Segmentation and feature extraction of sputum cell for early detection of lung cancer

Diagnosis of lung cancer in its primal stage is a major problem faced by the medical world. For that proper details are needed from the images, which can only be obtained by a good segmentation method. However, many common forms of techniques are available in market and their major drawback is the accuracy of segmentation of the nucleus from the ROI and also the time consumed for the same. Since this pose to be a great problem, most of the techniques shrink to this phase only. In this paper we introduce a new type of cell image segmentation which works on the PAP stained sputum cytology images. This allows a very simple formulation, obviating the need for additional methods. The subsequent phase of feature extraction and classification is also done accordingly. Due to its simplicity the algorithm is fast and very robust. Our method demonstrates on sputum cytology images.