Temple Wearable Review: Does Deepinder Goyal's Device Really Measure Brain Blood Flow?

Commentary · By Dr Saurabh Kumar, MD (Psychiatry, AIIMS New Delhi) · Consultant Neuropsychiatrist, Healthy Minds India, Patna · 18 October 2026

I am a consultant neuropsychiatrist in Patna. At Healthy Minds India I measure and change brain physiology every working day, with rTMS, tDCS and fNIRS, so I have no quarrel with the idea that cerebral blood flow is worth measuring, or that a wearable might one day do it well. My concern is sequence. A measurement technology earns the right to make claims by being validated, and the validation comes first.

This Temple wearable review audits the published evidence against the standard framework for wearable biometrics, lists the public claims that outrun that evidence, argues that the way those claims were made is an ethical problem rather than a stylistic one, and sets out the legal exposure this creates in India and abroad. I am not a lawyer and nothing here is a finding that any law has been broken. I am a clinician explaining why I will not let this number into a consultation yet.

Temple wearable review in brief: quick answers

What is Deepinder Goyal's Temple brain wearable, and what does it claim to measure?

Temple is a temple-worn optical wearable from Temple Private Limited, promoted almost entirely through the social media accounts of its founder, Deepinder Goyal, who co-founded Zomato and its parent company Eternal. He stepped back from food delivery early in 2026 and within weeks had raised about US$54 million for the new venture.¹

The hardware is not exotic. Multi-wavelength reflectance photoplethysmography (PPG, the same light-based pulse sensing used in a smartwatch), an accelerometer and a skin-temperature sensor sit in a small housing over the front of the temple, held by an adhesive grip. The founder has himself compared the optical sensor to the one in a smartwatch.² What is new is where it sits and what the app says about the signal. Two proprietary outputs are offered:

• Brain Flow Index, marketed as BrainFlow, presented as a continuous stand-in for cerebral blood flow.
• Entropy, a number from 1 to 250 that refreshes every second, described as a readout of metabolic and sympathetic demand, the real-time cost of being alive.³⁴

Temple wearable price in India and launch

The Temple wearable price in India is expected to be ₹72,000 to 80,000 with lifetime software access, and shipping is promised before the end of 2026.³ Temple's own terms of sale describe it as a wellness product rather than a medical device, and warn that its accuracy is not intended to match professional medical or scientific instruments.

Does the Temple wearable measure brain blood flow? Start with the anatomy

The front of the temple sits over the frontal branch of the superficial temporal artery. That artery is a branch of the external carotid, and it supplies the scalp, not the brain. A reflectance PPG sensor pressed against the skin there looks through a few millimetres of tissue at most. What it sees is scalp perfusion: the pulse wave arriving from the heart, shaped by local sympathetic tone, skin temperature and how hard the device is pressing.

That is not a trivial signal. Scalp blood flow changes with posture, exercise and emotion, and some of those changes run in parallel with changes in the brain because both answer to the same systemic drivers. But parallel is not identical. The one scientific question Temple must answer is whether a surface signal at the temple carries any cerebral information that could not be obtained from the wrist or the finger.

The one experiment that bears directly on this question in the flow domain, blocking the superficial temporal artery during optical measurement over the scalp, found that a substantial fraction of the optical flow signal at short sensor spacings came from outside the skull.⁵ Temple's own preprint, discussed below, points the same way. The anatomy is not an argument against the product. It is the reason the burden of proof sits where it sits.

Ten months of escalating claims

The claims have grown over 2026, and the pattern matters as much as any single statement. Scepticism was first told to wait, then told it had missed its turn.

A sound study made to carry an unsound claim

The use made of the mortality cohort is the clearest example in this story. The 12-year follow-up that the company relayed measured total cerebral blood flow with phase-contrast MRI in older adults, once, at baseline, in the basilar and internal carotid arteries, and then related that value to later deaths.⁸ So: whole-brain, absolute, volumetric, in people in their seventies and eighties.

Temple reports a relative, uncalibrated optical index at one spot on the scalp, continuously, in young consumers. Nothing in the cohort study tells us that a change in that index means anything, that it can be improved, or that improving it would change the outcome. There is also the familiar problem that the association is exactly what you would expect if low cerebral blood flow were a marker of accumulated vascular disease rather than a cause of death; the cohort authors adjusted for cardiovascular risk and were careful about mechanism.

Turning their finding into people with higher cerebral blood flow have twice the shot at a longer life, and placing that sentence beside a product photograph, is the ecological fallacy performed as advertising. The broader idea behind the product, the founder's Gravity Ageing Hypothesis that gravity-driven falls in brain perfusion drive ageing, is described even in sympathetic press coverage as a hypothesis.³ I have not found a peer-reviewed statement of it anywhere.

Is the Temple wearable scientifically validated? Where the evidence stands

The Digital Medicine Society's V3 framework separates three questions that a product launch tends to blur.¹⁴ A later extension adds usability.¹⁵ Marketing is honest only when it stays within the stage actually reached. Here is where Temple stands at each stage.

The one preprint

Temple has one preprint, posted to bioRxiv on 17 September 2026 and, according to its authors, submitted for peer review.¹⁰ All nine authors are company staff or, in the case of the senior author, a scientific adviser to the company. The study enrolled 44 men aged 20 to 32, giving 60 sessions across head-down tilt, stand-to-squat and stand-to-supine challenges. Women were excluded to reduce hair-related artefact.

The reference was a Kernel DevKit-3 time-domain NIRS system whose signals were run through a two-layer mathematical model to produce separate scalp-layer and brain-layer haemoglobin traces. The headline finding is that the Temple index tracked the brain-assigned layer more closely than the scalp-assigned layer, with median lag-optimised correlations of 0.84 to 0.91.¹⁰ The authors describe this with proper restraint: TD-NIRS was used to separate surface from deeper responses, not as a gold standard for flow, and the remaining systemic and superficial contributions are listed as unresolved.

The transcranial Doppler study in 23 adults that the company counts as its first validation has, so far as I can establish, not been posted in any form that allows scrutiny.³⁹ Entropy has no published study at all.

Validation is only as good as the reference, and here the reference is a model

The preprint's brain compartment is not a direct cortical measurement. It is what is left after a two-layer inversion that assumes a fixed 8 mm scalp-and-skull thickness and a brain-to-scalp absorption ratio of 1.5, and the authors concede the two layers remain partly correlated.¹⁰ This is the central, well-documented difficulty of measuring the brain with light from the surface. A systematic review concluded that tissue outside the brain does influence continuous-wave NIRS,¹⁶ and a clinical study found that scalp circulation markedly affected cerebral blood-volume indices in a high proportion of cases.¹⁷

The accepted partial remedy is short-separation regression, and the preprint's own numbers show how much work it does: correlations between the Temple index and the brain layer fell from 0.83 to 0.85 after heart-rate adjustment to 0.44 to 0.65 after short-separation adjustment.¹⁰ The authors call the superficial contribution part of the structure. A reader without a stake in the product would call it a large part.

There is also a problem of vocabulary. The product talks about brain blood flow. NIRS estimates haemoglobin concentration, a volume quantity, and transcranial Doppler measures red-cell velocity in a single artery. Flow is velocity multiplied by cross-sectional area, and neither reference supplies both. A 2022 position paper asked the field to stop saying cerebral blood flow velocity precisely because the conflation misleads,¹⁸ and flow inferred from velocity alone can err by 20% when vessel diameter changes.¹⁹ The preprint acknowledges all of this in its limitations. The company's public language does not.

"Only wearable validated against brain blood volume": validated against a model-derived estimate of a volume-related quantity, in a laboratory, in young men, during three manoeuvres, by the company and its adviser, pending peer review. All of those qualifiers are true. None of them appears in the post.

Temple wearable accuracy: the statistics, and who was left out

The preprint's statistics would not survive ordinary peer review intact. The 60 sessions come from 44 people, several of whom did more than one protocol, yet the tests treat sessions as independent; repeated measures call for a mixed model, and ignoring that structure usually makes p-values too small. Several tests are one-sided with no stated reason to disregard the opposite direction. Multiple secondary analyses were run without correction, then reported as though each stood alone.

The correlations quoted in the marketing are the best values found by sliding the two signals up to 30 seconds against each other. The zero-lag values, the fair numbers for a concurrent-validity claim, are 0.83, 0.70 and 0.80.¹⁰ Two signals that both respond to an abrupt posture change will correlate strongly whatever their origin, a point the authors themselves raise. Agreement was assessed on standardised traces, which strips out the scale and offset that Bland-Altman analysis requires;²⁰ what the plots do show is that the Temple response was smaller than the reference in every protocol, by 0.25 to 0.51 standard deviations.

No power calculation was performed. No test-retest reliability, intraclass correlation or smallest detectable change was reported, so a user has no way to tell a real change in the index from noise. For a device whose whole premise is watching the number move through the day, that is not a minor omission. Two design features bear directly on the people being sold the device:

1. 1Movement

   Consumer PPG accuracy drops sharply with rhythmic motion, the clearest finding from Bent and colleagues' comparison of six wrist wearables against ECG,²¹ and one that survives the dispute over how skin tone was classified in that work.²² The squat protocol is a motion challenge by design, the accelerometer-based artefact handling is undescribed, and contact pressure and sweat were not varied. A device marketed to elite athletes was validated while people stood still or lay down.

2. 2The sample

   Forty-four healthy men aged 20 to 32. Monk Skin Tone categories 3 to 8 were included but their distribution and relationship to signal quality went unexamined.¹⁰ Women, older adults, anyone with heart, blood-vessel or psychiatric illness, anyone on a drug that changes vessel tone: all absent. These are precisely the people in whom cerebral haemodynamics matters clinically, and precisely the people a longevity wearable will reach.

The Temple Entropy score: a biomarker by announcement

A biomarker, in the sense used by the FDA-NIH BEST glossary and every regulatory framework I know of, is a characteristic measured as an indicator of a biological process, with a defined context of use and evidence of analytical and clinical validity.²³ It is not discovered by naming a number on a home screen. The claim that resting Entropy tracks lifespan across the animal world alludes to the rate-of-living hypothesis, a contested idea with a long history of exceptions,²⁴ here stretched to cover a proprietary index no one outside the company has seen. Selling a longevity signal on that basis is not a stretch of the evidence; there is no evidence to stretch.

The one number offered, r = 0.93 against indirect calorimetry over more than 100 cardio sessions, cannot be examined at all, which is a different problem from being weak. We do not know what was measured, whether the correlation is within or between sessions, how many people the sessions came from, whether heart rate came from the same device or a chest strap, or how the comparison figure of 0.55 for heart rate was arrived at, given that heart rate tracks oxygen uptake closely during steady exercise in nearly every study I know of.

A metabolic cart measures whole-body oxygen consumption and carbon dioxide production. A temple-worn optical sensor measures neither. Whatever Entropy is, then, it is a model that predicts energy expenditure from pulse, temperature and movement, which is what every consumer wearable already does, with modest accuracy. The further claim that it is readable only at the temple is, as stated, unfalsifiable; taken literally it would mean the scalp vessels carry metabolic information that the wrist and finger arteries do not, and no reason for that has been offered. I would read a methods paper with interest. Until there is one, Entropy is a branded estimate of exertion.

Why this is an ethical problem, not start-up exuberance

I am reluctant to call any of this ordinary start-up enthusiasm that the market will sort out. The claims are about the brain. They are made to people who cannot check them, attached to a price that signals seriousness, by a person whose earlier success gives him a credibility the evidence has not yet earned.

What troubles me most is one particular move: posting a preprint and, the same day, treating it as settled. Posting to bioRxiv is good practice. Using the posting, before any reviewer has seen the paper, as the warrant for a worldwide superlative is something else. The paper's limitations section is careful; the post that quotes it is not; and peer review was the one step the marketing was timed to get ahead of. In January, criticism was met with the reply that it was too early to judge an unreleased device. In September, a preprint was treated as enough to close the question.

The disclaimers point the same way. Temple's terms say it is not a medical instrument. Its public messaging says cerebral blood flow is a proven biomarker for ageing, that low flow doubles mortality, and that the device measures it continuously. A buyer reads the second; the first is for the lawyers. Indian advertising guidelines are explicit that a disclaimer may not contradict the main claim of an advertisement or be used to correct a misleading impression.²⁵

Two further statements deserve mention because of whom they concern. The remark that PPG poses no risk to a developing brain is a safety assertion about children made without a paediatric study. Low-power optical sensing is very probably harmless, and that is not the point; the honest sentence was we have not tested this in children, and the line about aspiring Olympians reads as an invitation rather than a restriction. And no wearable anywhere has clinical validation for fitness-for-duty screening of pilots, where the cost of a false reassurance is obvious; floating the use now does marketing work and defers the evidence indefinitely.

None of this is new in kind. Wexler and Reiner said as much about consumer neurodevices in 2019,²⁶ and Ienca and colleagues set out the privacy and dignity problems of brain-related data gathered outside any clinical frame.²⁷ What is new, at least in India, is the combination: a celebrity founder, a form factor that makes the brain claim visually literal, a price that selects for people who will take the number seriously, and a press that has, with a few exceptions, run each post as news.

There is precedent for how this ends. In 2016 the US Federal Trade Commission settled with the makers of Lumosity, a brain-training programme sold on claims of improving cognition and delaying age-related decline; the company paid two million dollars, had to notify its subscribers, and was barred from future claims without competent and reliable scientific evidence.²⁸ The problem was not that brain training was implausible, but that the claims came first and the evidence was sought later, with consumers' anxiety about ageing as the lever. Closer to home, the Patanjali litigation reached the Supreme Court because a company with a charismatic founder and a national following kept making health claims after being told, repeatedly, that it could not substantiate them.²⁹

The legal exposure, in India and abroad

I am not a lawyer, and nothing here is a finding that any law has been broken. What I can do is set out the questions the conduct raises and the instruments under which a regulator or court would consider them.

Consumer protection

Section 2(28) of the Consumer Protection Act 2019 defines a misleading advertisement to include one that falsely describes a product, is likely to mislead consumers as to its nature or quality, or deliberately conceals important information.³⁰ The Central Consumer Protection Authority's 2022 guidelines add that an advertisement must not exaggerate the accuracy, scientific validity or practical usefulness of a product, must not suggest its claims are universally accepted where informed scientific opinion is significantly divided, and must be capable of objective substantiation on demand. They apply to every medium including social media, and the Act makes endorsers as well as manufacturers liable, so a founder posting in his own name is both.²⁵³⁰

Penalties under section 21 run to ₹10 lakh for a first contravention and ₹50 lakh thereafter, with the possibility of an endorsement ban, and section 89 provides for imprisonment of up to two years for a false or misleading advertisement prejudicial to consumers.³⁰ These are small sums for a company that raised US$54 million at a valuation near US$190 million; the consequences of a regulatory notice are not measured in rupees.

Is Temple a medical device in India? The medical device rules

The Medical Devices Rules 2017 define a medical device by intended use, which includes monitoring of disease and investigation of a physiological process, and since 2020 all devices fall within the regulatory perimeter.³¹ The CDSCO's 2026 guidance on medical device software draws the wellness boundary clearly: wellness software may claim to sustain or improve general health but must not refer to disease, disorder or pathological condition, and software intended to report physiological values for screening or monitoring is not wellness software.³² Temple's terms place it on the wellness side of that line; its marketing walks across it. Ageing is not a disease, but mortality risk is a clinical outcome, cerebral blood flow is a physiological process, and continuous monitoring is monitoring. A regulator asked to classify the device would look at what it is sold to do, not at what the terms say it is not.

The Drugs and Magic Remedies (Objectionable Advertisements) Act 1954 is an awkward fit for a passive monitor, but it was the vehicle for the Supreme Court's 2024 judgment in Indian Medical Association v Union of India, the Patanjali matter, in which the Court made clear that misleading health advertising is a matter of public concern requiring enforcement under both the 1954 and 2019 Acts.²⁹ Any company making brain and longevity claims in India after August 2024 is operating in that climate.

Abroad

The US FDA's January 2026 general wellness guidance now explicitly allows non-invasive optical sensing to estimate physiological parameters outside device regulation, provided the product makes no disease, diagnostic or clinical-management claim.³³ Readiness screening for pilots is a clinical-management claim, and mortality risk is a disease-referenced claim. The Federal Trade Commission separately requires competent and reliable scientific evidence for health claims before they are made, and treats a disclaimer buried in terms of sale as no defence.³⁴

The data question

This is likely to become more pressing than the advertising question. Temple collects a continuous signal from the head, labels it as brain data, and offers lifetime cloud access as part of the purchase price. Whatever the signal actually contains, the label matters, because users have been told the number reflects their brain and will treat it accordingly. India's Digital Personal Data Protection Act 2023 does not yet carve out neural data as a special category, but the international direction is clear. Chile amended its constitution in 2021 to protect mental integrity and neural data, and UNESCO's member states adopted a Recommendation on the Ethics of Neurotechnology in November 2025 that asks governments to treat neural data as sensitive, to prohibit its commercial exploitation without informed consent, and to subject consumer neurotechnology to evidence and transparency requirements.³⁵ A company that claims to read the brain should publish its data-protection policy before it ships, not after it is asked.

Temple Brain Flow Index at a glance: what is claimed, what is shown

The version of this product that could be sold tomorrow

In fairness, there is a version of Temple that could ship without departing from the evidence, and it is worth saying what it would look like. The company could say it has built a small temple-worn optical sensor that records a usable PPG signal through daily activity, which is a real engineering achievement. It could say that in one laboratory study of healthy young men, the index from that signal rose and fell with a model-derived deeper NIRS compartment during posture changes, more closely than it tracked the surface compartment, and that the study is under review. It could add that the index is not a measure of cerebral blood flow, that its behaviour in women, older adults and patients is unknown, that the device has no clinical use, and that the studies needed to find out are being funded. I think that would be a more interesting product, not a less interesting one.

What would satisfy a clinician

Critique that cannot say what would satisfy it is merely complaint. The standard I would apply is the one I apply to any device considered for my own clinic:

1. Peer-reviewed publication of both the TD-NIRS and transcranial Doppler studies, with mixed-model analyses, two-sided tests and zero-lag correlations reported as primary.
2. A flow-sensitive reference such as diffuse correlation spectroscopy, calibrated perfusion imaging or Doppler with vessel diameter measured, alongside beat-to-beat blood pressure and end-tidal carbon dioxide, a design already shown feasible in a wearable.³⁶
3. Test-retest reliability and smallest detectable change measured in daily life rather than on a tilt table.
4. Validation in women, older adults, patients with cardiovascular, cerebrovascular and psychiatric illness, and people on vasoactive drugs, with skin-pigmentation distribution reported.
5. A defined context of use, a threshold, and a clinical or performance outcome.
6. A methods paper for Entropy.
7. Independent replication by a group with no financial or advisory tie to the company.

A company with US$54 million in hand can do all of this. Most of it should have been done before the first post.

Is the Temple wearable worth buying? What this means in the consulting room

For Indian clinicians the matter is more immediate than regulation. Many of us already see patients who arrive with a continuous glucose trace, a sleep score or a heart-rate-variability readout and ask what it means for their mood or their memory. A number labelled brain blood flow, worn on the head and priced like a medical investigation, will carry more weight in that conversation than any of those, and it will carry it without a single clinical study behind it.

My question has been whether a wearable may be sold on a brain claim before the claim has been tested. Temple's answer, in practice, has been yes: announce, raise, post, preprint, pre-order. Mine is no, and not because I doubt that temple-worn optics will turn out to be useful. I rather expect they will. The objection is to the order of events, which is the difference between a measurement and a story about one, and the people who pay ₹75,000 for the story are the ones who will find out which it was.

The position for clinicians is simple enough. Temple has no validated clinical use. It should not inform a decision about diagnosis, treatment, screening or fitness for duty, and a patient who brings a reading should be asked what it is and how it was obtained rather than what it means. I would ask journal editors handling the preprint to review it with the marketing in view, since readers will not separate the two. For regulators, the material is on the public record. For the company, the remedy is also simple, and in its own interest: publish first.

This article is a commentary on publicly available marketing and scientific material. It is not medical or legal advice, and it is not a finding that any law has been broken. Readers with concerns about their own health should consult a qualified clinician.

Frequently Asked Questions

• Does the Temple wearable measure brain blood flow?

  Not in any validated sense. Its sensor reads scalp blood flow over the superficial temporal artery. One company preprint in 44 young men shows the index co-varies with a model-estimated deeper NIRS signal during posture changes. That is a volume-related estimate, not a flow measurement, and it has not been peer-reviewed or independently replicated.

• Is the Temple wearable scientifically validated?

  Not yet. The only published evidence is one company preprint in 44 healthy men aged 20 to 32, posted to bioRxiv in September 2026 and not yet peer-reviewed. There is no independent replication, no reliability data and no clinical study, and the Entropy score has no published study at all.

• What is the Temple wearable price in India?

  The expected price is ₹72,000 to 80,000, including lifetime access to the app, with shipping promised before the end of 2026. Final pricing had not been confirmed when this review was written.

• Is Temple a medical device in India?

  Its terms of sale call it a wellness product. Under the Medical Devices Rules 2017, classification turns on intended use, and marketing that speaks of monitoring a physiological process and mortality risk sits on the medical side of CDSCO's 2026 wellness boundary. Whether a regulator would classify it that way has not been tested.

• What is Temple's Entropy score?

  A proprietary number from 1 to 250, refreshed every second, described by the company as the real-time cost of being alive. No paper, preprint or methods description has been published. From what the sensor can physically measure, it is most likely an estimate of exertion built from pulse, skin temperature and movement.

• Can a brain wearable reading change my psychiatric diagnosis or treatment?

  No. No consumer brain wearable currently has clinical validation for diagnosis, treatment selection or monitoring in psychiatry. If you bring a reading to your psychiatrist, expect to be asked what the device is and how the number was obtained, not what it means for you.

• Is the Temple wearable worth buying?

  Not on the evidence available today, if you are buying it for what the marketing promises. As an engineering product, it records a usable optical signal from the temple. As a measure of brain blood flow or longevity, it has no peer-reviewed, independent or clinical validation. Wait for the published studies, or buy it knowing you are paying for an experiment.

• How accurate is the Temple wearable?

  Nobody outside the company can say. The preprint reports no test-retest reliability, no smallest detectable change and no error against a true flow measurement, so there is no way to tell a real change in the index from noise. Motion, sweat and contact pressure, the usual enemies of optical sensors, were not tested.

• Deepinder Goyal ka Temple device kya hai?

  Temple ek chhota optical sensor hai jo kanpati (temple) par chipka kar pehna jaata hai. Yeh smartwatch jaisi PPG light technology se nabz aur khoon ka bahaav padhta hai, aur app mein usse "brain blood flow" ka number dikhata hai. Abhi tak iska koi peer-reviewed ya clinical study prakashit nahin hua hai.

• Temple wearable kitne ka hai?

  India mein iska expected daam lagbhag ₹72,000 se 80,000 hai, jismein app ka lifetime access shaamil hai. Company ne 2026 ke ant tak shipping ka vaada kiya hai. Itna paisa dene se pehle yeh jaan lein ki iska koi peer-reviewed ya clinical validation abhi tak nahin hai.

• Are there validated ways to assess brain function in a clinic?

  Yes. Clinical interview and standardised rating scales remain the core. Where indicated, tools such as qEEG, fNIRS, neuropsychological testing and MRI have defined uses and known limits. At Healthy Minds India in Patna these are used alongside, never instead of, a proper clinical assessment.

About this article

Written and medically reviewed by

Dr. Saurabh Kumar

• MD (Psychiatry) AIIMS New Delhi
• Associate Professor & HOD, Dept. of Psychiatry, RDJM Medical College, Muzaffarpur
• Ex-Assistant Professor, PMCH Patna
• Consultant, Healthy Minds India, Patna

Published

18 October 2026

Format

Commentary

Focus

Consumer brain wearables and the evidence behind them

Sources

36 references

• Conflicts of interest: the author has no financial, advisory or personal relationship with Temple Private Limited, Eternal Limited, Kernel, Boston University or any manufacturer of competing wearables, and received no funding for this work.
• Dr Saurabh Kumar's research on rTMS, treatment-resistant depression and OCD has been cited over 600 times across about 59 publications.
• This article is a commentary on publicly available marketing and scientific material. It is not medical or legal advice, and it is not a finding that any law has been broken.
• Written for general education. It does not replace an individual assessment by a qualified clinician and does not diagnose any condition. Diagrams are original illustrations prepared for this article.

© 2026 Healthy Minds India, Patna · Commentary, October 2026