Did you know?
Bone looks white on an X-ray because dense bone absorbs much of the radiation, while air absorbs little and looks black.
Source: RadiologyInfo.org (RSNA & ACR) (opens in a new tab)
How medical imaging works
Decision support for diagnostic labs
A second reader for every study. The doctor stays in the loop.
Your radiologists read first. Our AI then analyses the same study independently and, when it sees something worth a second look, shows its reasoning. The radiologist accepts or overrides — and always makes the final call.
- Radiologist reads first
- Reasoning you can inspect
- Every decision logged
Doctor-in-the-loop
Six steps, one principle: the physician decides
The second reader is designed around how radiologists already work. It never sees the study first, never signs a report, and never acts without a clinician's decision.
Step 1 of 6 · System
Study arrives
A new imaging study reaches the lab's existing worklist from PACS. Nothing changes about how it is acquired or routed.
StudyRadiologistSecond readerDecisionAudit
The AI never signs a report. Only the radiologist can finalise a decision.
Radiology second reader
When it disagrees, it shows you why
A flag on its own isn't useful. Each one arrives with the evidence behind it, laid out so a radiologist can confirm or dismiss it quickly and confidently.
- Bounding boxes and heat regions show exactly where the model looked.
- Plain-language reasoning explains what prompted the flag.
- Measurements are shown for the reader to verify, not to accept blindly.
- Accept or override in one step; overrides capture a short rationale.
Second reader — reasoning view
Read comparison
DisagreementRadiologist
No focal finding recorded in Region A.
Second reader
Localised change suggested in Region A.
Reasoning
- 1. Texture in the boxed region differs from the surrounding area.
- 2. The heat region shows where the model focused its attention.
- 3. Measured extent is shown for the reader to verify.
Awaiting the radiologist's decision.
Illustrative example — not real patient data.
Genomic health profile
Whole genome sequencing summary
Prepared for physician review
Variant summary
| Classification | What the reviewer sees | Status |
|---|---|---|
| Flagged for physician review | Evidence summary, inheritance and references | Review required |
| Uncertain significance | Current evidence and reclassification notes | Clinician judgement |
| Likely benign | Listed for completeness | For information |
Health profile
Summary of relevant findings, written for clinicians.
Diet & lifestyle
Guidance considerations for discussion with the patient.
Treatment suggestions
Options for the physician to review — not a prescription.
Interpretation and any clinical action rest with the ordering physician.
Illustrative example — not real patient data.
Genomics
From sequence to a report physicians can act on
Whole genome sequencing and targeted mutation analysis, organised into clear health profiles with diet and lifestyle guidance and treatment-plan suggestions — all prepared for physician review, never sent as instructions.
Lab operations dashboard
Today
Awaiting read
Ordered by priority
Second read complete
Ready for sign-off
Flags to review
Reasoning attached
Signed
Logged to audit trail
For labs
Built to fit the lab you already run
We work with diagnostic labs as partners. The second reader sits alongside your PACS, RIS and LIS, follows your reading protocols, and gives lab leadership a clear view of flags, decisions and follow-ups.
- No change to how studies are acquired or routed.
- Operational dashboards for worklists, flags awaiting review and sign-off.
- A partnership model with clinical onboarding and ongoing support.
Trust & safety
Accountability is designed in, not added on
Clinical AI earns trust slowly. We build for it with clear roles, inspectable reasoning and a record of every decision.
Physicians decide
The software offers decision support only. Every finding is accepted, edited or overridden by a qualified clinician, and only a clinician can sign a report.
Transparent reasoning
When the second reader raises a flag it shows its work: the region, the overlays and a plain-language explanation, so the reader can judge it on its merits.
Full audit trail
Reads, flags, decisions and rationales are recorded in an append-only log that supports clinical governance and quality review.
Privacy by design
Data handling is designed around minimisation, access control and encryption, and is configured to meet each partner lab's own policies and obligations.
Audit trail
Study received
09:02 · System
Routed from PACS to the reading worklist
Radiologist read submitted
09:15 · Radiologist
Initial impression recorded before any AI output was shown
Second reader analysis completed
09:15 · Second reader
Independent analysis of the same study
Disagreement flagged
09:16 · Second reader
Reasoning, overlays and region references attached
Radiologist decision
09:21 · Radiologist
Flag reviewed; decision and written rationale captured
Report signed
09:24 · Radiologist
Final report signed by the radiologist
Each record is append-only and attributable.
Illustrative example — not real patient data.
Science
The science behind Clarity
Short, cited explainers on imaging and genomics. Every fact links to a source we actually read.
Did you know?
On the CT scale, water is defined as 0 Hounsfield units and air as −1000. Readers “window” this range to see different tissues.
Source: AMIA Annual Symposium Proceedings (opens in a new tab)
Reading a medical imageDid you know?
Human DNA has about 3 billion bases, and more than 99 percent of them are the same in all people.
Source: MedlinePlus Genetics (U.S. National Library of Medicine) (opens in a new tab)
DNA and the genomeDid you know?
The exons — the protein-coding pieces of DNA that exome sequencing reads — are thought to make up about 1 percent of the genome.
Source: MedlinePlus Genetics (U.S. National Library of Medicine) (opens in a new tab)
Sequencing explainedSee the second reader in your own workflow
Book a walkthrough with our team. We'll show how doctor-in-the-loop decision support fits alongside your existing systems, and answer your clinical, technical and governance questions.