“High quality” is not a measurement. It is a conclusion that should follow from several independent measurements at different biological scales.
A procedure log can show whether a protocol was followed. CT can examine the whole patient. Electron microscopy can reveal synapses and cell membranes in tiny tissue samples.
None is sufficient alone. Together, they provide a much stronger account of what was preserved, what may have been damaged and what remains unknown.
The questions a quality system should answer
A serious quality assessment should separate at least six questions:
- How much warm and cold ischaemia occurred before effective protection?
- Did circulation and cooling reach the brain quickly enough?
- Did cryoprotective agent reach the target concentration, and was its distribution uniform?
- Did ice formation, oedema, shrinkage or thermal stress produce visible damage?
- Were neuronal membranes, synapses and other ultrastructural features retained?
- Can another expert inspect the data, methods and deviations?
Each question requires a different instrument. A good result on one axis cannot erase a failure on another.
Layer one: timeline and physiological conditions
The record begins before perfusion. Relevant timestamps include deterioration, dispatch, legal pronouncement, team arrival, cooling, cardiopulmonary support, surgery and the start of cryoprotective perfusion.
Temperature should be recorded over time and, where possible, at several body sites. A cooling curve contains far more information than one final reading.
The record should also include medications, ventilation, cardiopulmonary support and major interruptions. These variables affect circulation and the biological burden accumulated before vitrification.
They support an S-MIX estimate of ischaemic exposure. S-MIX is a model, not a tissue assay, so it should remain one component of the case profile.
Layer two: perfusion telemetry
During cryoprotective perfusion, the system can record pressure, flow, temperature, solution volume, vascular resistance and the concentration entering and leaving the patient.
Refractive index is commonly used as a practical proxy for cryoprotectant concentration. The terminal value shows what reached the sampled circuit outlet, not every microscopic region.
Pressure and flow curves can expose obstruction, leakage or changing vascular resistance. Visible oedema, brain shrinkage and surgical observations add anatomical context.
These are important process metrics. They cannot prove uniform tissue penetration because damaged or obstructed vessels may leave regional areas underprotected.
Layer three: standardized whole-body CT
CT is the main non-destructive post-preservation measurement currently available for an entire human patient.
It measures X-ray attenuation. With calibrated reference data, attenuation patterns can estimate cryoprotectant distribution and identify regions consistent with insufficient concentration, ice or gross structural change.
CT can also reveal large cracks, gas, major swelling or shrinkage and asymmetries between regions. It cannot resolve cells, membranes or synapses.
Why scanning temperature matters
CT values depend partly on physical density, and density changes with temperature. Published CT research has demonstrated that temperature can shift measured attenuation values.
Comparing scans taken at different temperatures can therefore confuse a measurement difference with a preservation difference.
Tomorrow.bio scans every cryopreserved human patient while the patient remains submerged in liquid nitrogen at approximately -196°C.
This gives every scan the same cryogenic endpoint and removes temperature as a major source of between-case variation in cryoprotective density estimates.
The scan protocol and calibration still need to remain consistent. Standardizing temperature does not remove every scanner, reconstruction or interpretation variable.
Tomorrow.bio reports that it is currently the only human cryopreservation organization applying this submerged, same-temperature CT protocol to every patient.
Its published quality programme reports CT scanning every patient at -196°C and colour-coding regions by inferred cryoprotectant penetration and ice.
The logical asymmetry matters. A poor CT result is evidence of a problem. A good result supports macroscopic cryoprotection, but cannot establish microscopic preservation.
Layer four: electron microscopy and brain ultrastructure
Electron microscopy reaches the scale that CT cannot.
It can show neuronal and glial membranes, axons, dendrites, myelin, mitochondria, capillaries, synaptic vesicles, synaptic clefts and postsynaptic densities.
It can also reveal ice-related voids, compressed tissue, membrane discontinuities, organelle swelling and other patterns consistent with osmotic or preparation injury.
This matters because long-term memory is thought to depend partly on persistent neural connectivity, ensembles of synaptic strengths and molecular or subcellular features.
A 2025 survey of neuroscientists found broad support for structural contributions to memory, while also showing uncertainty about the critical scale and sufficient detail.
Preserving recognizable synapses is therefore more informative than preserving only the brain’s gross shape. It still does not prove that every memory-relevant feature survived.
From a micrograph to a synaptic map
A single electron micrograph shows one extremely thin plane. It can demonstrate local structural quality, but it cannot map a circuit.
Serial-section electron microscopy or focused-ion-beam scanning electron microscopy can image successive layers and reconstruct a small tissue volume in three dimensions.
Researchers can then trace neurites and identify the synapses connecting them. This is the basis of synaptic-resolution connectomics.
Current methods can map small volumes, animal brains and selected human tissue volumes. They cannot non-destructively map every synapse in an intact cryopreserved human brain.
The distinction prevents an easy overclaim: electron microscopy can test whether sampled synapses and local circuits are structurally preserved. It cannot certify the complete human connectome.
Tomorrow.bio’s neural microsample programme
Tomorrow.bio asks members for separate consent to collect up to three small samples from selected brain or spinal cord regions for ultrastructural analysis.
The samples are intended to support electron microscopy and protocol improvement. Sampling is optional and is not required for a member to receive cryopreservation.
The published consent information states that samples are chosen from regions considered less critical for memory, identity and personality.
A microsample provides direct evidence at high resolution, but only for its location. Sampling strategy, processing artifacts and blinded scoring all affect how confidently it generalizes.
This is why whole-body CT and electron microscopy complement each other. CT supplies broad coverage; electron microscopy supplies local depth.
Layer five: cooling and thermal stress
After perfusion, the cooldown record should include sensor locations, temperatures, rates, holds and deviations through the glass-transition region.
Cooling too slowly can permit ice in underprotected regions. Cooling too aggressively can produce large thermal gradients and mechanical stress.
CT can detect some macroscopic fractures after cooldown. Microscopic cracking and molecular damage may remain below its resolution.
The final temperature therefore matters less than the complete thermal history and what the post-cooldown measurements show.
Layer six: histology, chemistry and functional research
Electron microscopy is not the only destructive research tool. Light microscopy can assess tissue architecture, while biochemical markers can investigate cell membranes, proteins and injury pathways.
In experimental tissue, rewarming followed by electrophysiology, metabolism or viability testing provides evidence that structure alone cannot.
These methods are valuable for validating and improving protocols. Most cannot be applied comprehensively to a human patient intended for continued storage.
Laboratory success also does not substitute for case data. Controlled animal tissue and a real human case can differ in ischaemia, disease, scale and transport history.
Storage quality is part of preservation quality
A strong initial procedure is not enough if long-term conditions are poorly controlled.
Quality assurance should include patient identity, chain of custody, storage position, liquid-nitrogen level, temperature monitoring, alarm history, vessel inspection and documented maintenance.
These checks do not improve the original ultrastructure. They establish whether the achieved state has remained continuously protected.
See how that responsibility is separated and maintained in the long-term storage facility.
Quality should remain a profile, not one score
A case can have low estimated ischaemia and uneven cryoprotectant delivery. It can have strong CT density yet disappointing ultrastructure in a sample.
It can also show excellent local electron microscopy while leaving unsampled regions uncertain.
The defensible output is a profile: timeline, thermal data, perfusion telemetry, CT distribution, ultrastructural evidence, procedural deviations, storage history and uncertainty.
That profile should be published in sufficiently detailed case reports, compared across cases and used to change protocols when recurring weaknesses appear.
The remaining limitations are discussed in technical challenges for high-quality preservation.
TL;DR: Cryopreservation quality requires several measurements: case timing, perfusion data, controlled cooling, cryogenic CT, neural electron microscopy when consented, and auditable storage records.
Estimate preservation quality
Explore how timing and circumstances can change the quality of a cryopreservation procedure.
Loading the interactive tool...
Further reading