Detecting Rhinosinusitis With an Electronic Nose Based on Differential Mobility Spectrometry

Research output: Book/ReportDoctoral thesisCollection of Articles


The sense of smell can potentially be used to diagnose diseases. However, the sense of smell in humans has lower sensitivity and discrimination capability compared to animals such as dogs. Probably the most well-known example of using a dog’s sense of smell is drug detecting dogs at airports. Research studies have demonstrated that dogs can identify samples acquired from patients with cancer or infection.

The molecules detected by sense of smell are volatile organic compounds (VOCs). Nowadays, it is possible to measure these compounds, for example, with an electronic nose (eNose). An eNose is a device that analyses gas-phase mixtures and produces a measurement signature that represents the spectrum of the molecules found in the mixture. Differential mobility spectrometry (DMS) is not a traditional eNose but produces comparable information and, in addition, has a higher sensitivity. DMS has not previously been used in otorhinolaryngologic studies.

Acute rhinosinusitis (ARS) is typically caused by a virus or bacteria. As both produce similar symptoms, differentiating them based on symptoms or clinical examination is a challenging task, which often leads to the overprescription of antibiotics. Chronic rhinosinusitis (CRS) involves, for instance, nasal blockage and discharge lasting at least 12 weeks. As many other rhinologic diseases cause similar symptoms, the definite diagnosis of CRS warrants computed tomography imaging, which is not available in primary care. Therefore, there is a need for a rapid, accurate and non-invasive method to diagnose ARS and CRS.

This dissertation examines the diagnostics of ARS and CRS with DMS and consists of four studies. First, five common rhinosinusitis bacteria in vitro were analysed with DMS. Second, maxillary puncture and aspiration of the contents were performed on patients with ARS. The acquired pus was analysed with DMS, and the results were compared to the traditional bacterial culture of the pus. Third, nasal air from volunteers was aspirated into collection bags using a pump built for this purpose. The results were then compared to room air samples and the feasibility of the method was evaluated. Fourth, patients with CRS without nasal polyps and patients with deviated nasal septum were studied. Aspirated nasal air was collected in the same way as in the third study and analysed with DMS. The ability of DMS to distinguish patients in different groups was then evaluated. The data analysis employed in the studies involved machine learning methods which were used to examine the sensitivity and specificity of DMS to distinguish samples.

The results reveal that DMS can separate common rhinosinusitis bacteria in vitro with very good accuracy. Furthermore, DMS shows very good accuracy to distinguish bacterial positive and bacterial negative samples compared to bacterial cultures. The method used for aspirating the nasal air and subsequent analysis with DMS proved to be a useful method. The nasal air samples were perfectly distinguished from room air samples. In addition, DMS demonstrates good accuracy to discriminate patients with CRS without nasal polyps from patients with deviated nasal septum by analysing nasal air.

The studies in this dissertation were pilot studies and the results are affected by small sample size. However, cross-validation provides confidence of the reliability of the classifier. The studies demonstrate that DMS can analyse different sample types and distinguish groups from each other. The aspiration of nasal air was shown to be practicable and can be used in further studies of rhinologic diseases.
Original languageEnglish
Place of PublicationTampere
ISBN (Electronic)978-952-03-2721-7
Publication statusPublished - 2023
Publication typeG5 Doctoral dissertation (articles)

Publication series

NameTampere University Dissertations - Tampereen yliopiston väitöskirjat
ISSN (Print)2489-9860
ISSN (Electronic)2490-0028


Dive into the research topics of 'Detecting Rhinosinusitis With an Electronic Nose Based on Differential Mobility Spectrometry'. Together they form a unique fingerprint.

Cite this