AI body-scan era 2026 — Midjourney Medical, OpenBioHack, full-body MRIs, and the new market for 'scan yourself, then what?'

What it claims

• 'Catch cancer early', 'know your body', 'longevity medicine for everyone', 'preventive imaging'
• Midjourney Medical / Rundown AI coverage — 'step into a spa for a body scan', AI-rendered body visualisation as a wellness experience
• OpenBioHack — 'open-source biohacking', DIY measurement stack, self-experiment infrastructure
• CGMs and 100-biomarker panels — 'metabolic health for non-diabetics', 'optimise your glucose curve'

What the label is not telling you

• The incidentaloma problem is the structural fact of healthy-person full-body imaging. A 2017 ACR white paper and multiple subsequent meta-analyses put the rate of incidental findings on a screening MRI in asymptomatic adults at 8–86% depending on the protocol — most of them clinically irrelevant, most of them requiring follow-up to confirm they are clinically irrelevant. The follow-up cycle (repeat imaging at 3 / 6 / 12 months, possible biopsy, definitely months of waiting) is the under-priced cost of a Prenuvo / Ezra / Neko purchase. For a regulated user with a high health-literacy baseline, this is manageable. For an anxious baseline, a recently-bereaved user, a postpartum or perimenopausal window, or any acute-stress phase, the scan is a sympathetic event that can take a year to come back from.
• Prenuvo, Ezra, Neko Health — the premium full-body MRI tier. The imaging is real and the protocol is sound. Neko (Daniel Ek's company) has the most considered clinic experience and is the most honest about the limits of what a scan can tell you. Prenuvo has the largest marketing footprint and the most documented incidental-finding controversy. The category is appropriate for adults with a meaningful cancer family history, a specific concern the standard healthcare system has dismissed, or a high health-literacy + regulated-baseline pairing that can metabolise an ambiguous result without it becoming the next six months of the nervous system.
• Midjourney Medical (preview, 2026) — the category's most interesting and most autonomically risky entrant. A body-scan-as-spa-experience, with AI-rendered visualisations of the inside of the user's body. The user-experience design is excellent — the spa frame addresses exactly the cold-clinical-imaging objection that has kept full-body MRI a niche purchase. The under-discussed risk is the same as Instagram's filter problem at a deeper layer: a generative model rendering 'you, inside' is a model rendering an interpretation, not a measurement. The line between 'this is your scan' and 'this is what the model thinks your scan should look like' is the most important consent question in the category, and it is not yet settled. Treat preview claims as preview claims; wait for the peer-reviewed validation studies before letting this category make decisions about your body.
• Function Health and the 100-biomarker panel tier. Function (Mark Hyman-backed) and similar services run a panel covering metabolic, hormonal, inflammatory, micronutrient and emerging biomarkers twice a year, with AI-assisted interpretation. The clinical case for the breadth is genuinely contested — the conventional medicine view is that screening healthy adults on 100 biomarkers will generate false positives at a rate that makes the panel net-harmful. The longevity-medicine view is that early metabolic drift is invisible on the 8-biomarker NHS / GP panel and visible on the 100-biomarker one. Both views are defensible. The autonomic question is the same as for imaging: how does your nervous system handle an ambiguous result?
• CGMs for non-diabetics (Levels, Lingo, Stelo, Nutrisense). The 2024–2026 mass-marketing of CGMs as a wellness purchase for healthy adults sits in the same family. The data is real, the visualisations are compelling, and for a metabolically curious adult with a regulated baseline, two weeks of CGM data is a useful one-time education in personal glucose response. The harm pattern, well-documented across our coaching cohort, is the user who turns a two-week experiment into a 12-month surveillance habit, becomes glucose-anxious around social meals, restricts food categories on the basis of post-prandial spikes that are physiologically normal, and arrives in a more dysregulated relationship with eating than they started with. Use as a 14-day diagnostic; do not wear continuously.
• OpenBioHack and the open-source DIY tier. The repo (github.com/OpenBioHack) is a real and useful resource for technical users who want to run a measurement stack outside the commercial wellness-tech ecosystem — open hardware, open analysis pipelines, no vendor lock-in, no data-monetisation. The honest scope: this is a technical product for technical users. The same caveats about incidentaloma, false positive, and autonomic cost of ambiguous data apply, with the added complexity that the user is now also their own clinician. Appropriate for engineers with a methodical baseline; inappropriate for any user already in measurement-anxiety territory.
• The category-wide pattern: no scan, panel or device measures your recovery capacity. The honest under-discussed variable across this entire market is not 'do you have the data?' — it is 'do you have the autonomic capacity to act on the data without the data becoming the new stressor?' Every product in this review answers the first question. None of them answer the second. The answer to the second question is the Kokorology Journal — the state instrument that tells you whether your week of CGM data, your full-body MRI report, your 100-biomarker panel and your OpenBioHack experiment are net-regulating or net-dysregulating in your month.

Effect on the nervous system

Diagnostic instrumentation is sympathetic by design. The waiting period for a result, the ambiguity of an incidentaloma, the cognitive load of integrating a 100-row biomarker report, the four-hourly glance at a CGM curve — each of these activates the threat-detection system the products promise to soothe. For users with a regulated baseline and a specific question, the category is genuinely useful and the data justifies the activation. For users in any acute load phase (burnout, perimenopause, postpartum, post-COVID, grief, an existing anxiety pattern, recent diagnosis in a family member), the category routinely makes things worse by adding ambiguous, partially-actionable, high-emotional-salience data to a system that is already over-recruiting threat detection. The decision is not 'is this product good?' — most of them are well-built. The decision is 'is my nervous system in a state where one more high-resolution data stream is regulating or dysregulating?' The journal is the only instrument that answers that.

Who it might suit

Full-body MRI (Prenuvo / Ezra / Neko) — adults with meaningful family cancer history, a dismissed-but-persistent specific concern, or a high health-literacy + regulated baseline. Once, not annually, unless a clinician recommends. Function-tier 100-biomarker panel — adults working with a longevity-literate clinician who will integrate the data, not chase every flag. CGM — 14-day diagnostic for metabolically curious adults with a regulated baseline, then off. OpenBioHack — engineers and methodical self-experimenters with the technical capacity to interpret and the autonomic baseline to absorb ambiguity. Midjourney Medical — wait for peer-reviewed validation; treat current preview as preview.

Who should skip it

Avoid the entire scan / panel / device wave during active burnout, perimenopause/menopause with sleep disruption, postpartum, post-COVID recovery, acute grief, recent family diagnosis, or any active anxiety disorder. The autonomic cost of ambiguous data lands hardest in exactly the states most likely to seek the data. Avoid continuous CGM wear — it converts a useful diagnostic into a daily surveillance loop that disrupts relationship with food in measurable ways. Avoid annual full-body MRI 'screening' in the absence of a clinical reason — the incidentaloma cycle compounds. Avoid Midjourney Medical-class generative body-visualisation as a basis for any clinical decision until the peer-reviewed validation work is published — until then, the rendering is interpretation, not measurement.

Bottom line

Honest hierarchy: (1) Start with state, not data. A Kokorology Journal week will tell you whether your nervous system has the recovery capacity to use a high-resolution body scan without the scan becoming the next month's load. If the answer is no, fix that first. (2) One thoughtful intervention at a time. A 14-day CGM, or a one-time full-body MRI with a meaningful clinical question, or a 100-biomarker panel integrated with a longevity-literate clinician — pick one, integrate the result, then stop. (3) Treat Midjourney Medical and the AI body-scan tier as preview technology, not as a decision-grade product, until the validation work catches up. (4) OpenBioHack if you are an engineer who wants to own your stack; otherwise the consumer products do the job. (5) The single best longevity intervention in this review is not in this review — it is sleep, walking, real food, real relationships, and a journal that tells you when you have stopped doing any of the four. Start with the free Starter Guide, then the Journal. The burnt-out Anchor covers the recovery-capacity question this category does not.