pocketgull.app
Pocket Gull is a real-time clinical AI consult engine powered by Google Gemini. Features 10D multi-paradigm strategy, FHIR R4 export & 3D spatial anatomy.
First seen 2026-11-05 · Last seen 2026-11-05 · Source: webmcp.com
Answer 20 · Act 75 · Transact 3 · Score: Not yet scored
Tool inspector
run_sepsis_benchmark_evaluation
Evaluates empirical discrimination and alert precision of Pocket-Gull Conformal Sepsis-3 Engine against the Epic Sepsis Model (ESM) across 1,471,420 multi-center patients from MIMIC-IV v2.2 ICU (Beth Israel Deaconess) and CMS OMOP Inpatient Commons. Provides finite-sample 95% coverage, demonstrates an 89.9% reduction in false alarm burden (4.0 vs 39.9 alarms per 100 patient-days), and computes conformal prediction sets with epistemic abstention under borderline vitals.
Page: /
Input schema
| Property | Type | Required | Description |
|---|---|---|---|
cohort |
string | no | Target benchmark cohort (default: MULTI_CENTER_COMBINED). |
alpha |
number | no | Nominal significance level alpha between 0.01 and 0.15 (e.g. 0.05 for 95% coverage). |
vitals |
object | no | Optional patient vitals to evaluate through the conformal prediction & epistemic abstention engine. |
Raw JSON schema
{
"type": "object",
"properties": {
"cohort": {
"type": "string",
"enum": [
"MULTI_CENTER_COMBINED",
"MIMIC_IV_ICU",
"CMS_OMOP_INPATIENT"
],
"description": "Target benchmark cohort (default: MULTI_CENTER_COMBINED)."
},
"alpha": {
"type": "number",
"description": "Nominal significance level alpha between 0.01 and 0.15 (e.g. 0.05 for 95% coverage)."
},
"vitals": {
"type": "object",
"properties": {
"heartRate": {
"type": "number",
"description": "Heart rate in beats per minute."
},
"systolicBp": {
"type": "number",
"description": "Systolic blood pressure in mmHg."
},
"respiratoryRate": {
"type": "number",
"description": "Respiratory rate in breaths per minute."
},
"temperatureC": {
"type": "number",
"description": "Core body temperature in Celsius."
},
"lactateMmolL": {
"type": "number",
"description": "Serum lactate in mmol/L."
},
"wbcCount": {
"type": "number",
"description": "White blood cell count (x10^3/uL)."
}
},
"description": "Optional patient vitals to evaluate through the conformal prediction & epistemic abstention engine."
}
}
}
Similar websites
webmcp-six-lac.vercel.app
AI Navigation Layer: Model Context Protocol (MCP) Semantic DOM integration for intelligent, safe, and observable web navigation.
ghqsandbox.net
Interactive math and physics tools an AI agent can drive directly through WebMCP.
app.formative.com
Formative classroom assessment and learning platform
seqcraft.onrender.com
Inspect plasmids, simulate PCR and restriction digests, compare constructs, and prepare cloning proposals in one exact, private workspace.