PC-Based Chromatic Monitoring of E coli in Urine

A. T. Sufian and C. R. Jones

Introduction

Chromatic monitoring has been used for preliminary testing of infection in urine using a portable system. The system was based upon a combination of a webcam w'ith a personal computer (Deakin et al„ 2014) to provide a convenient and economic means for primary medical care. The approach has been show n to be capable of characterising complex samples of urine and its components.

Chromatic System

The chromatic system consisted of optically addressing a urine sample in a transparent cuvette w'ith polychromatic illumination produced by a tuneable laptop computer visual display unit (VDU) screen. The cuvette was placed between the VDU screen and a webcam which captured an image of the cuvette and computer screen (Figure 4.1a). The image captured by the webcam was transferred directly to the host laptop computer for recording and chromatic analysis. The VDU screen illumination could be tuned and controlled via the host computer. It involved adjusting the gains of the R, G. В screen channels to optimise performance via trial and error, which led to gains of 0.7 for R. G and В for this particular application. The webcam was adjusted so that the full illumination area provided on the VDU screen filled the camera image area. The screen areas on either side of the cuvette were tuned via trial and error for providing a means for any image illumination correction and to stabilise the webcam response.

Tests were performed w'ith polychromatic light from the VDU screen transmitted through the cuvette and sample in the presence of ambient light and also with the VDU illumination replaced by a black card so that only ambient light was present. Figure 4.1b shows examples of images of the cuvette only captured by the webcam. Figure 4.1b (i) TRANS is the image of a normal urine sample with VDU transmission, whilst Figure 4.1b (i) REFL is the same sample w'ith ambient light in the absence of VDU illumination. Figure 4.1b (ii) TRANS and REFL are the corresponding images with a highly contaminated urine sample. R. G. В outputs for the urine sample with and without VDU illumination were extracted from the images for further processing.

Chromatic Maps and Processing

Figure 4.2 show's the optical transmission spectra of uncontaminated urine (Vekatratnam and Lents, 2011) and unspecified E. coli in sterilised de-ionised w'ater (Alupoaei et al., 2004) over the wavelength

Portable computer-based chromatic monitoring system,

FIGURE 4.1 Portable computer-based chromatic monitoring system, (a) Layout of the PC monitoring system with cuvette illumination source, sample and webcam, (b) Examples of images of the monitored urine samples with illumination, (i) Negative UTI sample, (ii) Positive UTI sample. Trans - with transmitted screen Illumination; Refl - ambient light, no screen illumination (Deakin et al.. 2014).

range 200-800 nm. Also shown in Figure 4.2 are the responses R, G, В of human vision receptors on which electronic camera detectors are based, covering only the wavelength range 400-650 nm. There remain differences between the two spectra in the 400-650 nm range.

The chromatic parameters used for monitoring urine samples by Deakin et al. (2014) are R and B, leading to chromatic maps of R: B. Since the system needed to have the capability for use under normal ambient illumination, it was necessary to obtain images first with VDU illumination, followed by images without VDU illumination [e.g., Figure 4.1b (i) and (ii)]. Thus, the sample transmission parameters were

The resultant VDU parameters were further normalised with respect to a water sample to provide a scale range (0-1), that is

Optical spectra of uncontaminated urine and E. coli in water (Deakin et al„ 2014). whilst the ambient light parameters were likewise normalised

FIGURE 4.2 Optical spectra of uncontaminated urine and E. coli in water (Deakin et al„ 2014). whilst the ambient light parameters were likewise normalised

The chromatic maps used for checking the urine samples were R(TRANS)n versus B(TRANS)n and R(REFL)n versus B(REFL)n.

Chromatic Analysis Results

An illustration of test results for ten clinical urine samples is given in Figure 4.3a for VDU illumination (R(TRANS)n : B(TRANS)n) and Figure 4.3b for reflection (R(REFL)n : B(REFL)n).

The VDU illumination map shows that the clinical urine samples lie above the R(TRANS) n = B(TRANS)n locus, whereas the E. coli in the pure urine locus lies below the R(TRANS) n = B(TRANS)n locus. Comparison with knowm bacterial growth urine samples indicates that urine with bacterial growth

The ambient light (scattered light) map shows results for the ten urine samples tested along with the E. coli pure urine locus. These results show that all sampled points lie above the R(REFL) n = B(REFL)n locus. Significant bacterial growth samples lie in the G sector.

Chromatic results for 200 urine samples were reported and compared with urine culture results. The comparison showed that the number of true positive infection results was 73, and the number of true negatives was 66, that is, a total of 139 correct results. The number of false positives was 55 and false negatives 6, that is, only 6 samples from the 200 (3%) were incorrectly indicated as not being infected. Of the 55 false positive samples, 49 had some bacterial growth, albeit below the chosen

R : В chromatic maps for ten clinical urine samples, (a) R(TRANS)

FIGURE 4.3 R : В chromatic maps for ten clinical urine samples, (a) R(TRANS): B(TRANS) map for VDU illumination, (b) R(REFL): B(REFL) map for ambient light (Deakin et al., 2014).

urinary tract infection (UTI) cutoff of 105 cfu/mL. Thus, the approach provided a good fail-safe capability for preliminary assessment of E. coli levels.

References

Alupoaei, С. E., Olivares, J. A., and Garcia-Rubio, L. H. (2004) Quantitative spectroscopy analysis of prokaryotic cells, vegetative cells and spores, biosens. Bioelectron. 19, 893-903.

Deakin, A. G., Jones, G. R., Spencer, J. W., Bongard, E. J., HGal, M., Sufian, A. T., and Butler, С. C. (2014) A portable system for identifying urinary tract infection in primary care using a PC-based chromatic technique. Physiol. Meas. 35, 793-805.

Vekatratnam, A. and Lents, N. H. (2011) Zinc reduces the detection of cocaine, methamphetamine, and THC by ELISA. Urine Testing J. Anal. Toxicol. 35, 333-340.

Chromatic Monitoring of Honey Samples

A. T. Sufian and G. R. Jones

Introduction

Floney is a natural sweet substance produced by honeybees worldwide (Figure 5.1). It consists of a complex mixture of water, various sugars, amino acids, enzymes, vitamins and minerals (White, 1975). The composition is influenced not only by natural factors (geographical origins, botanical sources, environment, climate) (Anklam, 1998) but also by processing, handling and storage (Martin and Bogdanov, 2002). Monitoring such a complex liquid is important to the food industry (Bogdanov et al., 1999; Martin and Bogdanov, 2002; Pilizota and Tiban, 2009) to ensure genuine quality and to identify fraudulent imitations and adulteration which affect the integrity of the honey market (Pilizota and Tiban, 2009; El-bialee et al., 2013; Roshan et al., 2013).

Low-cost, traditional methods (taste, smell, visual observation) are not sufficiently reliable for identifying fraudulent and adulterated honeys, whilst sophisticated optical measurement techniques (Gonzales et al., 1999; Isengrad et al., 2001; Terrab et al., 2002; Nanda et al., 2003; El-bialee et al., 2013; Ozbalci et al., 2013; Roshan et al., 2013) require expensive instrumentation (e.g., spectroscopic, colorimetric, polarimetry, refractometry) and are not easily adapted for field use. Methods based upon physicochemical procedures are time consuming and require a multiplicity of tests with expensive equipment and laboratory infrastructure (Anklam, 1998; Bogdanov et al., 1999; Martin and Bogdanov, 2002).

However, optical chromaticity provides a cost-effective means for the preliminary optical monitoring of honey samples with a portable computer (PC) being deployable at remote honey- producing areas as well as consumer sources (Sufian, 2014). Chromatic methods have been deployed for the combined analysis of optical transmission, polarisation and fluorescence signals from various honeys and for producing chromatic maps from which the honey condition may be checked. Such a PC-based system is described, along with chromatic processing to provide a preliminary indication of honey quality.

Map of distribution of world production of honey as of FAO Statistics 2018

FIGURE 5.1 Map of distribution of world production of honey as of FAO Statistics 2018.

Portable Computer-Based Chromatic Monitoring System

The chromatic monitoring system for monitoring honey samples is a more sophisticated form of the PC-based urine monitoring system (Chapter 4). It involves not only chromatic optical transmission but also chromatic fluorescence and polarisation. Figure 5.2a shows a photograph of such a system, whilst Figure 5.2b show's a schematic diagram of the system layout. A honey sample is contained in a transparent cuvette located between a visual display unit (VDU) screen of a portable computer (which is one illumination source) and a webcam (which detects the honey image). The VDU screen, webcam and computer were interconnected to form the basis of the compact, self-contained, portable and cost-effective unit.

Optical transmission tests w'ere performed with polychromatic light produced by the VDU screen and controlled via the computer software (Chapter 4). An image of the cuvette and surrounding screen area w'as captured by the webcam (Figure 5.3a) and transferred to the computer for chromatic processing.

Optical polarisation tests were performed with a polarisation filter positioned between the VDU screen and the honey-containing cuvette, whilst a second polarising filter w'ith its polarising axis suitably rotated was placed between the cuvette and webcam (Figure 5.2).

Optical fluorescence was observed with two light-emitting diodes (LEDs) to provide short- wavelength monochromatic light (405 nm wavelength) for addressing the honey-containing cuvette (Figure 5.2). The two LEDs were inclined at 45 degrees to the camera line of sight on the camera side of the cuvette so that back-scattered fluorescent light could be captured by the webcam together with the VDU illumination just at either side of the sample, while VDU light transmitted through the sample was blocked using a black card.

 
Source
< Prev   CONTENTS   Source   Next >