An estimated 160,000 people each year develop melanomas, the most dangerous form of skin cancer, and as many as 48,000 die worldwide every year from some form of skin disease according to the recent World Health Organization (WHO) report. As such, there is a critical need to develop improved diagnostic procedures to reduce these deaths. Because such large numbers of cases are only expected to increase over the next few decades, the object of this application is to develop an improved skin cancer diagnostic system that can have a significant and positive impact on this population, which is the subject of this proposal. The central hypothesis of this application is that single-scattered, polarized light spectroscopic methods combined with multiple-scattered, unpolarized light spectroscopy provide unprecedented tissue functional information and cellular structures for rapid noninvasive diagnosis of the skin cancer. This hypothesis has been formulated on the basis of strong preliminary data produced with our combined Multiple Scattered Light Spectroscopy (MSLS)/Polarized Light Spectroscopy (PLS) prototype laboratory system. The rationale for the proposed research is that a combination of MSLS/PLS technology can be used to accurately reflect morphologies in specific diseased layers of skin. MSLS mathematically models the multilayered structure of the skin using light propagation, whereas PLS physically (not mathematically) discriminates between multiple layers of tissue. This proposed system, achieved highly accurate results for skin cancer detection in a pilot clinical study, providing a specificity of 91%, relative to a specificity of 80% and 82% provided by using PLS and MSLS methods individually. Thus the proposed research is fundamental to the essential part of NIH's mission that pertains to the development new systems to reduce the burden of disease. Guided by strong preliminary data, this hypothesis will be tested by pursuing four specific aims. These involve: 1) Developing an optical spectroscopic system combining single/multiple-scattered light measurements; 2) Implementing reconstruction algorithms to improve parameter extractions based on more accurate Hb/HbO2 calculation models, non-spherical particle scattering models and multi-modal particle size distribution models; 3) Testing the system using simulation experiments, and human subjects; 4) Developing algorithms for skin cancer diagnosis by correlating optical signatures with the pathophysiologic parameters using histomorphometric techniques. The approach is innovative because it utilizes the combined single and multiple-scattered light spectroscopic methods for skin cancer diagnosis. The proposed research is significant, because it is expected to advance a novel spectroscopic method for skin cancer detection by focusing on improving feature extraction algorithms and enhancing classification software. This research is expected to significantly contribute to the early skin cancer screening and detection by maximizing cure rates and reducing and avoiding biopsies. ? ? ?
| Wang, James Z; Zhang, Yuanyuan; Dong, Liang et al. (2014) G-Bean: an ontology-graph based web tool for biomedical literature retrieval. BMC Bioinformatics 15 Suppl 12:S1 |
| Li, Lin; Zhang, Qizhi; Ding, Yihua et al. (2014) Automatic diagnosis of melanoma using machine learning methods on a spectroscopic system. BMC Med Imaging 14:36 |
| Song, Xuebo; Li, Lin; Srimani, Pradip K et al. (2014) Measure the Semantic Similarity of GO Terms Using Aggregate Information Content. IEEE/ACM Trans Comput Biol Bioinform 11:468-76 |
| Du, Zhidian; Li, Lin; Chen, Chin-Fu et al. (2009) G-SESAME: web tools for GO-term-based gene similarity analysis and knowledge discovery. Nucleic Acids Res 37:W345-9 |
