Case Study smartphone-based-recognition-of-heart-failure-by-means-of-microelectromechanical-sensors

2024 Release

Smartphone-Based Recognition of Heart Failure by Means of Microelectromechanical Sensors

François Haddad,Antti Saraste,Kristiina Santalahti,Mikko Pänkäälä,Matti Kaisti,Riina Kandolin,Piia Simonen,Wail Nammas,Kamal Jafarian Dehkordi,Tero Koivisto,Juhani Knuuti,Kenneth W. Mahaffey,Juuso Blomster

DOI: 10.1016/j.jchf.2024.01.022

Executive Summary

This multicenter study evaluated the feasibility of smartphone-based seismocardiography (SCG) and gyrocardiography (GCG) for detecting heart failure (HF) using embedded microelectromechanical sensors (MEMS). Algorithms derived from accelerometer and gyroscope signals achieved high diagnostic accuracy (AUC 0.95, sensitivity 85%, specificity 90%) across diverse clinical subgroups. The findings suggest that smartphone-based cardiac motion analysis could enable remote HF detection and monitoring with minimal hardware requirements.

“This study shows that smartphones can detect heart failure with high accuracy using built-in motion sensors, offering a simple and non-invasive way to monitor heart health remotely.”

Answer Machine Insights

Q: What was the diagnostic accuracy of the smartphone-based algorithms for HF detection?

The algorithms achieved an accuracy of 89%, with sensitivity of 85% and specificity of 90%.

Across both HF cohorts, the algorithms had an area under the receiver operating characteristic curve of 0.95 with a sensitivity of 85%, specificity of 90%, and accuracy of 89% for the detection of HF.

Q: How did gyroscope signals compare to accelerometer signals in HF detection?

Gyroscope signals provided significantly higher diagnostic accuracy than accelerometer signals.

The AUCs were significantly higher for the gyroscope signals compared with the accelerometer signals.

Key Results

The smartphone-based algorithms achieved an AUC of 0.95, sensitivity of 85%, and specificity of 90% for HF detection.
Gyroscope signals provided higher diagnostic accuracy compared to accelerometer signals, emphasizing the importance of rotational dynamics in cardiac motion analysis.