Healthspan, Lifespan, and WELLcareVibrome® biosignatures as a functional measurement layer

Vibrome® science treats the body as a resonant mechanical system and measures its output directly. Conventional sensors largely report biochemical concentrations at a moment in time, or movement-derived proxies such as step count and heart-rate variability. Vibrome® measurement instead quantifies the resonance and decay characteristics of biological sound and vibration — how energy propagates, where it damps, and how rhythmic order is preserved or lost. The clinical anchor for the term is eurhythmia, the regular rhythm of healthy physiology, against which arrhythmia and other disorder are defined. Level 42 AI originated the quantitative music-of-the-body framing and remains its principal developer.

Evidence: definitional + instrumentation described in peer-reviewed literature

This is the platform's central research programme, and it should be described as such. The hypothesis is that progressive stiffening of connective tissue and declining mechanotransductive responsiveness produce measurable changes in how vibration propagates and damps through the body, and that these changes track functional aging more closely than chronological time does. What is established is that the instrument measures mechanical signal with high bandwidth and repeatability. What is not yet established, in peer-reviewed form, is a validated mapping from those measurements to a biological-age estimate with known accuracy against accepted reference clocks. Treat current outputs as functional-state indices under validation, not as a certified biological age.

Evidence: research hypothesis — validation in progress

They measure a different class of thing and are best understood as orthogonal rather than substitutive. Inflammatory markers report circulating concentrations that reflect processes already underway, are subject to assay turnaround, and are typically sampled episodically. Vibroacoustic measurement reports mechanical function in real time and can be repeated at no marginal cost. The reasonable claim is complementarity: a functional layer that indicates whether something is changing now, alongside a biochemical layer that characterizes what is changing. Head-to-head comparative accuracy against standard panels for specific outcomes has not been published and should not be asserted.

Evidence: mechanistic reasoning; comparative accuracy unpublished

Published Vibrome® respiratory-infection research reports vibroacoustic measurements that distinguished RT-PCR-confirmed infection in participants without symptoms or with limited symptoms. Detection in an asymptomatic participant does not by itself establish how far in advance of symptoms a signal changes. Extending these findings to chronic degenerative change or aging remains a hypothesis, not a demonstrated capability. Disease-specific use remains investigational, and these measurements do not replace confirmatory testing.

Evidence: published respiratory-infection findings; chronic-decline and lead-time validation outstanding.

The imPulse UNA transducer is specified at 0.1 Hz to 24 kHz: infrasound, the audible band, and a limited extension above 20 kHz. Wider bandwidth allows mechanical features outside the range of ordinary listening to be investigated. Device bandwidth, signal quality and validated clinical usefulness are separate properties. The 0.1 Hz to 160 kHz figure refers to the ambient-sensor platform on the roadmap, not to this transducer. Confirm the specification for the particular device and configuration being evaluated.

Evidence: imPulse UNA device specification; ambient-platform roadmap distinguished from the imPulse UNA range.

Guided acquisition takes approximately 20 to 60 seconds at each auscultation point, across an 11-point protocol. The chestpiece is placed over light clothing, with controlled posture and breathing instructions where required; no gel or sample collection is involved. A measurement at one point is not a complete scan. Total appointment time depends on the number of positions, setup, signal-quality checks and repeat measurements. Confirm the applicable protocol with the research or deployment team.

Evidence: imPulse UNA acquisition protocol and product description; total scan time is protocol-dependent.

The research approach uses Structural Machine Learning: models are synthesized as interpretable programs within a domain-specific grammar, with penalties for unnecessary complexity and feature count. The published COVID-19 case study compares this approach with deep-learning baselines on small samples and reports better out-of-sample performance for the structural approach in that evaluation. This supports investigating constrained models when data are limited; it does not establish that they outperform deep learning in every setting. Performance must be assessed on independent data for each intended use.

Evidence: peer-reviewed methodological case study. GECCO 2023 publication.

The mechanism has two components. Mechanically, the body is a coupled fluid–structure system: airflow and blood flow interact with pressure gradients, viscosity and tissue forces across deformable networks, so changes in tissue stiffness, perfusion distribution or airway impedance alter how energy propagates and dissipates. Biologically, mechanotransduction is the established process by which cells convert mechanical stimuli into biochemical signalling, which provides a principled reason to expect mechanical state and cellular state to be coupled. Tensegrity is used as the structural model connecting the two. Readers should note that tensegrity is a well-established conceptual framework in tissue mechanics rather than a standardized measurable quantity, and claims should be framed accordingly.

Evidence: established mechanism + platform-specific model

In principle yes, and this is a structural advantage rather than a marketing claim: fluid mechanics and wave propagation are species-general, so the governing physics of the measurement does not change between a rodent, a dog, a pig and a human, even though anatomy, scale and normative ranges do. This supports a non-terminal translational evidence engine in which signatures can be developed and cross-validated across in vitro, rodent, companion-animal, livestock and human contexts. Species-specific normative baselines, acquisition geometry and validation are required for each target species and cannot be assumed to transfer.

Evidence: physical first principles; species-specific validation required

Studies can evaluate defined aspects of function and health over shorter periods, while keeping those outcomes separate from claims about longer life. Repeated mechanical measurements offer a candidate intermediate layer for observing whether function changes during an intervention. To become a validated surrogate for a clinical outcome, a measurement needs evidence that its changes reliably predict that outcome in the relevant context. That validation remains outstanding for Vibrome®-derived longevity indices. A change in a candidate measurement alone is not proof of improved healthspan or survival.

Evidence: research rationale; longevity-surrogate validation outstanding.

Primary care is largely organized around discrete encounters and around detecting disease once present. Consumer wellness operates continuously but often without objective validation or scope discipline. WELLcare is the space between: continuous objective measurement, prioritized rather than exhaustive action, and explicit referral boundaries with licensed clinical care. Level 42 AI's protocols are organized as point-of-care for clinical settings, point-of-need for field and community deployment, and point-of-habitat® for ambient measurement in the environments where people actually live. The distinguishing commitment is that WELLcare output is function-first and is never presented as diagnosis.

Evidence: category definition

The constraint is not data volume. An engaged consumer may already carry several body-worn devices and interact with several more at home daily, and the resulting problem is prioritization, not scarcity: nobody can act on every health variable simultaneously, and which variable matters depends on the individual's sleep, obligations, diet, stress and even season. The argument for an additional measurement is therefore specific — it must be orthogonal to what wearables already capture and must resolve to a prioritized action. Vibrome® measurement is orthogonal in that it reports mechanical and fluid-dynamic function rather than movement-derived or optically-derived proxies. If it cannot produce a ranked, actionable output, it adds to the noise rather than reducing it.

Evidence: category reasoning

Coaching converts information into behavior, and its two scarce inputs are prioritization and accountability. An objective, repeatable measurement serves both: it identifies which variable is moving, and it creates a shared factual record that a coach and client review together, replacing self-report with observation. Practitioners consistently describe accountability — knowing a specific conversation is coming — as the mechanism that changes behavior where information alone does not. Measurement makes that conversation concrete. The platform's role stops at measurement and prioritization: it does not diagnose, prescribe, or substitute for licensed clinical judgment, and a responsible coaching workflow includes explicit referral triggers.

Evidence: practice reasoning; scope boundary stated

The risk is presenting a plausible mechanism or promising early result as an established benefit. A passive measurement platform does not intervene pharmacologically, but inaccurate or overinterpreted outputs can still lead to poor decisions. The appropriate response is to distinguish published findings, preliminary internal analyses and research hypotheses. Published respiratory-infection findings do not validate aging measurement, and neither investment nor recognition substitutes for outcome evidence. Vibrome®-derived healthspan and biological-age measures remain under validation.

Evidence: stated evidence policy and research limitations.

The Wellthbeing dashboard connects to compatible devices through WebUSB and WebHID in supported Chromium-based browsers. This approach can reduce the need for a separate native application. It still depends on compatible hardware, browser and operating-system support, and the deployment's connectivity requirements. WebUSB requires a secure context and user permission to grant device access; permission does not necessarily need to be granted again for every session. Confirm the supported configuration before deployment.

Evidence: product architecture; browser requirements documented in the WebUSB documentation.

The platform is designed around cryptographically authenticated device sessions and to support regional data residency. A deployment assessment should verify authentication, encryption, access controls, retention, hosting region and processor agreements for the actual configuration. These design intentions do not establish HIPAA, GDPR or other compliance. Request the applicable assessment, its date and scope, and the relevant contracts before processing sensitive data. Network identifiers used for authentication also need to be addressed in the deployment's privacy documentation.

Evidence: platform architecture; deployment controls and compliance assessment require confirmation.

Real-time inference can help an authorized study team identify poor recordings or site-specific acquisition drift during enrollment rather than after database lock. Model development and confirmatory evaluation must remain separate: access to unblinded results needs controlled roles, and model selection, locking and analysis procedures should be specified in advance. Updating a model against confirmatory test outcomes would compromise an independent evaluation. This is an operational design proposal, not evidence that the platform has already improved trial outcomes.

Evidence: operational design; study governance and prospective evaluation required.

Validation compares a locked model with an independent reference standard appropriate to the target condition, using data not used for model development. Published Vibrome® respiratory research uses RT-PCR-confirmed infection as its comparator. Evaluation should report cohort characteristics, sensitivity, specificity, confidence intervals and the handling of unusable recordings. A held-out evaluation reduces leakage risk but does not make leakage impossible. Each new indication needs its own appropriate reference standard and validation; respiratory-infection findings cannot establish biological-age accuracy.

Evidence: published respiratory methodology; indication-specific validation required.

The modelling approach uses constrained model grammars, complexity penalties, invariance objectives, held-out data and classifier ensembles. These mechanisms aim to reduce reliance on patterns that exist only in the training sample. The Structural Machine Learning case study provides evidence within its evaluated cohort, rather than a general guarantee of robustness. Patient-level data separation, independent external testing and evaluation across sites and acquisition conditions are needed to assess whether a model generalizes to its intended population.

Evidence: peer-reviewed methodological case study. GECCO 2023 publication; external validation remains important.

Vibrome® research evaluates thresholds under simulated disease-prevalence scenarios. Threshold selection changes the balance between missed cases and false positives, while prevalence affects predictive values and expected downstream testing demand. High overall accuracy in a low-prevalence setting can occur even when sensitivity is inadequate, so accuracy alone is insufficient. Simulated prevalence adjustment is not evidence of successful deployment in every population. Any proposed threshold needs prospective validation, calibration review and appropriate clinical and regulatory governance for the intended use.

Evidence: prevalence simulation in Vibrome® research; local prospective validation outstanding.

Tuberculosis screening and triage is an investigational application studied under NCT04923958. Care-cascade research describes losses caused by delayed care-seeking, inaccessible testing and incomplete follow-up. A repeatable passive measurement is being investigated as a way to support assessment closer to where people live. A trial registration documents a study, not demonstrated diagnostic performance. Published viral respiratory-infection results do not establish TB accuracy, and suspected TB still requires an appropriate clinical workup and confirmatory testing.

Evidence: registered study and care-cascade systematic review; Vibrome® TB application investigational.

Severity assessment is a research objective alongside disease detection. Preliminary internal severity analyses have not yet been published for independent review of their cohort, reference standard or uncertainty. These analyses should not be treated as established performance. A severity measure needs independent validation against defined clinical assessments and relevant outcomes. Detecting a disease-associated pattern does not automatically establish severity, prognosis or the appropriate treatment.

Evidence: preliminary internal work; peer-reviewed severity validation outstanding.

Repeated functional measurements could be investigated for tracking recovery and identifying changes that warrant clinical review after treatment. Post-TB surveillance is proposed as a research direction, including recovery, residual impairment and possible recurrence. These are distinct outcomes and each needs a defined comparator and prospective evaluation. Vibrome® measurements have not been established as a recurrence test or a replacement for scheduled follow-up, microbiological testing or clinical assessment.

Evidence: research rationale and proposed application; post-treatment clinical utility unestablished.

Contact Level 42 AI to discuss the intended research or wellness use, device configuration, onboarding and current availability. Ask about the Founding Partner programme, including its current allocation, price and update commitments. Request a written quotation specifying hardware, software access, support, updates, recurring charges and permitted uses. Participation or purchase does not establish diagnostic authorization or validate a longevity endpoint.

Evidence: current pricing and contract terms are confirmed in a written quotation. Contact Level 42 AI.

A practice should assess purchase cost, software and support charges, staff time, setup, repeat recordings, expected utilization and any revenue assumptions. Acquisition needs no reagents or sample collection, but absence of consumables does not make the total cost of a scan zero. No published ROI study or independent health-economic evaluation is available yet. Reimbursement, savings, payback periods and replacement of existing tests should not be assumed. A commercial decision needs a documented model using the practice's own costs and intended workflow.

Evidence: economic rationale; ROI and reimbursement unestablished.

Level 42 AI’s strategy is to develop commercial research, longevity and veterinary applications whose revenue could help sustain lower-margin global-health deployments. That is a proposed funding model, not proof of profitable operations or a completed transition. Assessing it requires evidence of customer demand, repeat use, delivery costs, support capacity and the economics of the intended global-health programmes. Commercial adoption and scientific validation are separate milestones; neither should be inferred from a strategy statement.

Evidence: company strategy statement; commercial viability not yet established.

Companion-animal, equine and livestock applications could investigate respiratory mechanics and repeated functional measurements in species that cannot report symptoms verbally. Shared physical principles motivate these applications, but anatomy, behavior, acquisition geometry and normal ranges differ across species. Each intended use requires species-specific validation and appropriate veterinary oversight. These applications are in development; animal results cannot be assumed to validate a human clinical use, and applicable requirements must be assessed for each market.

Evidence: applications in development; species-specific validation required.

The platform can be assessed against human evidence, scientific novelty, scalability, safety, data governance and commercial feasibility, while separately asking whether it has demonstrated an effect on healthspan-relevant outcomes. This is a self-assessment, not an independent finding.

Evidence: criteria self-assessment.

Signal quality is affected by body habitus, breathing pattern, comorbid lung disease, motion and ambient noise, and mitigation depends on standardized acquisition, repeated-measurement averaging, automated quality control and emphasis on system-level rather than isolated local features. Published performance is for acute respiratory infection; transfer to chronic degenerative processes and to biological-age estimation is hypothesized, not demonstrated. Normative reference ranges across age, sex, body composition and ancestry are still being established, which limits the interpretability of a single measurement relative to a within-person trajectory. Independent external replication by groups unaffiliated with the developer has not yet been published, and that is the single most valuable missing piece of evidence.

Evidence: stated limitations