Extracting Patterns of Chemical Information from Differential Mobility Spectrometry Measurements under Varying Conditions of Humidity and Temperature
Philipp Müller 1∗ , Gary A. Eiceman 2 , Anton Rauhameri 1 , Anton Kontunen 1,3 , Antti Roine 1,3 , Niku Oksala 1,3,4 , Antti Vehkaoja 1 , and Maiju Lepomäki 1,5 · 2026
Read Extracting Patterns of Chemical Information from Differential Mobility Spectrometry Measurements under Varying Conditions of Humidity and Temperature by Philipp Müller 1∗ , Gary A. Eiceman 2 , Anton Rauhameri 1 , Anton Kontunen 1,3 , Antti Roine 1,3 , Niku Oksala 1,3,4 , Antti Vehkaoja 1 , and Maiju Lepomäki 1,5 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-19572.txt.
Introduction
Differential Mobility Spectrometry is a fast, cheap tool for pulling facts from gas samples with tricky volatile organic compounds, including surgical smoke. Philipp Müller and seven co-authors note one snag: readings hinge on humidity and, less so, heat, making matches across settings hit-or-miss. Pinning settings to fixed levels is often out of reach.
The paper studies 1,852 readings of surgical smoke from heated porcine adipose and muscle tissue, mapping how shifting dampness and heat affect results. Grasping that impact matters for matching data gathered under varied settings. The work confirms the link to surroundings and puts numbers on it for such samples.
Catalog record
| CREATOR | Philipp Müller 1∗ , Gary A. Eiceman 2 , Anton Rauhameri 1 , Anton Kontunen 1,3 , Antti Roine 1,3 , Niku Oksala 1,3,4 , Antti Vehkaoja 1 , and Maiju Lepomäki 1,5 |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-19572
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Extracting Patterns of Chemical Information from Differential Mobility Spectrometry Measurements under Varying Conditions of Humidity and Temperature Philipp Müller 1∗ , Gary A. Eiceman 2 , Anton Rauhameri 1 , Anton Kontunen 1,3 , Antti Roine 1,3 , Niku Oksala 1,3,4 , Antti Vehkaoja 1 , and Maiju Lepomäki 1,5 Abstract . Differential Mobility Spectrometry (DMS), also known as Field Asymmetric Ion Mobility Spectrometry, is a rapid and affordable technology for extracting information from gas phase samples containing complex volatile organic compounds, and can therefore be used for analyzing surgical smoke. One obstacle to its widespread application is the dependence of DMS measurements on humidity and, to a lesser degree, temperature, making comparison of data measured under different environmental conditions arbitrary. The commonly used solution is to regulate these environmental conditions to some predefined humidity and temperature levels. However, this approach is often unfeasible or even impossible. Therefore, in this paper we analyzed a dataset of 1 852 DMS measurements of surgical smoke evaporated from porcine adipose and muscle tissue to get an understanding of the impact of varying humidity and temperature on DMS measurements. Our analysis confirmed clear dependence of the measurements on these two factors. To overcome this challenge, we fitted regression models to raw and normalized DMS measurement data. Subsequently, these models were used for estimating DMS measurements for known tissue types based on recorded humidity and temperatures. Our test suggests that it is possible to estimate DMS measurements of surgical smoke from porcine adipose and muscle tissue under specific environmental conditions by standardizing DMS measurements separation voltage-wise and training multivariate regression models on the normalized data, which is the first step in removing the need for standardized measurement conditions. † † footnotetext: ∗ corresponding author: [email protected] 1 Sensor Technology and Biomeasurements, Faculty of Medicine and Health Technology, Tampere University, FI-33720 Tampere, Finland 2 Department of Chemistry and Biochemistry, New Mexico State University, Las Cruces, NM 88003-8001, USA 3 Olfactomics Ltd, Kampusareena, Korkeakoulunkatu 7, FI-33720 Tampere, Finland 4 Vascular Centre, Tampere University Hospital, Central Hospital, P.O. Box 2000, FI-33521 Tampere, Finland 5 Department of Pathology, Fimlab Laboratories, Arvo Ylpön katu 4, FI-33520 Tampere, Finland Version: 23 February 2026 This work has been submitted to Springer for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible. 1. Introduction Ion mobility spectrometry (IMS) is an emerging and promising technique for molecular analysis of complex mixtures of chemicals, which ionizes gas-phase molecules and measures drift times of the resulting ions in a carrier gas under the influence of electric fields. Due to differences in the ions’ molecular weights, charge, and the shape among compounds, their mobilities differ [ 1 ] , resulting in different measurements for different volatile organic compounds (VOCs) [ 2 ] . Therefore, IMS can, unlike mass spectrometers, distinguish molecules with the same mass but different shape. Due to numerous choices of carrier gases and designs for the electric fields, multiple IMS configurations exist [ 3 ] . Differential mobility spectrometry (DMS), also known as Field Asymmetric Ion Mobility Spectrometry, operate at atmospheric pressures and yield high degree of selectivity [ 3 ] . In DMS an electric field consisting of an asymmetric oscillating high intensity field and a low static field component is used to obtain information on the electric field-mobility dependence of ions [ 4 ] . DMS sensors have gained popularity over recent years. DMS measurements of evaporated VOCs as input data for various machine learning techniques has been shown to enable accurate identification of, for example, chemicals [ 5 ] , diseases in cultivars [ 6 ] , and differentiation of porcine tissue types [ 7 , 8 , 9 ] as well as distinguishing between cancerous and healthy tissue based on surgical smoke generated from cutting the tissue with a surgical knife or evaporating it with a laser [ 10 , 11 ] . However, a challenge to the accurate identification based on DMS measurements is that all classifiers require stable environmental conditions. Changes in humidity and temperature impact the mobility of ions [ 12 , p. 250 ff.] , and hence result in different sensor responses of the same VOCs under different environmental conditions. If stable humidity, temperature, and air pressure cannot be ensured, the performance of any classifier relying on DMS measurements will suffer considerably. A descriptive example of this risk was observed in [ 13 ] . In this study, human tissue samples were identified based on DMS measurements of intraoperatively gathered surgical smoke samples in 20 breast cancer surgeries. The relative humidity varied between 12% and 37% over the 20 surgeries, and the linear discriminant analysis classifier, which yielded accuracies between 93 and 95% in [ 7 ] , only achieved a mean accuracy of 44%. Other studies also revealed variations in environmental conditions inside operating rooms. For example, in [ 14 ] and [ 15 ] the authors observed temperatures between 19°C and 23°C and relative humidities between 55% and 65% over 23 breast surgeries and 22 head and neck surgeries respectively. Studies in [ 13 ] – [ 15 ] highlighted that regulating humidity and temperature to some predefined standard conditions, in operating rooms, is often unfeasible or even impossible. In this paper, to our knowledge for the first time, the aim was to standardize the measurements rather than the measurement conditions. Therefore, we first investigated the effect of changing humidity and temperature on DMS measurements of surgical smoke from porcine adipose and skeletal muscle tissue. We then tested two regression methods for estimating the DMS measurements of both tissue types under given environmental conditions and compared the estimates with actual DMS measurements at these conditions. Previous work on predicting DMS measurements include, e.g. [ 16 , 17 ] . In [ 16 ] the authors predicted DMS curves for MeOH and MeCN diluted in water using a random forest regression model and nitrogen as carrier gas. They did not investigate the impact of environmental conditions. In contrast, we used purified, compressed air as carrier gas and predicted DMS measurements of surgical smoke, which is more complex than the diluted chemicals used in [ 16 ] . In [ 17 ] authors described how to predict VOCs DMS measurement by modeling dynamic ion-solvent cluster formation within the DMS sensor. Again, nitrogen was employed as carrier gas and the impact of changing humidity and temperature on DMS measurements was not studied. The complexity of the ionization chemistry of mixtures presented a major challenge in our work, which was further highlighted by vapor modifier effects (particularly water). This might explain why the earlier works mentioned in the previous paragraph focused on pure chemicals rather than chemical mixtures. The complexity of surgical smoke was shown, for example, in Weber and Spleiß [ 19 ] who detected main volatile organic substances from eleven chemical groups in surgical smoke from porcine muscle tissue. Albrecht et al. [ 20 ] detected approximately 150 VOCs and identified 15 substances in emission from laser surgery and electrosurgery for liver and fat tissue; Spleiß et al. [ 21 ] identified 59 substances in emissions form surgical CO 2 laser of liver and muscle tissue; Barrett and Garber