GC-MS system is versatile in determining a wide range of VOCs [6]. Gas chromatograph has a high separating efficiency [24] and separates the chemical mixture into pulses of pure chemicals based on their volatility, or ease with which they evaporate into a gas. The MS identifies the VOCs and quantifies them based on their structure [25].

In the Electronic nose (E-Nose) as portrayed in Fig. (3), the VOCs adsorb onto a sensor where a change in conductivity, color or oscillation of a crystal is detected by the sensing system of the E-nose. The E-nose responds to only a mixture of compounds in the sample. The pattern recognition system interprets and detects the high levels of VOCs causing Lung cancer [10, 17, 26-28]. The drawbacks include the extensive preparation of the breath samples, lack of quantitative data and calibration [17, 29].

Quartz microbalance [30] contains piezoelectric sensors. These sensors use highly sensitive quartz crystals for the measurement of the mass and viscosity of the VOCs that interact within its thin films. Molecular interaction between target and ligand is detected by the sensor. The resistance at resonance, changes with the viscosity of the VOCs in contact with the crystal surface, which leads to a frequency shift [31-33]. Difficulty lies in the selection of sensitive polymers for their application as thin films over the sensors and also in the control of the repetition of the coating process [17, 34].

The colorimeter [35] distinguishes among the VOCs by its composite response. The Colorimetry sensor array uses a diverse range of chemically responsive dyes.

Adsorption of VOCs causes these dyes to change their colours depending upon their chemical environment which is then scanned and converted to a number [17, 36].

The IMs consists of an ionizing source that breaks the VOCs molecules into ions. The detection process in the IMS, operated in the positive and negative modes, is based on the movement of ions in a drift gas that occurs at the ambient pressure under the influence of stable electric field, in the range from 100 to 350 Vcm^-1, towards a Faraday plate [38], thus generating an electrical signal. The electrical signals from each VOC combine to produce an ion spectrum [17, 38, 39].

Cyranose 320 consists of 32 carbon polymer sensors that respond to exhaled breath, where a change in sensor resistance is caused by the adsorption of the VOC mixtures of the breath onto the sensors. The exhaled breath sample of each individual results in a pattern of sensor responses, which is a characteristic of each individual’s smell print. This helps in distinguishing between the breath of a healthy individual and of a cancer stricken person [39].

The screening techniques described above aren’t cost-effective and are slow. They require the use of complex instruments for the dehumidification of the breath biomarkers. Many of these techniques are difficult to use, thus creating a need for highly experienced analysts for the operation and the interpretation of the results [17]. This creates a need for less complicated breath analysers as shown in Fig. (4), that are highly effective for the mass screening of individuals for lung cancer.

Gold nanoparticles are highly sensitive in the detection of biomarkers at lower concentration levels [41]. They are fully biocompatible [42] and no scattering occurs when mono-layered gold nanoparticles are used. All interactions with respect to the gold particles occur at the surface. The electron density change in the surface causes a maximum shift in the surface Plasmon absorption, which helps in the detection of VOCs. It thus enables the use of noble metal nanoparticles as sensitive sensors [43].

Mono-layered capped gold nanoparticles are thermally air stable of reduced dispersity, have well controlled size and are well spaced [1, 44]. As the concentration of gold nanoparticles increases, there is an increase in the dc conductivity. When the concentration reaches to a moderate level, the value of the dc conductivity becomes steady. Toluene solvent suspensions and Gold nanoparticle/chloroform are used in the preparation of Mono-layered gold nanoparticles as there is an increase in electrical conductivity due to the presence of shorter molecular chain at the end of the Au nanoparticles [45]. These monolayered Au particles control the properties at interface and have a high surface area-to-volume ratio, which makes them different and much better than the bigger Au molecules. The limit of identification of VOCs by the nanoelectrode is lower than their detection at the macrosized electrode. This is mainly due to the presence of a higher ratio between the faradic and capacitive currents [46].

The electrodes present in the Chemiresistors, are coated with a conducting polymer film. The thin electrically- conductive polymer films swell due to the chemical reaction of the volatile organic chemicals with that of the films, thus causing a change in electrical resistance across the film. The chemiresistor can be used repeatedly as it regains its original condition by the release of the chemicals in the environment [48].

Metallic nanoparticles are synthesized in a two-phase system which is composed of aqueous and non-aqueous solutions. These solutions are combined with techniques of extraction of ions. The metallic monolayered nanoparticles are then coated with a hydrophobic layer of alkanethiols. First the ions from an aqueous solution are transferred with the help of a phasetransfer reagent to a toluene solution. Then the reduction of gold is done with an aqueous borohydride solution which is later capped with thiols [44]. Lastly the gold nanoparticles are repeatedly extracted and purified from thiols [1].

As shown in Fig. (5), fourteen different chemiresistive MCNP films deposited on micrometric electrical transducers, are combined onto a sensor array. They are insensitive to the water molecules in the exhaled breath and are placed inside a test chamber [49]. Deposition of 20 circular gold electrodes on thermal oxide silicon wafer is done by the use of an electron beam evaporator. Post-deposition, the gold nanoparticles are then extracted from their chloroform solution by sonication [1].

The sensor array is first tested by evacuating the chamber and exposing it to a mixture of artificially designed VOC biomarker breath patterns. The procedure is repeated and the signals obtained, are scrutinized using a standard PCA algorithm [1, 49].

Analyzing the relationship between multiple variables within a multi-dimensional data, requires a statistical method such as PCA, which can reduce the dimension of the data using linear feature extraction technique. Each sensor data is retained and arranged into a matrix format where the sensor types are arranged in rows and the measurement numbers are arranged in columns. The percentage-resistance changes and slopes are used in the calculation of the covariance matrix and its values of the Eigenvector and Eigen value. The two principle components discriminate between non-overlapping healthy and lung cancer group. This is used in the interpretation of PCA results, in the score plot containing the projection of two cluster data onto the two dimensional planes [40]. The data is then confirmed by testing the breath patterns using Solidphase Microextraction (SPME) and GC-MS.

The highly sensitive and accurate GC-MS, with 87% reproducibility, are rarely used because they are expensive, difficult to use and require highly experienced analysts to operate them and interpret the results [1, 17].

Electronic nose cannot obtain quantitative data as the sensors are not selective. If more than one type of vapour is present, a quantitative output from one sensor is likely to be erroneous [50] and therefore the device cannot be calibrated.

Quartz microbalance gives an accuracy of 80%, sensitivity of 85% with 100% specificity. Their main drawback lies in the stability and precision due to the requirement for the pre-concentration of the breath samples collected [51, 52].

The sensitivity of the Colorimeter is 73.3% and specificity 72.4%. They are generally less preferred due to their low sensitivity and the inability to work in the presence of humidity and the need for the pre-concentration of the collected breath [17].

The accuracy of Cyranose 360 is 85% but is not used as it’s slow, costly, requires complex instruments and cannot detect VOCs in the presence of humidity [17].

The electronic nose containing chemiresistors coated with gold nanoparticles have an accuracy of around 88%. The long chain thiols used in the synthesis of the monolayered gold particles differ in their polarity and vapor pressure. This provides good stability to the device for a period of three months [54]. Change in the electrical resistance of the film occurs due to the absorption of the breath onto the sensor array. As a result, these chemiresistors can identify the pattern of each breath biomarker and therefore quantitative data can be obtained [51]. Pre-concentration of the breath samples is not required due to the hydrophobic nature of the thiol coating on the Au nanoparticles. It has a greater reproducibility of 90%, which proves that is more accurate, promising and convenient to be used as a method for lung cancer detection at its early stages [1].

The advantages of this technique include a high accuracy in discriminating between malignant and benign disease. The breath analysis test can be repeated in short intervals, without the requirement for dehumidification. Large volume of sample can be collected and the test is easy to perform. It can be tested on smokers as well as non-smokers and has a great potential as a screening test. The limitations of exhaled breath analysis include the lack of recommended guidelines for sampling of exhaled breath. Also a need arises, to bring down the cost of the gold nanoparticles. Collection and analysis of the samples using techniques such as mass spectrometry and infrared spectroscopy also hinder in the faster analysis of the breath biomarkers.