A NANOSENSOR BASED PORTABLE ARTIFICIAL NOSE TO DETECT LUNG CANCER S. Udina*1, A. Gaara2,E. Vrouwe3, C. Jaeschke1, J. Mitrovics1, H. Haick2 1
2
JLM Innovation GmbH, Vor dem Kreuzberg 17, 72070 Tübingen, Germany Russell Barrie Nanotechnology Institute, Technion – Israel Institute of Technology, Haifa 32000, Israel 3 Micronit Microfluidics, Colosseum 15, 7521 PV Enschede, The Netherlands *
[email protected]
ABSTRACT Breath analysis has a great potential for rapid non-invasive cancer screening and other disease monitoring applications. The design decisions and first experimental results for a portable point-of-care breath analysis instrument for lung cancer detection are presented in this work. Index terms– Breath analysis, breath sampling 1.
Design convenient user interfaces which are suitable for medical environment (clinical trials), but also for the research environment (exploratory data analysis). Minimization of overall size
INTRODUCTION
Breath analysis has a great potential for rapid non-invasive cancer screening and other disease monitoring applications. Over the last years, breath analysis has emerged to a very fast developing and growing field, but it has often relied in remote breath sampling and offline GC-MS (also GC-IMS, PTR-MS) analysis. This approach can introduce unwanted effects [1], and is logistically inefficient requiring the managing, storage and transport of patient samples. Direct breath sampling has clear advantages and ultimately the point-of-care breath sampling instrument is potentially the most efficient approach, as long as the analysis instrument acquisition and maintenance costs can be kept low enough. The availability of such instruments is nowadays extremely limited at the research level, and practically nonexistent at a commercial level. In this work we want to show design decisions, and measurement results with special detail in the breath sampling subsystem, for a point-of-care breath analysis instrument within the framework of project LCAOS, in hope of providing some useful tips and guidelines, which should help standardize a field suffering greatly from lack of standardized sampling methods [2]. 1.1 Design challenges The task of building a breath analyzing system faces many design challenges. Specifically: Prevent hygienic issues like contamination of the breath samples by condensation or particles in breath. Ensure high speed of the fluidic circuit to allow fast sampling and clean-up times, needed for point-of-care applications. Detection of extremely small features in relatively large signal responses of markers in breath (high accuracy of the data acquisition electronics). Baseline stabilization by ensuring that the sensors work under stable environmental conditions, namely at a controlled, stable temperature. - Avoid contamination of breath by VOCs released by the fluidics [1] - Sampling of end-tidal (aka alveolar) air. Avoidance of unsafe contact with the instrument
Figure 1. Front image of the designed instrument. The breath sampling port enables direct sampling the patients’ breath. Outer box is kept 1-2 °C above ambient temperature, while inner electronics and fluidics are kept at T>37°C to avoid condensation. The instrument is connected via USB to a PC interface which presents the data. The design described next intends to provide an answer to all these challenges. 2.
INSTRUMENT DESCRIPTION
Fig. 1 presents an image of the designed instrument. The device is regulated in temperature by an integrated temperature control system. The small size ensures portability and ease-ofuse for point of care applications.
2.1 Breath sampling strategy Capture of end tidal breath has been identified as a key factor in the performance of breath analysis tests [3]. The presence of the flow sensor ensures that sample capture takes place at the end of the exhalation process. Moreover the ≈60 ml total volume glass tube provides some seconds of end-tidal breath buffering which can actually be sampled repeatedly while still keeping significant sample quality. A micropump was used to transfer the sample into the extremely low volume volume nanosensor microchambers, leading to extremely fast subsecond fluidic transport times. The use of single-use exchangeable mouthpieces at the breathing port prevents any hygienic issues said mouthpiece is equipped with a non-return valve avoiding
backflow of sample at the end of the exhalation process. The extremely short fluidic paths and the use of highly inert materials (glass tube, Teflon tubing, stainless steel connectors) ensure minimal contamination of the sample. The microfluidic pump is placed behind the chambers to minimize any contamination coming from it..
Figure 2. a) research software interface including full instrument configuration. b) clinical software interface oriented to clinical users, with improved patient data input 2.2 Software interfaces For such a new instrument coming into the rather new field of breath analysis, the software interfacing has a number of challenges. In particular the needs for exploratory research test measurements and the needs of clinical software interfaces are often contradictory, in the sense that the research interface aims for maximum information and simplicity, and the clinical interface aims for simplicity of use, and low-training operation. For this reason two different software interfaces have been designed, as shown in Fig. 2. The research interfaces includes complete instrument configuration in a Labview ® Interface. The clinical interface is a simplified interface with improved patient information registration. 3.
RESULTS AND DISCUSSION
Figure 3 shows an example response from a selected resistive gold nanoparticle (GNP) sensor to sampled breath. The response shows many different effects happening at the interfaces. The fast response of the nanosensor and the fluidic transport system allows for an extremely good interpretation and identification of effects, which would otherwise be impossible with other reference technologies, like GC, due to sampling frequency limitations. The presented measurements shows sampling transients related to different effects: tiny fluidic leaks, baseline drifts and sample evolution over short periods of time. The system provides valuable data and reference tools to standardize the breath sampling procedure in compact instruments. Still, there is much interpretation and measurement
work to perform in order to establish the optimal measurement parameters for the various applications, and in particular for lung cancer detection.
Figure 3. Detailed example of operation of the breath sampling system. A resistive GNP sensor is used to identify the phases of the sampling process. Different numbered areas correspond to: ① Pump is off, baseline ambient air inside the sensor chamber while patient is blowing into the breath sampling port. ② Pump is turned on after patient exhalation is finished, end-tidal breath enters the sensor chamber immediately. ③ Pump is stopped and the breath sample is measured by the sensors, after a brief transient, a stationary signal is reached. ④ pump is turned on again showing three different interesting effects, a first down-peak shows tiny leakage of ambient air into the fluidics tubing, then a small buffer volume provided by the fluidics tubing presents again the initial breath sample to the sensors. ⑤ with the pump still on, a strong overshoot indicates that the new sample coming from the buffer area of the breath sampling tube has reached the sensors, change in signal probably indicates a change in VOC composition due to time evolution of the VOCs at the glass tube buffer area, or perhaps a sample temperature effect.⑥After pump is stopped, the sensor reaches almost instant stationary signal with the resampled breath. ⑦cleanout cycle after removal of the mouthpiece shows initial very fast recovery of baseline, followed by slow baseline drift which can be attributed to release of less volatile VOCs, or simply a temperature effect. 4.
ACKNOWLEDGEMENT
The research leading to these results has received funding from the FP7-Health Program under the LCAOS (grant agreement no. 258868). Dr Jens Herbig is acknowledged for his valuable technical input. 5.
REFERENCES
[1] Bos LDJ, Wang Y, Weda H, Nijsen TME, Janssen APGE, Knobel HH, Vink TJ, Schultz MJ, Sterk PJ “A simple breath sampling method in intubated and mechanically ventilated critically ill patients” Respir Physiol Neurobiol, vol. 191, pp.67-74, 2014 [2] Kim K.-H., Jahan S. A., Kabir E., “A review of breath analysis for diagnosis of human health” TrAC, vol. 33, pp 1-8, 2012 [3] Amann, A., Miekisch, W., Pleil, J., Risby, T., Schubert, J.,0. “Methodological Issues of Sample Collection and Analysis of Exhaled Breath, Exhaled Biomarkers” European Respiratory Society Journals Ltd., pp. 96–114, 2010
