AI Transparency
How E-Labus AI works, its limitations, and what you should know
Important: AI Is Not Infallible
AI systems can and do make mistakes. While our technology is designed to provide helpful health insights, it is not perfect. Low-probability outcomes can and occasionally do occur. No AI system achieves 100% accuracy.
The analysis provided by E-Labus is for informational and educational purposes only. It does not constitute medical advice, diagnosis, or treatment recommendations. Always consult with a qualified healthcare provider before making any health decisions.
How E-Labus AI Works
Data Sources
Our AI analyzes data you provide: uploaded lab documents (PDFs, images), manually entered values, and data synced from connected healthcare providers via FHIR. We do not access external medical records without your explicit authorization.
AI Model Technology
We use large language models (LLMs) trained on medical literature and clinical knowledge. For extraction, we use computer vision models to read lab documents. For analysis, we use clinical reasoning models with deterministic settings (temperature: 0) to ensure reproducible results when the same data is analyzed.
Reference Ranges & Trends
Your values are compared against the reference ranges printed on your lab report and tracked over time so you can see how each marker changes. This is educational context, not a validated clinical tool — always review results with your provider.
Accuracy & Limitations
What Can Go Wrong
- OCR Extraction Errors: Text from scanned or photographed documents may be misread, leading to incorrect values.
- Unusual Biomarker Patterns: AI may misinterpret rare conditions or atypical presentations it hasn't encountered frequently.
- Missing Context: AI doesn't know your full medical history, current symptoms, medications (unless entered), or lifestyle factors.
- Low-Probability Events: Even with 95%+ accuracy, 1 in 20 analyses may contain significant errors. Plan accordingly.
Data Quality Matters
Garbage in, garbage out. The quality of AI analysis depends entirely on the quality of input data:
- • Clear, high-resolution document scans produce better results
- • Manually verify critical or unexpected values
- • Complete health profiles improve personalization accuracy
- • More historical data points enable better trend analysis
Confidence Scoring Explained
Each analysis includes a confidence score (0-100%) indicating how reliable the AI considers its output. Here's what different levels mean:
70%+ (High Confidence)
Sufficient data points, clear patterns, and consistent with known medical literature. Standard disclaimers apply.
40-69% (Moderate Confidence)
Some data gaps, unusual patterns, or conflicting indicators. Interpret with additional caution and consider professional consultation.
Below 40% (Low Confidence)
Significant data limitations or highly unusual values. Treat as preliminary only. Do not make health decisions based on this analysis without professional verification.
What E-Labus Does NOT Do
No Diagnosis
We identify patterns and flag potential concerns. We do not diagnose diseases or conditions. Only licensed healthcare providers can diagnose.
No Prescriptions
We share educational information only. We do not prescribe or recommend medications, supplements, or treatments. Consult your doctor.
Not a Replacement
E-Labus is a tool to help you understand your health data. It is not a replacement for regular check-ups, professional medical advice, or clinical judgment.
No Emergency Care
If you're experiencing a medical emergency, call emergency services immediately. E-Labus is not designed for acute or emergency situations.
What E-Labus Does Well
- Organize and visualize lab results from multiple sources in one place
- Track trends over time to spot patterns you might miss
- Provide educational context about what biomarkers mean
- Keep your medications and labs organized together for discussion with your doctor
- Help you prepare informed questions for healthcare appointments
- Securely share your data with healthcare providers you authorize
Questions?
If you have questions about our AI technology, data handling, or methodology, please contact us:
Convelabs Technology
Email: support@e-labus.com