How CardioIQ Works

The logic behind it comes from a clinic, not a lab

Most AI health tools are built by engineers who license some medical content afterward. CardioIQ was built the other way around: the clinical reasoning came first, from a physician who reads exactly these panels every week, and the software was built to carry that reasoning at scale.

So every flagged value in your report can be traced back two ways: to a specific guideline, and to a physician whose name is actually on it.

Why "optimal" is a different question than "normal"

A standard lab report flags you against everyone else who got tested — the reference range is really just "where most people fall." Optimal is a different question: given your own risk profile, what level keeps your arteries on the safest trajectory over the next several decades? Three ideas shape how CardioIQ answers that:

• The bar moves with your risk. ESC/EAS guidelines set stricter cholesterol targets as cardiovascular risk climbs — so your "optimal" isn't a universal number, it's calibrated to you specifically.
• Some risk hides from a basic panel. ApoB counts the actual atherogenic particles; Lp(a) reflects inherited risk. Neither shows up on a routine cholesterol test, and either can change the picture entirely.
• It's about decades, not a single draw. Plaque builds from cumulative exposure — so getting a number lower, earlier, and keeping it there matters more than any one snapshot.

None of this is invented in-house — it's the same evidence base preventive cardiologists already use, made legible for your specific numbers. The full framework → · Why ApoB →

From your panel to your report: five steps

Nothing in between is a black box:

1. Your numbers go in. Your panel and relevant risk factors get structured into one consistent profile.
2. Each marker gets mapped to risk. A physician-built clinical knowledge graph links every value to what it actually means for cardiovascular risk.
3. The reasoning gets checked against evidence. Every link is anchored to a current guideline or a named trial — not a general "studies show."
4. The report is written like a clinician thinks. Finding → your optimal → why it matters → what could change it → what to ask your doctor → where CardioIQ's role ends.
5. It stays current. When guidelines or major trials update, the evidence base behind your report updates with them.

The sources this is built on

Every judgment traces to a named source, not a vague reference:

• ESC/EAS dyslipidaemia guidelines (2019, with the 2025 Focused Update) — LDL / ApoB targets by risk category, with risk assessed via SCORE2/SCORE2-OP
• EAS consensus statements — Lp(a), and ApoB as a preferred marker
• ACC/AHA primary-prevention guidance and the PREVENT risk equations
• ADA and KDIGO standards where cardiometabolic and kidney risk overlap
• Named cardiovascular outcomes trials, cited directly

When a guideline position changes, the report changes with it.

Where CardioIQ stops

Part of building this responsibly is being specific about its edges:

A report that's upfront about what it can't do is one worth trusting for what it does.