A real-time web-based telemedicine framework based on AI and IoMT for emergency triage and initial diagnostics: the...

Rapid medical decision-making for emergency and chronic conditions remains a global challenge, especially in under-resourced and remote settings. Traditional triage models often rely on narrowly focused algorithms or limited sensor inputs, which can hinder timely diagnosis and treatment.

This web-based telemedicine system helps with emergency triage and initial diagnosis by using organised clinical rules based on medical guidelines and approved by doctors.

TeleMedQuick integrates Internet of Medical Things (IoMT) devices with a rule-based expert inference engine comprising 76,229 clinical rules. These rules were developed through a combination of medical guideline reviews and direct consultations with certified emergency physicians. The system evaluates vital signs, symptoms, demographics, and patient history for conditions such as stroke, diabetes, hypertension, respiratory disorders, and heart attacks.

The system was evaluated on a medically annotated dataset of 750 patients under expert review. It achieved a triage accuracy of 99.1%, confirmed through expert validation and performance metrics, including F1-scores across all urgency levels. Rule design minimises symptom overlap and allows understandable, rapid decisions.

As an expert system, TeleMedQuick bridges the gap between IoMT sensing and clinical reasoning in telemedicine. It enables scalable, real-time triage and initial diagnostic support with validated transparency, making it suitable for prehospital care, especially in low-access or high-demand contexts.