Alexandros Kyriakides

A Novel Method for Bacterial UTI Diagnosis Using Raman Spectroscopy

The current state of the art on bacterial classification using Raman and Surface Enhanced Raman Spectroscopy (SERS) for the purpose of developing a rapid and more accurate method for urinary tract infection (UTI) diagnosis is presented. SERS, an enhanced version of Raman offering much increased sensitivity, provides complex biochemical information which, in conjunction with advanced analysis and classification techniques, can become a valuable diagnostic tool. The variety of metal substrates used for SERS, including silver and gold colloids, as well as nanostructured metal surfaces, is reviewed. The challenges in preprocessing noisy and complicated spectra and the various methods used for feature creation as well as a novel method using spectral band ratios are described. The various unsupervised and supervised classification methods commonly used for SERS spectra of bacteria are evaluated. Current research on transforming SERS into a valuable clinical tool for the diagnosis of UTIs is presented. Specifically, the classification of bacterial spectra (a) as positive or negative for an infection, (b) as belonging to a particular species of bacteria, and (c) as sensitive or resistant to an antibiotic are described. This work can lead to the development of novel technology with extremely important benefits for public health.

Classification of bacterial samples as negative or positive for a UTI and antibiogram using surface enhanced Raman spectroscopy

Urinary tract infection (UTI) diagnosis requires an overnight culture to identify a sample as positive or negative for a UTI. Additional cultures are required to identify the pathogen responsible for the infection and to test its sensitivity to antibiotics. A rise in ineffective treatments, chronic infections, rising health care costs and antibiotic resistance are some of the consequences of this prolonged waiting period of UTI diagnosis. In this work, Surface Enhanced Raman Spectroscopy (SERS) is used for classifying bacterial samples as positive or negative for UTI. SERS spectra of serial dilutions of E.coli bacteria, isolated from a urine culture, were classified as positive (105-108 cells/ml) or negative (103-104 cells/ml) for UTI after mixing samples with gold nanoparticles. A leave-one-out cross validation was performed using the first two principal components resulting in the correct classification of 82% of all samples. Sensitivity of classification was 88% and specificity was 67%. Antibiotic sensitivity testing was also done using SERS spectra of various species of gram negative bacteria collected 4 hours after exposure to antibiotics. Spectral analysis revealed clear separation between the spectra of samples exposed to ciprofloxacin (sensitive) and amoxicillin (resistant). This study can become the basis for identifying urine samples as positive or negative for a UTI and determining their antibiogram without requiring an overnight culture.

Complete urinary tract infection (UTI) diagnosis and antibiogram using surface enhanced Raman spectroscopy (SERS)

There are three stages to a complete UTI diagnosis: (1) identification of a urine sample as positive/negative for an infection, (2) identification of the responsible bacterium, (3) antibiogram to determine the antibiotic to which the bacteria are most sensitive to. Using conventional methods, all three stages require bacterial cultures in order to provide results. This long delay in diagnosis causes a rise in ineffective treatments, chronic infections, health care costs and antibiotic resistance. In this work, SERS is used to identify a sample as positive/negative for a UTI as well as to obtain an antibiogram against different antibiotics. SERS spectra of serial dilutions of E. coli bacteria mixed with silver nanoparticles, showed a linear correlation between spectral intensity and concentration. For antibiotic sensitivity testing, SERS spectra of three species of gram negative bacteria were collected four hours after exposure to the antibiotics ciprofloxacin and amoxicillin. Spectral analysis revealed clear separation between bacterial samples exposed to antibiotics to which they were sensitive and samples exposed to antibiotics to which they were resistant. With the enhancement provided by SERS, the technique can be applied directly to urine samples leading to the development of a new, rapid method for UTI diagnosis and antibiogram.

Classification of Raman Spectra using Support Vector Machines

The classification of Raman Spectra is useful in identification and diagnosis applications. We have obtained Raman Spectra from bacterial samples using three different species of bacteria. Before any form of classification can be carried out on the Raman Spectra it is important that some form of normalization is used. This is due to the nature of the readings obtained by the acquisition equipment. The method used for normalization greatly affects the accuracy of the results. We have carried out experiments using Support Vector Machines and the correlation kernel. Our observations have led us to the hypothesis that the correlation kernel is “self-normalizing” and gives satisfactory results without the need of any other normalization technique.

Urinary tract infection diagnosis and response to antibiotics using Raman spectroscopy

Urinary tract infection diagnosis and antibiogram require a 48 hour waiting period using conventional methods. This results in ineffective treatments, increased costs and most importantly in increased resistance to antibiotics. In this work, a novel method for classifying bacteria and determining their sensitivity to an antibiotic using Raman spectroscopy is described. Raman spectra of three species of gram negative Enterobacteria, most commonly responsible for urinary tract infections, were collected. The study included 25 samples each of E.coli, Klebsiella p. and Proteus spp. A novel algorithm based on spectral ratios followed by discriminant analysis resulted in classification with over 94% accuracy. Sensitivity and specificity for the three types of bacteria ranged from 88-100%. For the development of an antibiogram, bacterial samples were treated with the antibiotic ciprofloxacin to which they were all sensitive. Sensitivity to the antibiotic was evident after analysis of the Raman signatures of bacteria treated or not treated with this antibiotic as early as two hours after exposure. This technique can lead to the development of new technology for urinary tract infection diagnosis and antibiogram with same day results, bypassing urine cultures and avoiding all undesirable consequences of current practice.

ECG analysis in the Time-Frequency domain

The Electrocardiogram (ECG) has been established as a powerful diagnostic tool in medicine which provides important information about the patients heart condition. The correct identification of the QRS complexes is a fundamental step in every automated or semi-automated ECG analysis method. A major problem that is often encountered in automatic QRS detection is the presence of artifacts in the ECG data, which cause considerable alterations to the signal. In this work, the objective was to develop a method, based on Time-Frequency Analysis (TFA), which would be able to automatically detect and remove artifacts in order to increase the reliability of automatic QRS detection. The TFA method used for the analysis of the ECG data, was based on a time-varying Autoregressive (AR) model whose solutions were obtained using Burgs method. The algorithm could detect and remove 95.6% of the artifact areas and correctly identify 92.0% of QRS complexes (322 out of 335 annotated QRS complexes). The proposed method was compared with one of the most commonly used methods in ECG analysis, which is based on the use of wavelets. The wavelet-based method resulted in an accuracy of QRS detection of 65.3% mainly due to the large number of false positive detections in the regions of artifact.

UTI diagnosis and antibiogram using Raman spectroscopy

Urinary tract infection diagnosis and antibiogram require a 48 hour waiting period using conventional methods. This results in ineffective treatments, increased costs and most importantly in increased resistance to antibiotics. In this work, a novel method for classifying bacteria and determining their sensitivity to an antibiotic using Raman spectroscopy is described. Raman spectra of three species of gram negative Enterobacteria, most commonly responsible for urinary tract infections, were collected. The study included 25 samples each of E.coli, Klebsiella p. and Proteus spp. A novel algorithm based on spectral ratios followed by discriminant analysis resulted in classification with over 94% accuracy. Sensitivity and specificity for the three types of bacteria ranged from 88-100%. For the development of an antibiogram, bacterial samples were treated with the antibiotic ciprofloxacin to which they were all sensitive. Sensitivity to the antibiotic was evident after analysis of the Raman signatures of bacteria treated or not treated with this antibiotic as early as two hours after exposure. This technique can lead to the development of new technology for urinary tract infection diagnosis and antibiogram with same day results, bypassing urine cultures and avoiding all undesirable consequences of current practice.

Classification of Raman spectra of bacteria using rank order kernels

The range of applications of Raman-based classification has expanded significantly, including applications in bacterial identification. In this paper, we propose the use of Rank Order Kernels to classify bacterial samples. Rank Order Kernels are two-dimensional image functions which operate on two-dimensional images. The first step in the classification therefore, is to transform the Raman spectra to two-dimensional images. This is achieved by splitting the spectra into segments and calculating the ratio between the mean value of each and every other segment. This creates a two-dimensional matrix of ratios for each Raman spectrum. A similarity metric based on rank order kernels operating on the two-dimensional matrices is then used with a nearest neighbor algorithm for classification. Our results show that this method is comparable in accuracy to other methods which were used previously for the same data set.

Isolated word endpoint detection using time-frequency variance kernels

A major challenge in developing endpoint detection systems is the presence of background noise. We have developed a hybrid method for performing endpoint detection which is based on spectrogram estimation using LPC and a detection process based on imaging operations on the spectrogram. High-variance regions in the spectrogram, captured by variance kernels, can be used to accurately determine the endpoints of speech. This hybrid approach to endpoint detection is robust to various types and levels of background noise. Compared with two other publicly-available methods, our approach performs favorably.

Identification and Antibiotic Sensitivity of UTI Pathogens Using Raman Spectroscopy

Conventional methods of Urinary Tract Infection (UTI) diagnosis require determining the concentration and identity of the involved bacteria, as well as their susceptibility to various antibiotics, the so-called antibiogram. Such assays require repeated culturing of a sample and need at least 48 hours in order for bacterial colonies to be grown, counted, and exposed to antibiotics. However, the patient cannot remain untreated during this rather prolonged period before definitive diagnosis of the suspected infection becomes available. As a result, physicians prescribe broad spectrum antibiotics prior to antibiogram availability. This practice has many undesirable consequences, both short term and long term: (i) unsuccessful treatments leading to chronic infections, (ii) increased health care costs, and, most importantly, (iii) increased antibiotic resistance by a growing number of bacterial strains (Gruneberg, 1994; Casadevall, 1996; Cosgrove, S. & Carmeli, 2003; Alanis, 2005). Given these concerns, it is obvious that rapid and accurate identification of UTI pathogens as well as determination of their susceptibility to antibiotics would offer significant clinical benefits. Such methodologies are currently being developed and include the promising application of Raman spectroscopy for the diagnosis of UTIs. Recently, rapid diagnosis methods based on polymerase chain reaction (PCR) have been developed in order to bypass the need for culturing (Mothershed & Whitney, 2006) as well as to identify genes that confer antibiotic resistance (Rolain et al., 2004). Although such PCR assays are fast and very sensitive, they typically require species and strain specific probes that may or may not be available for a particular organism. In addition, amplification methods, like PCR, suffer from contamination problems, complex interpretation of results, as well as high costs. Mass Spectrometry is another method that has been proposed as an alternative approach for bacterial diagnostics without culturing (Chen et al., 2008). However, like the PCR approach, Mass Spectrometry also depends on prior knowledge of the pathogen under study and suffers from increased complexity and cost. Vibrational spectroscopies, like Raman spectroscopy, have been used, for the last few years, to detect bacteria with minimal sample manipulation (Maquelin et al., 2000; Schuster, 2000a, 2000b). Classification of bacterial species, as well as of subspecies, has been achieved with great accuracy and speed, especially with Surface Enhanced Raman Spectroscopy (SERS) (Kneipp et al., 2006) which allows enhancement of the inherently weak Raman signal. More