Clinically Diagnose Asthma and Monitor Its Severity Using an Ultrasensitive Chemiresistive Nitric Oxide (NO) Gas Sensor via Exhaled Breath Analysis Assisted by Pattern Recognition

Fractional
exhaled nitric oxide (FeNO) is widely recognized
as
a reliable biomarker for asthma. FeNO sensors can help diagnose asthma
and monitor its severity. In this study, an ultrasensitive chemiresistive
gas sensor, sensitive to the key breath biomarkers of asthmanitric
oxide (NO) and H2Swas fabricated using Ag-decorated
ZnO. The sensor exhibits detection limits of 5 ppb for NO and 50 ppb
for H2S, and it can discriminate 10 ppb NO and 60 ppb H2S from the exhaled breaths. Clinically, a total of 80 exhaled
breath samples were collected and tested, including 40 from asthma
patients (APs) and 40 from healthy control subjects (HCs). The AP
group was effectively distinguished from the HC group using a pattern
recognition algorithm (PCA), attributed to the sensor’s beneficial
cross-sensitivity to asthma biomarkers. A diagnostic model distinguishing
asthma from non-asthma was constructed using the support vector machine
(SVM) algorithm, achieving an overall accuracy, sensitivity, and specificity
of 0.81, 0.88, and 0.75, respectively. The area under the curve (AUC)
value for all subjects in the receiver operating characteristic (ROC)
curve was 0.92. The severity of asthma in three inpatients was monitored
using the clinical evaluation method of diurnal peak expiratory flow
(PEF) variation, alongside our sensor. The sensor’s response
values exhibited a strong correlation (r = −0.74
(p < 0.05)) with the diurnal PEF variation values.
To validate the sensor’s diagnostic capability, six breath
samples from both HCs and APs were tested simultaneously using our
sensor and a commercial electrochemical NO sensor utilized clinically.
With r = −0.98 (p < 0.05)
and R2 = 0.94, a strong linear relationship
between two types of response values was observed, confirming the
sensor’s accuracy and reliability in detecting NO concentrations
in exhaled breath. Theoretical adsorption models of NO on the surface
of the sensor were constructed using DFT calculations to elucidate
the mechanisms driving the sensor’s ultrasensitivity. Overall,
the sensor demonstrates a significant potential for use in clinical
practice to diagnose asthma and monitor its severity.