Expanding the dynamic range of x-ray videodensitometry using ordinary image digitizing devices.

While x-ray images on film have a wide dynamic range of up to three optical density units (o.d.) or 1:1000 variation in transparency, many ordinary digitizing video systems have a limited dynamic range of ~1.4 o.d. in actual practice. The digital images thus either lack details in the dark areas or are saturated and bloom in the bright areas and have a poor contrast ratio. We present here an object-dependent digital image processing method to overcome this drawback by appropriately selecting and correcting pixels drawn from two images acquired from the same radiograph at different light levels. A calibrated gray scale is used to compensate for the nonlinearity of the video system and derive the true optical density for each pixel. The method should find applications in digital mammography, teleradiology, archiving of radiographs, and videodensitometry in general.