Spaces:
Sleeping
Sleeping
File size: 2,035 Bytes
f52ffa0 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 |
from typing import Any, Dict, List
def apply_rules(
patient_context: Dict[str, Any],
model_output: Dict[str, Any],
) -> Dict[str, Any]:
"""
Simple rule set over ECG model outputs.
"""
label = model_output.get("label")
score = float(model_output.get("score", 0.0))
hr = model_output.get("hr")
hr_int = int(hr) if hr is not None else None
explanations: List[str] = []
alert_level = "none"
age = patient_context.get("age")
has_prior_stroke = bool(patient_context.get("has_prior_stroke"))
# Arrhythmia label with high confidence
arrhythmia_labels = {"arrhythmia", "suspected_afib", "afib"}
if label in arrhythmia_labels and score >= 0.85:
alert_level = "escalate"
explanations.append("High-confidence arrhythmia detected.")
elif label in arrhythmia_labels and score >= 0.6:
alert_level = "notify"
explanations.append("Arrhythmia suspected; monitor closely.")
# Heart rate based thresholds
if hr_int is not None:
if hr_int >= 140:
alert_level = "escalate"
explanations.append("Severe tachycardia (hr >= 140).")
elif hr_int >= 120 and alert_level == "none":
alert_level = "notify"
explanations.append("Tachycardia (hr >= 120).")
# Patient risk factors escalate one level
if has_prior_stroke and alert_level != "none":
alert_level = "escalate"
explanations.append("Prior stroke history: escalate.")
elif has_prior_stroke and alert_level == "none":
alert_level = "notify"
explanations.append("Prior stroke history: monitor closely.")
if isinstance(age, int) and age >= 75 and alert_level == "notify":
alert_level = "escalate"
explanations.append("Age >= 75 with concerning signal: escalate.")
# If no triggers, add baseline explanation
if not explanations:
explanations.append("No rule-based alerts triggered.")
return {"alert_level": alert_level, "explanations": explanations}
|