Handheld Erythema and Bruise Detector

Visual inspection of intact skin is commonly used when assessing persons for pressure ulcers and bruises. Melanin masks skin discoloration hindering visual inspection in people with darkly pigmented skin. The objective of the project is to develop a point of care technology capable of detecting erythema and bruises in persons with darkly pigmented skin. Two significant hardware components, a color filter array and illumination system have been developed and tested. The color filter array targets four defined wavelengths and has been designed to fit onto a CMOS sensor. The crafting process generates a multilayer film on a glass substrate using vacuum ion beam splitter and lithographic techniques. The illumination system is based upon LEDs and targets these same pre-defined wavelengths. Together, these components are being used to create a small, handheld multispectral imaging device. Compared to other multi spectral technologies (multi prisms, optical-acoustic crystal and others), the design provides simple, low cost instrumentation that has many potential multi spectral imaging applications which require a handheld detector. INTRODUCTION — PURPOSE Visual inspection of intact skin is a common means to identify and characterize bruising and erythema. In people with darkly pigmented skin, the presence of melanin greatly hinders visualization of discoloration. A need exists for clinically usable technology that assists clinicians in assessing dark skin. Previous Bruise Detection Methods During investigations of potential child and elder abuse, clinicians and forensic practitioners assess bruises to document shape, extent and estimate age. Visual inspection in vivo and photography of bruises are two conventional methods employed in clinical practice. However they are qualitative, subjective, inaccurate and hence unreliable. Statistics indicate that their best accuracy is only about 50% [1-4]. Further, their failure rate is even much higher if bruises happened on darkly pigmented skin. Recently spectroscopy has emerged to improve the reliability of bruise detection [5, 6]. It infers the fractional contents of various chromophores (oxy-hemoglobin (oxyHb), deoxy-hemoglobin (deoxyHb), bilirubin, melanin) by measuring the optical reflectivity of one sample point at a time of bruised region for a continuous range of wavelength at fine steps (e.g., at 2nm incremental step). It is a reliable tool to investigate the basic biochemical processes associated with bruised skin on both white and darkly pigmented skin. However, point spectroscopy can be too time consuming and tedious for clinical use since one must image multiple locations. Moreover, point spectroscopy is not conducive to create a distribution map of the chromophore concentration over time. Such a map is important since it contains the intrinsic features of the diseased skin such as its shape, size and age [7]. Previous Erythema Identification Methods Over the past decade, researchers have identified erythema identification using non invasive methods [8-10]. Additional studies [11, 12] analyzed erythema based on the changes in skin color using true color images. However, they did not report cases on darkly pigmented skin. Tissue Reflectance Spectroscopy (TRS) was clinically used to detect Stage I ulcers in subjects with different types of skin [13, 14]. They reported high sensitivity (85%) and specificity (75%). However, similar to the case of bruise detection, TRS is performed at one spatial point at a time which makes the practical clinical diagnostic procedure tedious, time consuming and laborious. Medical Imaging 2008: Computer-Aided Diagnosis, edited by Maryellen L. Giger, Nico Karssemeijer Proc. of SPIE Vol. 6915, 69153K, (2008) · 1605-7422/08/$18 · doi: 10.1117/12.770718 Proc. of SPIE Vol. 6915 69153K-1 Multispectral Imaging Method and Applications Multi-spectral imaging has matured into a technology with many applications, including classification of difficult targets in defense [15], produce sorting [16], precision farming in agriculture [17], product quality online inspection in manufacturing [18], contamination detection in food industry [19], remote sensing in mining [20], atmospheric composition monitoring in environmental engineering, and early stage diagnosis of cancer and tumors [21-24]. With great progressive pace, multi-spectral imaging is replacing many visual or visual experience based diagnostic methods as the most important disease diagnostic tool and post treatment diagnostic tool (in medical imaging). Both TRS and multispectral imaging methods measure light reflectivity of diseased skin. The difference is that the former samples more densely in the spectral domain and more sparsely over the spatial domain than the latter. Images of millions of samples (e.g. 600x800 pixels for a typical low-cost CCD camera) can be conveniently obtained at a set of discrete wavelengths using band pass filters to remove unimportant spectra. This makes a multi-spectral imaging system a cost effective and efficient tool for detecting and characterizing bruises and erythema. Results of initial efforts of using multi-spectral imaging to study erythema and cutaneous inflammation in vivo [25, 26] indicate that it allows for the construction of quantitative maps relating to erythema and edema. The most extensive study of bruise age using indexes of bilirubin and hemoglobin over time were reported in Randeberg and colleagues [27]. They used multi-spectral imaging to detect and characterize bruises on light pigmented skin at spectral ranges of 400 – 1000nm (VNIR) and 900 – 1700nm (SWIR). However they did not report using this approach on darkly pigmented skin. Current multispectral imaging applications require either multiple exposures or extensive post-processing. Examples of these current technologies include filter wheels, generalized Lyot filter, electrical tunable filter, optic-acoustic based SmartSpectra, NASA’s Multi Spectral Imager. These devices are not well suited for point of care applications with respect to durability, portability and cost. This paper describes the design and development of the key components required of a handheld multispectral imager. A mosaic filter has been developed containing four filters synchronized to chromophores of interest in bruise and erythema detection. As a companion to this filter, an illumination system has been developed that bathes the area of interest with light consisting of the same wavelengths targeted by the mosaic filter. PRELIMINARY INVESTIGATION The initial aspect of this research characterized bruising and erythema using a multispectral imaging system based upon a filter wheel. This work identified a small set of filters that provide needed spectral information and fusion algorithms to achieve contrast enhancement [5, 6]. Using these findings, we focused attention on hardware compatible with the goal of developing a handheld point-of-care device capable of detecting erythema and bruising in people with varying skin tones.