Spaces:
Sleeping
Sleeping
File size: 14,910 Bytes
13d5ab4 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 |
"""
Analysis Synthesizer - Result Aggregation and Synthesis
Combines outputs from multiple specialized models
"""
import logging
from typing import Dict, List, Any, Optional
from datetime import datetime
logger = logging.getLogger(__name__)
class AnalysisSynthesizer:
"""
Synthesizes results from multiple specialized models into
a comprehensive medical document analysis
Implements:
- Result aggregation
- Conflict resolution
- Confidence calibration
- Clinical insights generation
"""
def __init__(self):
self.fusion_strategies = {
"early": self._early_fusion,
"late": self._late_fusion,
"weighted": self._weighted_fusion
}
logger.info("Analysis Synthesizer initialized")
async def synthesize(
self,
classification: Dict[str, Any],
specialized_results: List[Dict[str, Any]],
pdf_content: Dict[str, Any]
) -> Dict[str, Any]:
"""
Synthesize results from multiple models
Returns comprehensive analysis with:
- Aggregated findings
- Key insights
- Recommendations
- Risk assessment
- Confidence scores
"""
try:
logger.info(f"Synthesizing {len(specialized_results)} model results")
# Extract successful results
successful_results = [
r for r in specialized_results
if r.get("status") == "completed"
]
if not successful_results:
return self._generate_fallback_analysis(classification, pdf_content)
# Aggregate findings by domain
aggregated_findings = self._aggregate_by_domain(successful_results)
# Generate clinical insights
insights = self._generate_insights(
aggregated_findings,
classification,
pdf_content
)
# Calculate overall confidence
overall_confidence = self._calculate_overall_confidence(successful_results)
# Generate summary
summary = self._generate_summary(
classification,
aggregated_findings,
insights
)
# Generate recommendations
recommendations = self._generate_recommendations(
aggregated_findings,
classification
)
# Compile final analysis
analysis = {
"document_type": classification["document_type"],
"classification_confidence": classification["confidence"],
"overall_confidence": overall_confidence,
"summary": summary,
"aggregated_findings": aggregated_findings,
"clinical_insights": insights,
"recommendations": recommendations,
"models_used": [
{
"model": r["model_name"],
"domain": r["domain"],
"confidence": r.get("result", {}).get("confidence", 0.0)
}
for r in successful_results
],
"quality_metrics": {
"models_executed": len(successful_results),
"models_failed": len(specialized_results) - len(successful_results),
"overall_confidence": overall_confidence
},
"metadata": {
"synthesis_timestamp": datetime.utcnow().isoformat(),
"page_count": pdf_content.get("page_count", 0),
"has_images": len(pdf_content.get("images", [])) > 0,
"has_tables": len(pdf_content.get("tables", [])) > 0
}
}
logger.info("Synthesis completed successfully")
return analysis
except Exception as e:
logger.error(f"Synthesis failed: {str(e)}")
return self._generate_fallback_analysis(classification, pdf_content)
def _aggregate_by_domain(
self,
results: List[Dict[str, Any]]
) -> Dict[str, Any]:
"""Aggregate results by medical domain"""
aggregated = {}
for result in results:
domain = result.get("domain", "general")
if domain not in aggregated:
aggregated[domain] = {
"models": [],
"findings": [],
"confidence_scores": []
}
aggregated[domain]["models"].append(result["model_name"])
# Extract findings from result
result_data = result.get("result", {})
if "findings" in result_data:
aggregated[domain]["findings"].append(result_data["findings"])
if "key_findings" in result_data:
aggregated[domain]["findings"].extend(result_data["key_findings"])
if "analysis" in result_data:
aggregated[domain]["findings"].append(result_data["analysis"])
confidence = result_data.get("confidence", 0.0)
aggregated[domain]["confidence_scores"].append(confidence)
# Calculate average confidence per domain
for domain in aggregated:
scores = aggregated[domain]["confidence_scores"]
aggregated[domain]["average_confidence"] = sum(scores) / len(scores) if scores else 0.0
return aggregated
def _generate_insights(
self,
aggregated_findings: Dict[str, Any],
classification: Dict[str, Any],
pdf_content: Dict[str, Any]
) -> List[Dict[str, str]]:
"""Generate clinical insights from aggregated findings"""
insights = []
# Document structure insight
page_count = pdf_content.get("page_count", 0)
if page_count > 0:
insights.append({
"category": "Document Structure",
"insight": f"Document contains {page_count} pages with {'comprehensive' if page_count > 5 else 'standard'} documentation",
"importance": "medium"
})
# Classification insight
doc_type = classification["document_type"]
confidence = classification["confidence"]
insights.append({
"category": "Document Classification",
"insight": f"Document identified as {doc_type.replace('_', ' ').title()} with {confidence*100:.0f}% confidence",
"importance": "high"
})
# Domain-specific insights
for domain, data in aggregated_findings.items():
avg_confidence = data.get("average_confidence", 0.0)
model_count = len(data.get("models", []))
insights.append({
"category": domain.replace("_", " ").title(),
"insight": f"Analysis completed by {model_count} specialized model(s) with {avg_confidence*100:.0f}% average confidence",
"importance": "high" if avg_confidence > 0.8 else "medium"
})
# Data richness insight
has_images = pdf_content.get("images", [])
has_tables = pdf_content.get("tables", [])
if has_images:
insights.append({
"category": "Multimodal Content",
"insight": f"Document contains {len(has_images)} image(s) for enhanced analysis",
"importance": "medium"
})
if has_tables:
insights.append({
"category": "Structured Data",
"insight": f"Document contains {len(has_tables)} table(s) with structured information",
"importance": "medium"
})
return insights
def _calculate_overall_confidence(self, results: List[Dict[str, Any]]) -> float:
"""Calculate weighted overall confidence score"""
if not results:
return 0.0
confidences = []
weights = []
for result in results:
confidence = result.get("result", {}).get("confidence", 0.0)
priority = result.get("priority", "secondary")
# Weight by priority
weight = 1.5 if priority == "primary" else 1.0
confidences.append(confidence)
weights.append(weight)
# Weighted average
weighted_sum = sum(c * w for c, w in zip(confidences, weights))
total_weight = sum(weights)
return weighted_sum / total_weight if total_weight > 0 else 0.0
def _generate_summary(
self,
classification: Dict[str, Any],
aggregated_findings: Dict[str, Any],
insights: List[Dict[str, str]]
) -> str:
"""Generate executive summary of analysis"""
doc_type = classification["document_type"].replace("_", " ").title()
summary_parts = [
f"Medical Document Analysis: {doc_type}",
f"\nThis document has been processed through our comprehensive AI analysis pipeline using {len(aggregated_findings)} specialized medical AI domain(s).",
]
# Add domain summaries
for domain, data in aggregated_findings.items():
domain_name = domain.replace("_", " ").title()
model_count = len(data.get("models", []))
avg_conf = data.get("average_confidence", 0.0)
summary_parts.append(
f"\n\n{domain_name}: Analyzed by {model_count} model(s) with {avg_conf*100:.0f}% confidence. "
f"{'High confidence analysis completed.' if avg_conf > 0.8 else 'Analysis completed with moderate confidence.'}"
)
# Add insights summary
high_importance = [i for i in insights if i.get("importance") == "high"]
if high_importance:
summary_parts.append(
f"\n\nKey Findings: {len(high_importance)} high-priority insights identified for clinical review."
)
summary_parts.append(
"\n\nThis analysis provides AI-assisted insights and should be reviewed by qualified healthcare professionals for clinical decision-making."
)
return "".join(summary_parts)
def _generate_recommendations(
self,
aggregated_findings: Dict[str, Any],
classification: Dict[str, Any]
) -> List[Dict[str, str]]:
"""Generate recommendations based on analysis"""
recommendations = []
# Classification-based recommendations
doc_type = classification["document_type"]
if doc_type == "radiology":
recommendations.append({
"category": "Clinical Review",
"recommendation": "Radiologist review recommended for imaging findings confirmation",
"priority": "high"
})
elif doc_type == "pathology":
recommendations.append({
"category": "Clinical Review",
"recommendation": "Pathologist verification required for tissue analysis",
"priority": "high"
})
elif doc_type == "laboratory":
recommendations.append({
"category": "Clinical Review",
"recommendation": "Review laboratory values in context of patient history",
"priority": "medium"
})
elif doc_type == "cardiology":
recommendations.append({
"category": "Clinical Review",
"recommendation": "Cardiologist review recommended for cardiac findings",
"priority": "high"
})
# General recommendations
recommendations.append({
"category": "Data Quality",
"recommendation": "All AI-generated insights should be validated by qualified healthcare professionals",
"priority": "high"
})
recommendations.append({
"category": "Documentation",
"recommendation": "Maintain this analysis report with patient medical records",
"priority": "medium"
})
# Confidence-based recommendations
low_confidence_domains = [
domain for domain, data in aggregated_findings.items()
if data.get("average_confidence", 0.0) < 0.7
]
if low_confidence_domains:
recommendations.append({
"category": "Analysis Quality",
"recommendation": f"Lower confidence detected in {', '.join(low_confidence_domains)}. Consider manual review.",
"priority": "medium"
})
return recommendations
def _generate_fallback_analysis(
self,
classification: Dict[str, Any],
pdf_content: Dict[str, Any]
) -> Dict[str, Any]:
"""Generate fallback analysis when no models succeeded"""
return {
"document_type": classification["document_type"],
"classification_confidence": classification["confidence"],
"overall_confidence": 0.0,
"summary": "Analysis could not be completed. Document was classified but specialized model processing failed.",
"aggregated_findings": {},
"clinical_insights": [],
"recommendations": [{
"category": "Manual Review",
"recommendation": "Manual review required - automated analysis unavailable",
"priority": "high"
}],
"models_used": [],
"quality_metrics": {
"models_executed": 0,
"models_failed": 0,
"overall_confidence": 0.0
},
"metadata": {
"synthesis_timestamp": datetime.utcnow().isoformat(),
"page_count": pdf_content.get("page_count", 0),
"fallback": True
}
}
def _early_fusion(self, results: List[Dict]) -> Dict:
"""Early fusion strategy - combine features before analysis"""
pass
def _late_fusion(self, results: List[Dict]) -> Dict:
"""Late fusion strategy - combine predictions after analysis"""
pass
def _weighted_fusion(self, results: List[Dict]) -> Dict:
"""Weighted fusion strategy - weight by model confidence"""
pass
|