- B.S., University of North Carolina at Chapel Hill (1979)
- Ph.D., The Pennsylvania State University (1984)
Clinical and environmental applications of infrared spectroscopy, noninvasive blood glucose sensing, passive infrared remote sensing of environmental pollutants, biomedical applications of infrared imaging, and chemometrics.
Our research program focuses on the development of new analytical techniques in infrared spectroscopy for use in clinical and environmental applications. A major component of our work is the design of methods for both qualitative and quantitative determinations of analytes in complex sampling environments. We are combining state-of-the-art methods in spectroscopy with advanced computer-based data analysis in the solution of these problems.
We are developing environmental monitoring techniques based on the synergistic use of a single-point passive Fourier transform infrared (FT-IR) spectrometer coupled to an infrared line scanner imaging system. These passive remote sensors are configured to collect naturally occurring infrared radiance from the outdoor environment. The spectral features of airborne compounds are superimposed on this ambient infrared background emission as either absorption or emission bands. By use of appropriate data processing methods, these analyte signatures can be extracted and used to identify target chemicals, as well as estimate their concentrations. The combined use of an imaging line scanner system with a single-point spectrometer provides a unique capability for identifying chemical plumes and then interrogating them in detail. Example applications for this technology include regulatory monitoring of industrial stack emissions, chemical leak detection, pollutant monitoring at hazardous waste sites, and emergency response scenarios such as chemical plant accidents or terrorist incidents. Our work is focusing on new design concepts for these sensors, as well as the development of signal processing, pattern recognition, and image analysis methodology for use in data interpretation.
In collaboration with Professor Mark Arnold, we are pursuing the development of infrared-based chemical sensors for use in clinical applications. Our current efforts are directed to the use of near-infrared spectroscopy for monitoring glucose in blood and for measuring protein and urea during hemodialysis treatments. The near-infrared spectral region is particularly suited to these measurements because of the presence of transmission windows in which the strong background absorbance of water is reduced. Within these windows, combination and overtone bands of glucose and other analytes can be observed on top of the spectral background. For the glucose measurement, a noninvasive sensor is being developed that transmits near-infrared light through the dermis, followed by the application of chemometric methods to the resulting spectra to extract glucose information from the complex spectral background. Research is underway to improve and simplify the required spectroscopic instrumentation, as well as to develop the signal processing and calibration protocols necessary for implementing a robust analysis. If successful, the noninvasive glucose sensor will allow the direct measurement of blood glucose levels without requiring the collection of a blood sample. This technology will greatly benefit diabetic patients who must monitor their glucose levels several times per day.
A third area of interest lies in the application of FT-IR microscopic imaging to biomedical applications. This emerging technique couples an FT-IR spectrometer, infrared microscope, and multichannel focal plane array detector for use in acquiring infrared images of biological samples such as cells and tissues. By either transmitting infrared light through the sample or reflecting light from it, a single image acquisition can produce up to 16,384 infrared spectra corresponding to a 128128 grid of discrete locations across the sample. Spatial resolutions can be achieved down to the diffraction limit of less than 10 μm. Applications of interest for this technology include the automated diagnosis of disease state from spectra of tissues collected during biopsies and the correlation of spectral information in diseased tissue with treatment outcomes. A major issue in working with this technique is how to handle the tremendous volume of data acquired. We are pursuing a variety of research strategies in signal processing, data compression, feature extraction, numerical optimization, and pattern recognition in the design of tools for use in processing and extracting information from these infrared images.
- Wan, B.; Small, G. W. Wavelet Analysis Used for Spectral Background Removal in the Determination of Glucose from Near-Infrared Single-Beam Spectra; Analytica Chimica Acta 2010, 681, 63-70.
- Tarumi, T.; Amerov, A. K.; Arnold, M. A.; Small, G. W. Design Considerations for Near-Infrared Filter Photometry: Effects of Noise Sources and Selectivity; Applied Spectroscopy 2009, 63, 700-708.
- Tarumi, T.; Wu, Yuping; Small, G. W. Multivariate Calibration with Basis Functions Derived from Optical Filters; Analytical Chemistry 2009, 81, 2199-2207.
- Sulub, Y.; Small, G. W. Spectral Simulation Protocol for Extending the Lifetime of Near-Infrared Multivariate Calibrations; Analytical Chemistry 2009, 81, 1208-1216.
- Kramer, K. E.; Small, G. W. Digital Filtering and Model Updating Methods for Improving the Robustness of Near-Infrared Multivariate Calibrations; Applied Spectroscopy 2009, 63, 246-255.
- Kramer, K. E.; Small, G. W. Blank Augmentation Protocol for Improving the Robustness of Multivariate Calibrations; Applied Spectroscopy 2007, 61, 497-506.
- Sulub, Y.; Small, G. W. Spectral Simulation Methodology for Calibration Transfer of Near-Infrared Spectra; Applied Spectroscopy 2007, 61, 406-413.
- Kramer, K. E.; Small, G. W. Effects of Spectral Resolution on the Determination of Glucose in a Simulated Biological Matrix by Fourier Transform Near Infrared Spectrometry; Journal of Near Infrared Spectroscopy 2006, 14, 291-299.
- Sulub, Y.; Small, G. W. Simulated Radiance Profiles for Automating the Interpretation of Airborne Passive Multi-Spectral Images; Applied Spectroscopy 2008, 62, 1049-1059.
- Wan, B.; Small, G. W. Airborne Passive Fourier Transform Infrared Remote Sensing of Methanol Vapor from Industrial Emissions; The Analyst 2008, 133, 1776-1784.