Noninvasive Total Cholesterol Level Measurement Using an E-Nose System and Machine Learning on Exhaled Breath Samples
2024

Measuring Cholesterol Levels from Breath Using E-Nose and Machine Learning

Sample size: 151 publication Evidence: moderate

Author Information

Author(s): Paleczek Anna, Grochala Justyna, Grochala Dominik, Słowik Jakub, Pihut Małgorzata, Loster Jolanta E., Rydosz Artur

Primary Institution: AGH University of Krakow, Faculty of Computer Science Electronics and Telecommunications, Institute of Electronics

Hypothesis

Can an e-nose system combined with machine learning accurately measure total cholesterol levels from exhaled breath samples?

Conclusion

The study demonstrates that it is possible to develop a noninvasive device to measure total cholesterol levels from breath samples.

Supporting Evidence

  • The e-nose system achieved a mean absolute percentage error (MAPE) of 13.7% for the entire measurement range.
  • The model showed a better performance with a MAPE of 8% for cholesterol levels within the norm range (≤200 mg/dL).
  • Cholesterol levels in the study participants had a right-skewed distribution, with only a small number exceeding the norm.

Takeaway

Researchers created a device that can tell how much cholesterol is in your body just by smelling your breath, which is really cool and could help people stay healthy.

Methodology

Breath samples were collected from participants and analyzed using an e-nose system with various gas sensors, and machine learning algorithms were applied to predict cholesterol levels.

Potential Biases

Potential biases may arise from the self-reported data on health and lifestyle factors from participants.

Limitations

The study had a limited sample size of 151 participants, which may affect the generalizability of the results.

Participant Demographics

The study included 151 participants, with 92 women and 59 men, primarily over the age of 45.

Digital Object Identifier (DOI)

10.1021/acssensors.4c02198

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication