File size: 13,349 Bytes
28f1212 f2c113d 5d53fbf f2c113d e684a6c f2c113d e684a6c f2c113d e684a6c f2c113d 28e85d4 f2c113d 06e876f 28e85d4 7cdc1a2 28e85d4 f2c113d 06e876f 28e85d4 7cdc1a2 28e85d4 f2c113d 06e876f f2c113d 06e876f f2c113d e684a6c 06e876f e684a6c 06e876f f2c113d 28e85d4 06e876f f2c113d e684a6c f2c113d e684a6c f2c113d e684a6c f2c113d | 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 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 | # [Track A: Baseline]
"""
Agent Orchestrator — the brain of the CDS Agent.
Controls the multi-step pipeline:
1. Parse patient data (MedGemma)
2. Clinical reasoning / differential diagnosis (MedGemma)
3. Drug interaction check (OpenFDA + RxNorm APIs)
4. Guideline retrieval (RAG over ChromaDB)
5. Conflict detection (MedGemma)
6. Synthesis into CDS report (MedGemma)
Each step is a tool call. The orchestrator manages state, handles errors,
and streams step updates to the frontend via a callback.
"""
from __future__ import annotations
import asyncio
import time
import uuid
from datetime import datetime
from typing import AsyncGenerator, Callable, Optional
from app.models.schemas import (
AgentState,
AgentStep,
AgentStepStatus,
CaseSubmission,
CDSReport,
)
from app.tools.patient_parser import PatientParserTool
from app.tools.clinical_reasoning import ClinicalReasoningTool
from app.tools.drug_interactions import DrugInteractionTool
from app.tools.guideline_retrieval import GuidelineRetrievalTool
from app.tools.conflict_detection import ConflictDetectionTool
from app.tools.synthesis import SynthesisTool
# Type for the callback that streams step updates
StepCallback = Callable[[AgentStep], None]
class Orchestrator:
"""
Orchestrates the clinical decision support agent pipeline.
Usage:
orchestrator = Orchestrator()
async for step_update in orchestrator.run(case):
# stream step_update to frontend
...
result = orchestrator.get_result()
"""
def __init__(self):
# Initialize tools
self.patient_parser = PatientParserTool()
self.clinical_reasoning = ClinicalReasoningTool()
self.drug_interaction = DrugInteractionTool()
self.guideline_retrieval = GuidelineRetrievalTool()
self.conflict_detection = ConflictDetectionTool()
self.synthesis = SynthesisTool()
# State
self._state: Optional[AgentState] = None
@property
def state(self) -> Optional[AgentState]:
return self._state
def _create_steps(self, case: CaseSubmission) -> list[AgentStep]:
"""Define the pipeline steps based on the case configuration."""
steps = [
AgentStep(
step_id="parse",
step_name="Parsing Patient Data",
tool_name="patient_parser",
),
AgentStep(
step_id="reason",
step_name="Clinical Reasoning",
tool_name="clinical_reasoning",
),
]
if case.include_drug_check:
steps.append(
AgentStep(
step_id="drugs",
step_name="Drug Interaction Check",
tool_name="drug_interactions",
)
)
if case.include_guidelines:
steps.append(
AgentStep(
step_id="guidelines",
step_name="Guideline Retrieval",
tool_name="guideline_retrieval",
)
)
if case.include_guidelines:
steps.append(
AgentStep(
step_id="conflicts",
step_name="Conflict Detection",
tool_name="conflict_detection",
)
)
steps.append(
AgentStep(
step_id="synthesize",
step_name="Synthesizing Report",
tool_name="synthesis",
)
)
return steps
async def run(self, case: CaseSubmission) -> AsyncGenerator[AgentStep, None]:
"""
Run the full agent pipeline. Yields step updates as they happen.
This is the main entry point. Each step is executed sequentially,
with state flowing from one step to the next. Steps that don't
depend on each other (drug check + guidelines) run in parallel.
If a critical step (parse, reason) fails, subsequent dependent
steps are marked as SKIPPED to avoid cascading errors.
"""
case_id = str(uuid.uuid4())[:8]
steps = self._create_steps(case)
self._state = AgentState(
case_id=case_id,
steps=steps,
started_at=datetime.utcnow(),
)
try:
# ── Step 1: Parse patient data ──
yield self._mark_running("parse")
step = await self._run_step("parse", self._step_parse, case.patient_text)
yield step
if step.status == AgentStepStatus.FAILED:
# Can't continue without patient profile — skip remaining steps
for skipped in self._skip_remaining_steps("parse"):
yield skipped
self._state.completed_at = datetime.utcnow()
return
# ── Step 2: Clinical reasoning ──
yield self._mark_running("reason")
step = await self._run_step("reason", self._step_reason)
yield step
if step.status == AgentStepStatus.FAILED:
for skipped in self._skip_remaining_steps("reason"):
yield skipped
self._state.completed_at = datetime.utcnow()
return
# ── Step 3 & 4: Drug check + Guidelines (parallel) ──
parallel_tasks = []
if case.include_drug_check:
yield self._mark_running("drugs")
parallel_tasks.append(("drugs", self._step_drug_check))
if case.include_guidelines:
yield self._mark_running("guidelines")
parallel_tasks.append(("guidelines", self._step_guidelines))
if parallel_tasks:
results = await asyncio.gather(
*[self._run_step(sid, fn) for sid, fn in parallel_tasks],
return_exceptions=True,
)
for result in results:
if isinstance(result, Exception):
# Log but don't fail — graceful degradation
pass
else:
yield result
# ── Step 5: Conflict Detection ──
if case.include_guidelines:
yield self._mark_running("conflicts")
yield await self._run_step("conflicts", self._step_conflict_detection)
# ── Step 6: Synthesis ──
yield self._mark_running("synthesize")
yield await self._run_step("synthesize", self._step_synthesize)
self._state.completed_at = datetime.utcnow()
except Exception as e:
# Mark remaining steps as failed
for step in self._state.steps:
if step.status == AgentStepStatus.PENDING:
step.status = AgentStepStatus.FAILED
step.error = f"Pipeline aborted: {str(e)}"
raise
def _skip_remaining_steps(self, after_step_id: str) -> list[AgentStep]:
"""Mark all steps after after_step_id as skipped. Returns them for yielding."""
skipped = []
found = False
for step in self._state.steps:
if step.step_id == after_step_id:
found = True
continue
if found and step.status == AgentStepStatus.PENDING:
step.status = AgentStepStatus.SKIPPED
step.error = f"Skipped: prerequisite step '{after_step_id}' failed"
skipped.append(step)
return skipped
def _mark_running(self, step_id: str) -> AgentStep:
"""Mark a step as RUNNING and return it for immediate yielding."""
step = self._get_step(step_id)
step.status = AgentStepStatus.RUNNING
return step
async def _run_step(self, step_id: str, fn, *args) -> AgentStep:
"""Execute a single step, tracking status and timing."""
step = self._get_step(step_id)
step.status = AgentStepStatus.RUNNING
start = time.monotonic()
try:
await fn(*args)
step.status = AgentStepStatus.COMPLETED
except Exception as e:
step.status = AgentStepStatus.FAILED
step.error = str(e)
finally:
step.duration_ms = int((time.monotonic() - start) * 1000)
return step
def _get_step(self, step_id: str) -> AgentStep:
for step in self._state.steps:
if step.step_id == step_id:
return step
raise ValueError(f"Unknown step: {step_id}")
# ──────────────────────────────────────────────
# Step implementations
# ──────────────────────────────────────────────
async def _step_parse(self, patient_text: str):
"""Step 1: Parse raw patient text into structured profile."""
profile = await self.patient_parser.run(patient_text)
self._state.patient_profile = profile
step = self._get_step("parse")
step.input_summary = patient_text[:100] + "..." if len(patient_text) > 100 else patient_text
step.output_summary = f"Parsed: {profile.chief_complaint}, {len(profile.current_medications)} meds, {len(profile.lab_results)} labs"
async def _step_reason(self):
"""Step 2: Clinical reasoning over the structured patient profile."""
if not self._state.patient_profile:
raise RuntimeError("Patient profile not available — parse step must run first")
result = await self.clinical_reasoning.run(self._state.patient_profile)
self._state.clinical_reasoning = result
step = self._get_step("reason")
step.output_summary = (
f"{len(result.differential_diagnosis)} diagnoses, "
f"{len(result.recommended_workup)} recommendations"
)
async def _step_drug_check(self):
"""Step 3: Check drug interactions for current + proposed medications."""
if not self._state.patient_profile:
raise RuntimeError("Patient profile not available")
meds = self._state.patient_profile.current_medications
# Also include any medications proposed by the reasoning step
proposed_meds = []
if self._state.clinical_reasoning:
for action in self._state.clinical_reasoning.recommended_workup:
if "medication" in action.action.lower() or "prescribe" in action.action.lower():
proposed_meds.append(action.action)
result = await self.drug_interaction.run(meds, proposed_meds)
self._state.drug_interactions = result
step = self._get_step("drugs")
step.output_summary = f"{len(result.interactions_found)} interactions found"
async def _step_guidelines(self):
"""Step 4: Retrieve relevant clinical guidelines via RAG."""
if not self._state.clinical_reasoning:
raise RuntimeError("Clinical reasoning not available")
# Build query from the top diagnosis
top_dx = self._state.clinical_reasoning.differential_diagnosis
if top_dx:
query = f"{top_dx[0].diagnosis} clinical guidelines management"
else:
query = self._state.patient_profile.chief_complaint + " clinical guidelines"
result = await self.guideline_retrieval.run(query)
self._state.guideline_retrieval = result
step = self._get_step("guidelines")
step.output_summary = f"{len(result.excerpts)} guideline excerpts retrieved"
async def _step_conflict_detection(self):
"""Step 5: Detect conflicts between guidelines and patient data."""
result = await self.conflict_detection.run(
patient_profile=self._state.patient_profile,
clinical_reasoning=self._state.clinical_reasoning,
drug_interactions=self._state.drug_interactions,
guideline_retrieval=self._state.guideline_retrieval,
)
self._state.conflict_detection = result
step = self._get_step("conflicts")
n = len(result.conflicts)
if n == 0:
step.output_summary = "No conflicts detected"
else:
step.output_summary = f"{n} conflict(s) detected — {result.summary}"
async def _step_synthesize(self):
"""Step 6: Synthesize all tool outputs into a final CDS report."""
report = await self.synthesis.run(
patient_profile=self._state.patient_profile,
clinical_reasoning=self._state.clinical_reasoning,
drug_interactions=self._state.drug_interactions,
guideline_retrieval=self._state.guideline_retrieval,
conflict_detection=self._state.conflict_detection,
)
self._state.final_report = report
step = self._get_step("synthesize")
step.output_summary = "Clinical Decision Support report generated"
def get_result(self) -> Optional[CDSReport]:
"""Return the final report, if synthesis completed."""
if self._state:
return self._state.final_report
return None
|