In vivo fluorescence lifetime tomography.

Local molecular and physiological processes can be imaged in vivo through perturbations in the fluorescence lifetime (FLT) of optical imaging agents. In addition to providing functional information, FLT methods can quantify specific molecular events and multiplex diagnostic and prognostic information. We have developed a fluorescence lifetime diffuse optical tomography (DOT) system for in vivo preclinical imaging. Data is captured using a time-resolved intensified charge coupled device (ICCD) system to measure fluorescence excitation and emission in the time domain. Data is then converted to the frequency domain, and we simultaneously reconstruct images of yield and lifetime using an extension to the normalized Born approach. By using differential phase measurements, we demonstrate DOT imaging of short lifetimes (from 350 ps) with high precision (+/-5 ps). Furthermore, this system retains the efficiency, speed, and flexibility of transmission geometry DOT. We demonstrate feasibility of FLT-DOT through a progressive series of experiments. Lifetime range and repeatability are first measured in phantoms. Imaging of subcutaneous implants then verifies the FLT-DOT approach in vivo in the presence of inhomogeneous optical properties. Use in a common research scenario is ultimately demonstrated by imaging accumulation of a targeted near-infrared (NIR) fluorescent-labeled peptide probe (cypate-RGD) in a mouse with a subcutaneous tumor.

Monte Carlo simulation of light-tissue interaction: three-dimensional simulation for trans-illumination-based imaging of skin lesions.

Three-dimensional, voxel-based, and wavelength-dependent skin lesion models are developed and simulated using Monte Carlo techniques. The optical geometry of the Nevoscope with trans-illumination is used in the simulations for characterizing the lesion thickness. Based on the correlation analysis between the lesion thickness and the diffuse reflectance, optical wavelengths are selected for multispectral imaging of skin lesions using the Nevoscope. Tissue optical properties reported by various researchers are compiled together to form a voxel library. Tissue models used in the simulations are developed using the voxel library which offers flexibility in updating the optical properties and adding new media types into the models independent of the Monte Carlo simulation code.

Multi-spectral image analysis and classification of melanoma using fuzzy membership based partitions.

The sensitivity and specificity of melanoma diagnosis can be improved by adding the lesion depth and structure information obtained from the multi-spectral, trans-illumination images to the surface characteristic information obtained from the epi-illumination images. Wavelet transform based bi-modal channel energy features obtained from the images are used in the analysis. Methods using both crisp and fuzzy membership based partitioning of the feature space are evaluated. For this purpose, the ADWAT classification method that uses crisp partitioning is extended to handle multi-spectral image data. Also, multi-dimensional fuzzy membership functions with Gaussian and Bell profiles are proposed for classification. Results show that the fuzzy membership functions with Bell profile are more effective than the extended ADWAT method in discriminating melanoma from dysplastic nevus.

Multi-spectral imaging and analysis for classification of melanoma.

Wavelengths in the visible spectrum are selected for multi-spectral trans-illumination imaging of the skin lesions using the Nevoscope. The multi-spectral image data is analyzed using crisp and fuzzy partitioning techniques for classification of melanoma. It is shown that the multi-spectral images add the lesion depth and structural information to the superficial lesion characteristics obtained from the surface illumination images and hence, improve the sensitivity and specificity of melanoma diagnosis.

Classification of melanoma using tree structured wavelet transforms.

This paper presents a wavelet transform based tree structure model developed and evaluated for the classification of skin lesion images into melanoma and dysplastic nevus. The tree structure model utilizes a semantic representation of the spatial-frequency information contained in the skin lesion images including textural information. Results show that the presented method is effective in discriminating melanoma from dysplastic nevus. The results are also compared with those obtained using another method of developing tree structures utilizing the maximum channel energy criteria with a fixed energy ratio threshold.