A method for automated detection, monitoring and treatment of dysplasia by analyzing 3D reconstructed images of cells obtained from a specimen includes a biological specimen classifier that classifies cells from the sputum specimen as normal or abnormal. If abnormal cells are detected, then the abnormal cells are further classified as pre-cancerous or cancerous. Pre-cancerous cells are further sub-classified as being of glandular origin or squamous origin (dysplastic cells). This information would be used to define patient therapy so that if the cells are classified as dysplastic, then a cancer chemoprevention pharmaceutical like iloprost is administered to the subject over a predetermined time period to achieve a therapeutic dosage, and if only malignant cells were found or malignant and pre-cancerous cells were found, then next steps would involve patient triage to biopsy and surgery and possibly use of a cancer chemoprevention pharmaceutical.
