CORE TEMP ESTIMATION
A health monitoring app that estimates core body temperature using only heart rate data — a non-invasive alternative to traditional thermometers, built in collaboration with Dr. Udayraj.
Role
App Developer & Algorithm Engineer
Client
IIT Bhilai — Dr. Udayraj
Year
2024
Category
Health Tech / Research
Tech Stack
The Problem
Measuring core body temperature traditionally requires invasive methods (rectal, esophageal probes) or expensive infrared equipment. Wearable devices track heart rate continuously but don't estimate temperature. The research question: can we reliably derive core temperature from heart rate variability alone?
The Approach
Implemented signal processing algorithms that analyze heart rate patterns — variability, trends, and circadian rhythms — to estimate core body temperature in real-time. The app connects to standard fitness wearables via Bluetooth, processes the heart rate stream, and displays temperature estimates with confidence intervals. Validated against clinical thermometer readings.
How it was built
Studied the physiological relationship between heart rate variability and thermoregulation — literature review and data collection with Dr. Udayraj.
Developed the signal processing pipeline in Python — filtering, feature extraction from HRV data, and temperature estimation models.
Ported the algorithm to Dart/Flutter for real-time mobile processing with minimal battery impact.
Built the wearable integration layer — Bluetooth LE connection to fitness bands and smartwatches for continuous heart rate streaming.
Validated accuracy against clinical-grade thermometers across different activity levels and ambient conditions.
Impact
Visuals
Wearable Health
Core temperature estimation from heart rate data — connecting standard fitness wearables to clinical-grade insights.