Breaking barriers in Candida spp. detection with Electronic Noses and artificial intelligence

Abstract

The timely and accurate diagnosis of candidemia, a severe bloodstream infection caused by Candida spp., remains challenging in clinical practice. Blood culture, the current gold standard technique, suffers from lengthy turnaround times and limited sensitivity. To address these limitations, we propose a novel approach utilizing an Electronic Nose (E-nose) combined with Time Series-based classification techniques to analyze and identify Candida spp. rapidly, using culture species of C. albicans, C.kodamaea ohmeri, C. glabrara, C. haemulonii, C. parapsilosis and C. krusei as control samples. This innovative method not only enhances diagnostic accuracy and reduces decision time for healthcare professionals in selecting appropriate treatments but also offers the potential for expanded usage and cost reduction due to the E-nose's low production costs. Our proof-of-concept experimental results, carried out with culture samples, demonstrate promising outcomes, with the Inception Time classifier achieving an impressive average accuracy of 97.46% during the test phase. This paper presents a groundbreaking advancement in the field, empowering medical practitioners with an efficient and reliable tool for early and precise identification of candidemia, ultimately leading to improved patient outcomes.

Figure 1

Correlation graph of the results of applying the Nemenyi post-hoc test on the set of results for each model. In this type of graph, when it is farther from 1, the elements are more divergent; that is, they are statistically different.

Figure 2

Flow for sample identification and classification. First, control samples derived from the ATCC company are used to analyze and define the mycological diagnosis by the Laboratory of Medical Mycology. With this, the already cultivated species are identified and separated in Petri dishes. These cultures are then placed in the E-nose to identify the VOCs. With the collected data, pre-processing routines are executed to use the data already treated by the AI models. At the end, a species identification report is generated.

Figure 3

(a): Electronic Nose device used in experiments: (1) The Electronic Nose is packaged in a compact box; (2) The on-off switch activates it; (3) All connections are made of PTFE; (4) It has activated carbon filter and (5) PTFE filter; (6) Sample chamber also made of PTFE. (b): Example of samples of Candida albicans (URM8368) used to create the database. All were cultivated in Petri dishes using Sabouraud Dextrose agar culture medium. (c):E-Nose collection cycle. (1) Camera suction step (2) Sensor stabilization step (3) Camera cleaning (purge). (d): Data from the readings of each sensor over time for the samples of C. albicans. (e): Data readings from C. krusei after one day of culture. (f): Data readings from C. krusei after two days of culture.

Figure 4

Two dimensions from Principal Component Analysis (a) and Uniform Manifold Approximation and Projection for Dimension Reduction (b).