Back

Integration of Electrochemical Sensing and Machine Learning to Detect Tuberculosis via Methyl Nicotinate in Patient Breath

Jeppson, M. A.; Rasmussen, Z.; Castro, R.; Nalugwa, T.; Kisakye, E.; Mangeni, W.; Andama, A.; Jaganath, D.; Cattamanchi, A.; Mohanty, S. K.

2024-05-24 infectious diseases
10.1101/2024.05.23.24307746 medRxiv
Show abstract

Tuberculosis (TB) remains a significant global health issue; making early, accurate, and inexpensive point-of-care detection critical for effective treatment. This paper presents a clinical demonstration of an electrochemical sensor that detects methyl-nicotinate (MN), a volatile organic biomarker associated with active pulmonary tuberculosis. The sensor was initially tested on a patient cohort comprised of 57 adults in Kampala, Uganda, of whom 42 were microbiologically confirmed TB-positive and 15 TB-negative. The sensor employed a copper(II) liquid metal salt solution with a square wave voltammetry method tailored for MN detection using commercially available screen-printed electrodes. An exploratory machine learning analysis was performed using XGBOOST. Utilizing this approach, the sensor was 78% accurate with 71% sensitivity and 100% specificity. These initial results suggest the sensing methodology is effective in identifying TB from complex breath samples, providing a promising tool for non-invasive and rapid TB detection in clinical settings.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.