From Phil Cagle and colleagues: Moving toward bedside diagnosis of lung cancer?

http://spiedigitallibrary.org/jbo/resource/1/jbopfo/v16/i9/p096004_s1?isAuthorized=no

On-the-spot lung cancer differential diagnosis by label-free, molecular vibrational imaging and knowledge-based classification
J. Biomed. Opt. 16, 096004 (Sep 01, 2011); doi:10.1117/1.3619294
Liang Gao, Fuhai Li, Yaliang Yang, Jiong Xing, Ahmad A. Hammoudi, Hong Zhao, Yubo Fan, Kelvin K. Wong, Zhiyong Wang, and Stephen T. C. Wong
Weill Cornell Medical College, The Methodist Hospital Research Institute, Department of Systems Medicine and Bioengineering, Houston, Texas 77030
Rice University, Department of Bioengineering, Houston, Texas, 77005
Michael J. Thrall, Philip T. Cagle,
The Methodist Hospital and Weill Cornell Medical College, Department of Pathology and Laboratory Medicine, Houston, Texas 77030
Yehia Massoud,
Rice University, Department of Electrical and Computer Engineering, Houston, Texas, 77005

We report the development and application of a knowledge-based coherent anti-Stokes Raman scattering (CARS) microscopy system for label-free imaging, pattern recognition, and classification of cells and tissue structures for differentiating lung cancer from non-neoplastic lung tissues and identifying lung cancer subtypes. A total of 1014 CARS images were acquired from 92 fresh frozen lung tissue samples. The established pathological workup and diagnostic cellular were used as prior knowledge for establishment of a knowledge-based CARS system using a machine learning approach. This system functions to separate normal, non-neoplastic, and subtypes of lung cancer tissues based on extracted quantitative features describing fibrils and cell morphology. The knowledge-based CARS system showed the ability to distinguish lung cancer from normal and non-neoplastic lung tissue with 91% sensitivity and 92% specificity. Small cell carcinomas were distinguished from nonsmall cell carcinomas with 100% sensitivity and specificity. As an adjunct to submitting tissue samples to routine pathology, our novel system recognizes the patterns of fibril and cell morphology, enabling medical practitioners to perform differential diagnosis of lung lesions in mere minutes. The demonstration of the strategy is also a necessary step toward in vivo point-of-care diagnosis of precancerous and cancerous lung lesions with a fiber-based CARS microendoscope.

© 2011 Society of Photo-Optical Instrumentation Engineers (SPIE)