An approach for Lung Cancer Pattern Recognition in Computer Aided Detection Systems

A computerized medical image analysis technology suffers from imperfection, imprecision and vagueness of the input data in medical images and its propagation in all individual components of the technology including image processing and pattern recognition. In addition to the above, a Computerized Medical Image Analysis System such as computer aided detection (CAD) technology deals with another source of uncertainty which comes from inter and intra uncertainties in medical diagnosis. These sources of uncertainties in disease pattern recognition in medical images incur to incorrect diagnosis which involves the human life. Therefore, modeling uncertainties in disease pattern recognition in computer aided detection systems is a very vital task to be taken into account in developing a CAD system. In this paper, the sources of the uncertainty in the design of a lung CAD system using CT images of thoracic will be addressed. A fuzzy model is proposed to tackle the problem of uncertainty in the pattern recognition component of the lung CAD system. The result is promising to improve the performance of nodule pattern diagnosis in the lung and improve medical diagnosis using CAD systems.