1. Intended Use
prajnaScan is designed as a clinical workflow support tool for use by licensed healthcare professionals in outpatient clinical settings. Its intended functions are:
- Structured symptom capture and patient intake at the point of arrival.
- AI-assisted urgency scoring and triage prioritisation.
- Generation of structured pre-diagnosis notes for review by the consulting physician.
prajnaScan is not intended for patient self-diagnosis. Raw AI outputs are never delivered directly to patients — only clinician-reviewed and approved summaries may be shared with patients at the physician's discretion.
Current scope: symptom triage and structured patient history. Multimodal health assessments (camera or sensor inputs) are not yet part of the product and will be introduced only after independent clinical validation.
2. Clinical Oversight
Every AI-generated output from prajnaScan — including triage scores, red-flag alerts, and pre-diagnosis notes — is subject to the following oversight requirements:
- Physician sign-off required: No clinical action may be taken based solely on a prajnaScan AI output. All outputs must be reviewed and confirmed by a licensed medical professional before guiding any diagnostic or treatment decision.
- Doctor-only access: Pre-diagnosis notes are accessible only in the doctor's consultation view. Front-desk staff, ward assistants, and patients cannot access these outputs.
- Audit trail: All AI outputs and physician decisions are logged with timestamps to support clinical governance and quality review.
3. Accuracy & Limitations
Current Capabilities
prajnaScan's triage engine operates on structured symptom questionnaires and rule-based clinical logic, augmented by ML-derived urgency scoring. It is validated against primary care presentations in Tier 2 & 3 clinic settings.
Known Limitations
- The system does not include physical examination findings unless manually entered by clinical staff.
- Camera, sensor, or biometric inputs are not supported in the current version.
- Performance may be lower for rare presentations, atypical symptom clusters, or conditions outside primary care scope.
- Outputs should always be interpreted alongside clinical history, examination, and professional judgment.
Confidence Thresholds & Escalation
The system flags outputs with a confidence score. Low-confidence outputs trigger explicit escalation notices in the doctor view, indicating that additional clinical assessment is required before acting on the AI suggestion.
4. Patient Safety Commitments
- Accuracy testing: Triage models are benchmarked against labelled clinical datasets before deployment and reviewed quarterly with pilot clinic partners.
- Bias monitoring: We actively monitor for differential performance across age, gender, and geography, and flag any detected disparities for clinical review.
- Safety-first flags: Red-flag conditions (e.g., suspected cardiac emergencies, sepsis indicators) are always surfaced at the highest priority regardless of overall triage score.
- No silent failures: If the system cannot process an intake with sufficient confidence, it explicitly instructs clinical staff to route the patient directly to the doctor without an AI-generated note.
5. Regulatory Status
prajnaScan currently operates as a clinical decision support software tool and does not fall within the definition of a Class B or higher medical device under the Medical Devices Rules 2017 (India). It functions as a software workflow aid that presents information to a licensed physician — all clinical decisions remain with the physician.
As prajnaScan's capabilities expand (including potential multimodal assessments), we will proactively evaluate applicable regulatory pathways with India's Central Drugs Standard Control Organisation (CDSCO) and pursue any required approvals before launching new diagnostic-adjacent features.
6. Disclaimer
prajnaScan is not a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of a qualified healthcare provider with any questions regarding a medical condition. Never disregard professional medical advice or delay in seeking it because of information generated by prajnaScan.
AlakhAgam India Pvt. Ltd. is not liable for clinical outcomes arising from:
- Clinical decisions made without physician review of AI outputs.
- Use of prajnaScan outside its intended clinical workflow context.
- Modification of system outputs by clinic staff without authorisation.
7. Contact
For clinical safety inquiries, adverse event reporting, or questions about our validation protocols:
AlakhAgam India Pvt. Ltd. — Clinical Safety
Email: clinical@alakhagam.com
Website: www.alakhagam.com