchers can nudge the visual cortex with an acoustic pulse and a participant reports seeing a flash, that is causal evidence, not just association. tFUS intends to switch circuits on or off during perception tasks; if visual awareness disappears when a particular region is perturbed, sensory theories of consciousness gain weight, while unchanged perception under sensory disruption would bolster higher-order or global workspace accounts. The field's big competing theories, global workspace theory, integrated information theory, higher-order accounts, have spent years debating each other in something of an empirical vacuum; controlled tFUS experiments could begin to score those debates with real data.
The stakes run well beyond academic theory. tFUS has already emerged as a promising non-invasive neuromodulation technique for disorders of consciousness, with unique ability to precisely modulate deep brain structures, particularly the thalamus, while remaining non-invasive. Preliminary clinical trials in both acute and chronic patients with disorders of consciousness have shown encouraging results, including diagnostic category shifts, improvements in behavioral responsiveness, and alterations in thalamo-cortical connectivity. If causal experiments in healthy participants can clarify which circuits are necessary for awareness, clinicians treating patients in minimally conscious or vegetative states gain a much sharper map of what to target. The same logic bleeds into AI: as large language models grow more behaviorally sophisticated and the sentience debate intensifies, any method that can establish necessary and sufficient neural conditions for consciousness in biological systems gives philosophers and regulators a more rigorous baseline to reason from. The researchers are also aware they are selling hope, not a solved problem. Michel is characteristically candid: "It's a new tool, so we don't really know to what extent it's going to work. But I feel there's low-risk and high reward: why wouldn't you take this path?" Significant challenges remain, including optimizing stimulation parameters, addressing variability in patient responses, and ensuring long-term safety. And no matter how precisely you poke a brain region, inferring that subjective experience has changed still depends on the participant's verbal report, a measure that carries its own philosophical baggage. Michel himself acknowledges that "there are very few reliable ways of manipulating brain activity that are safe but also work," and tFUS still needs to prove it belongs in that short list across a wide range of targets and populations. Michel is also helping build a broader research community around these questions: along with neuroscientist Earl Miller, he co-founded the MIT Consciousness Club to encourage cross-disciplinary work on consciousness across the Boston area, a sign that the team sees the tFUS roadmap as the beginning of an institutional shift, not a single paper.
What is genuinely new here is not the hardware, which has been developing for years in clinical imaging and therapy, but the research design philosophy it enables. In their roadmap paper, the MIT team explains how this method could finally let scientists test cause-and-effect in consciousness research, not just observe correlations. That methodological reorientation, from passive observation to active intervention in healthy human brains, is what separates this moment from the last two decades of fMRI-heavy consciousness science. Whether tFUS can deliver on its promise will depend on the experiments Freeman and Michel are now designing; but the very fact that those experiments are now designable, noninvasively, in healthy volunteers, with millimeter precision, at depth, is a structural change in what the field is allowed to try.
If a focused acoustic pulse can reliably switch off a person's perception of light without touching a single hair on their head, the field will finally have something it has lacked since the 1990s: a test, not just a scan.