Optical automation for sputum cytology

The use of high-resolution image analysis in the detection and classification of cellular atypias has been demonstrated in many instances. We have applied this technique to the classification of bronchial epithelial atypias from sputum. Our research indicates that, in classifying cells into one of five atypia stages, 90% of the cell classifications were within one category of the criterion stage, while 88.6% of the subjects were correctly diagnosed. These results indicate the feasibility of using high-resolution cell image analysis to monitor the progression of squamous cell carcinoma of the lung and, it is to be hoped, to permit the reversal of that process.