Voxel-native morphometry, at mouse scale.

Tumour volume, bone microstructure and body composition, pulled straight from µCT voxels and rolled up to study, arm, animal and time-point.

Request access All modules

01 · Your unit of analysis

Built around how a study is actually run.

A clinical imaging tool hands you one patient at a time, because that is all a clinic needs. A study is not a stack of patients. It is arms, randomisation, time-points and a comparison you specified before the first scan. MiceX is built on that shape — every measurement lands in a study, in an arm, on an animal, at a time-point, and rolls up without anyone rebuilding the spreadsheet.

MiceX studies view — one cardiac cohort with three arms and thirteen animals — One cohort, three arms, thirteen animals — each arm carrying its own render, its own n, and its own statistics.

02 · What it measures

Six disease areas, measured from the voxels.

Cardiac is in active build. The other five carry a phase number rather than a ship date — each one a specification with a measurement set attached, not a claim about today.

MiceX disease-area coverage — cardiac active, oncology, bone, metabolic, NAFLD/MASH and pulmonary with phase labels — The measurement sets per area — BV/TV, Tb.N and Tb.Th for bone, %TGI and doubling time for oncology, LAA-950 for pulmonary, BAT volume and L3 skeletal muscle area for metabolic. The phase label on each card is the honest part: where the module actually is →

03 · What it costs you today

The measurement is not the expensive part. The analyst is.

Manual segmentation

5–15 min

per tumour, per scan, by hand.

A 200-scan study

≈ $5,000

of analyst time at $150/hr loaded.

MiceX

Scanner pace

the same segmentation, at scanner throughput.

“Average HCT-116 doubling time across all my historical efficacy studies” becomes one query, not a quarterly analyst project.

MiceX workflow wedges beside the phased roadmap — The workflow wedges sit next to the roadmap that dates them. The analyst-hours saving is live today; the compliance surface — audit trail per measurement, versioned re-analyses, Part 11 e-signatures — is designed for and scheduled in Phase 5. We do not claim GLP or 21 CFR Part 11 compliance today, and no part of MiceX is validated for a regulated environment yet.

04 · Endpoints, and how we prove them

We generate the cohort where the answer is known, then check we recover it.

Before real oncology data exists, the pipeline runs against a synthetic xenograft cohort built with the dose–response written in — vehicle, low, mid, high, a known doubling time, a known slowdown per arm. The endpoints it produces have to come back matching what was built in. This is the Prove stage of the loop, pointed at a module before its data arrives.

Synthetic xenograft study — day-28 tumour volume, T/C and %TGI per arm, then dose–response effect sizes — A synthetic xenograft cohort, generated with a known dose–response so the endpoint pipeline can be checked against an answer we already have. Simulated data — the proof is that the statistics recover what was built in, not a result about a compound.

How the loop works Read the methodology

05 · What’s real today

Where this module actually is.

SHIPPING

Cardiac pipeline end to end — 13 animals, 3 arms, real µCT at 40 µm

Study / arm / animal / time-point model

Per-animal measurement with cohort percentiles

Structured export with provenance

SPECIFIED, NOT BUILT

Oncology pipeline on real data (synthetic cohort today)

Bone, metabolic, NAFLD/MASH and pulmonary measurement sets

Pre-clinical intake — scanner-native import

Audit trail, versioned re-analyses, LIMS integration

Cardiac is wired end to end. Everything else has a specification and a place in the queue. We would rather show you the queue than the mockup.

06 · What you receive

Every animal, with its record attached.

Per-animal table — 13 animals with arm, training/validation split, blood pool, body volume, bounding box and HU — Thirteen animals, each with its arm, its training or validation split, and its measured values — blood-pool and body volume, bounding box, blood and myocardium HU.

Every animal carries its acquisition record and where it sits in its arm — percentiles are descriptive, not a classification.

07 · Who it’s for

Pharma drug discovery

Efficacy studies where the imaging endpoint is the readout and the analyst queue is the bottleneck.

Pre-clinical CROs

The same measurement across every sponsor’s study, with the record attached.

08 · Request access

Tell us what your study measures.

Model, arms and time-points — we will tell you plainly what MiceX measures today and what is still in the queue.