One patient. Two independent routes. Three tiers of evidence.
ACI is a prognostic Clinical Decision Support System (CDSS): computational, deterministic and built on triangulation. A reading is only useful at the bedside when the clinician can see where it came from, so ACI computes every condition twice, by two different methods, and shows both.
A prognostic Clinical Decision Support System, built on triangulation.
ACI is prognostic software built on Clinical Decision Support System (CDSS) fundamentals. It is computational and deterministic, working on a triangulation model: the patient's condition is calculated by triangulating telemetry readings, biochemistry and imaging.
- Prognostic
- Computational
- Deterministic
- Advisory
Two instruments reading one patient.
Both routes triangulate telemetry, biochemistry and imaging, but by different methods. For each activated condition, ACI produces two readings and displays them side by side on a common 0–100 scale. They are never combined, averaged or reconciled into one value. Where they differ, the difference is shown as a stated quantity.
ACI prognostic model
Cadence · Continuous
A weighted average of telemetry at the basic level (each individual vital parameter) and the hybrid level (domain phenotypes and cross-domain phenotypes that combine vital parameters), plus the worst clinical score from biochemistry and from imaging, for the tests ordered against that specific condition. Each parameter is measured against the reference range for the patient's sex and age band.
Published clinical standard
Cadence · Every 4 hours
Triangulates the telemetry and clinical test items listed by the internationally approved and recognised clinical tool for the condition, such as GRACE, HEART or SOFA, using that tool's own published rule. It runs on a 4-hour cadence, the observation practice hospitals already follow, and the tool is named on screen against every reading. Standards include NEWS2, SOFA, GRACE 2.0, HEART, KDIGO, GOLD, GINA, CURB-65, NIHSS, CTCAE v5.0.
Three tiers of evidence. A fixed budget, closing at 100.
Route A's score for a condition is the Tier 1 composite plus the Tier 2 and Tier 3 contributions. Within a tier the worst finding governs; contributions are never summed, averaged or normalised across evidence classes. The tier structure belongs to Route A alone and is never applied inside Route B.
- Tier 1
Monitored parameters
Continuous physiological parameters from any compliant bedside monitoring source, scored as a composite within a fixed 80-point budget.
Up to 80 points - Tier 2
Biochemistry & laboratory
Graded Mild, Moderate, Marked or Acute on a disclosed sub-class ladder. The worst finding in the tier governs.
Up to 10 points - Tier 3
Diagnostic imaging
Graded on the same ladder as Tier 2. A study that excludes a structural cause counts as a valid finding scoring zero, not a missing result.
Up to 10 points
Three streams. One screen. Nothing quietly filled in.
- Stream A
General monitoring
Every monitored parameter shown continuously. 'Stable' only when all are within their normal range; otherwise the exact count of unstable parameters is shown.
- Stream B
Condition-specific patterns
Multi-parameter patterns within one clinical domain, such as the Cardiac Squeeze, read over a rolling window.
- Stream C
Cross-domain convergence
Patterns in different clinical domains that move together, read as one signal across domains.
No stream hides another. A missing value is shown as absent, never filled in.
Every reading is measured from a confirmed starting point.
NEWS2 is not valid in pregnancy or spinal cord injury. In those cases ACI computes no entry band, the physician sets the baseline on clinical assessment, and the record says so.
- Admit
The admission record and observations are entered, received from a connected monitoring source or entered by hand. An entry band is computed from NEWS2.
- Confirm
The admitting physician confirms the record and the baseline. An unconfirmed baseline is displayed as unconfirmed.
- Measure
Both routes start from the same baseline and land on the same 0–100 scale. Later readings measure travel from that point.
Fixed rules, versioned registers, no self-modification.
ACI is rule-based software. A displayed score changes because the patient's condition has changed, not because the software has changed.
Deterministic
Reference ranges, deviation curves, band boundaries and weights are fixed, human-authored and held in versioned registers. The same inputs give the same output, every time.
Reconstructable by hand
Any score can be rebuilt from the patient's values, the published clinical standard and the disclosed band tables.
Fail visible, never silent
A component that cannot deliver declares its state on screen. A lapsed stream is held, marked stale with the elapsed time, and then declared unavailable.
Never fabricate
No interpolation, no estimates, no values carried forward to fill a gap. An absent value is shown as absent, never as zero.
Δ = |A − B| / B × 100, shown while both routes are active. It is never capped, clamped or tuned, and neither route is adjusted to agree with the other.
Works with what the hospital already has.
ACI is hardware-agnostic. It needs no particular monitor, analyser or imaging system, is supplied with no hardware of its own, and reads data already generated in the patient's care.
Laboratory and electronic record data
Imaging findings from the hospital imaging system
Physiological data from any compliant bedside monitoring source
See ACI at the bedside.
Talk to the team behind ACI about a clinical demonstration, partnership, or due-diligence pack.