Automated Technique Based on Image Processing for Accurate Diagnosis of Finger Clubbing

Medical observation has revealed that lung cancer, congenital heart defects, chronic lung infection, infection of the lining of heart chambers and a large number of other disorders mostly related to the lung and the heart is accompanied by the phenomenon of the “clubbing of the fingers”. It is actually a distortion of the general shape of the finger that now takes a bulging appearance, resembling a drumstick or a bird's beak. Therefore finger clubbing can be treated as an early symptom of the onset of many such disorders. In this work, we have developed a software based on image processing techniques for the quantitative detection of finger clubbing. The image processing techniques undertaken in this work include lovi-bond angle detection and finger width profile monitoring by measuring Distal Phalange Depth-Inter Phalange Depth (DPD-IPD) ratio. Thus the system developed in this work provides an early manifestation of such disorders so that early diagnosis and effective treatment can be undertaken and the risk factor in such cases can be substantially reduced.