Objective    To explore the algorithm of the shrunk lung image restoration.   Methods    Based on 2-D section image

combining with the morphological characteristics of the living lung tissue

the direction of lung deformation and the boundary line that deformation reaches was assumed

and vein filling of specific direction in irregular areas was achieved through polar coordinates

Bresenham drawing method and interpolation method

etc.   Results   After the lung dilation

the image was clear

overcoming many shortcomings of the traditional image amplification

such as obscure and zigzag boundary

and the image reflected the location and size changes of various structures within the lung under specific movement trial.   Conclusion   The method

with image processing

achieves lung marking amplification

and maintains the adjacent relationship among the interior organization structures

which can effectively restore the lung from shrinkage condition to normal condition.