Malignant hepatic tumors are tumors that result in high rates of morbidity and mortality. Improvement of therapeutic efficacy and the survival rate from hepatic tumors is an important problem, and the key issue is early diagnosis of hepatic tumors. Conventional methods of imaging diagnosis for hepatic tumors, such as sonography 
, magnetic resonance imaging (MRI) 
, computed tomography (CT) 
, digital subtraction angiography (DSA) 
, X-ray microtomography, X-ray fluorescence imaging 
and other methods, have relatively high rates of diagnostic accuracy for advanced tumors, and their imaging resolutions are only at the millimeter level. However, the early diagnosis of hepatic tumors, especially in sub-clinical diagnosis, may be more difficult. It is easy to overlook small liver tumors of diameters less than 1 mm. Needle biopsy is the most commonly used method for some disease diagnoses. However, this technique is invasive 
, and some small hepatic tumors may be overlooked by needle biopsy, which results in serious consequences because the best time for treatment is missed.
There are many other areas of research in the early detection of hepatic tumors. For example, gold nanoparticles in material science are studied as a hepatic tumor contrast agent 
to enhance images for the location of nanometer-scale hepatocellular carcinomas. Low-angle X-ray scattering is applied in the detection of structural changes in the serum proteins of patients 
. This method may detect and treat early hepatic tumors. X-ray ILPCI is a type of imaging that is being researched for detecting early hepatic tumors.
As a new imaging method, X-ray phase-contrast imaging (XPCI) has high spatial resolution and density resolution, which can provide high contrast images by using the phase shift of the X-ray. The density resolution for C, H, O and other light elements is approximately 1,000 times higher than that of traditional X-ray absorption imaging. This technique can greatly improve the image quality of soft tissues, particularly at the interface of tissues, where the refractive index changes significantly 
. Therefore, soft tissue imaging using XPCI has some potential in clinical applications. The phase-contrast depends on X-ray coherent scattering rather than absorption; consequently, XPCI can reduce potential radiation damage to tissues 
. Recently, this technique has been widely used by researchers for imaging small animals. XPCI has four approaches: X-ray interferometer 
, diffraction enhanced imaging (DEI), ILPCI and X-ray grating interferometer 
. X-ray ILPCI 
has become a major focus of current research. Its imaging conditions are the simplest, making it more suitable for clinical applications than other methods 
. X-ray ILPCI has previously been used to study the soft tissues of both humans and small animals, such as mice, rats, rabbits and hamsters 
. The results have been satisfactory, and high resolution images have been obtained 
. X-ray ILPCI may be an alternative method for observing hepatic tumors without contrast agents, and it is totally noninvasive. The micro-CT image resolution of mouse liver tumors in animal experiments is 9 μm when a contrast agent is injected 
. The X-ray ILPCI image resolution of nude mouse liver tumors can be 3.7 μm without a contrast agent. MRI is limited by the magnetic strength, so its spatial resolution is difficult to increase. X-ray fluorescence imaging mainly images the fluorescence of heavy metals in biological tissues, so the heavy metal content of the tissue determines the effect of the X-ray fluorescence imaging. X-ray fluorescence imaging 
mainly reflects the heavy metal content rather than the specific shape of the tumor, and it does not clearly detect blood vessels. X-ray ILPCI can allow the observation of the tumor shape and the blood vessel distribution around a tumor. X-ray ILPCI and X-ray fluorescence imaging are still in the experimental stages, and it may be a long time before research advances to the clinical application stage.
Texture is the visual perception of local features on an image. Texture analysis of medical images is a sophisticated computer-aided technique that allows the detection of mathematical modes in the gray level distribution of pixels in digital images. It provides an objective characterization of the signal behavior of anatomical structures or pathological processes. The texture features of an image reflect the spatial distribution of the pixel properties, and they usually have irregular local and regular macroscopic characteristics. The texture features of a part of an image are closely related to the gray value changes in this region. A smooth region of an image contains pixels whose gray values are similar to one another, whereas the gray values of the pixels in a rough region differ dramatically 
. Images need to be preprocessed before the extraction of texture features. Image preprocessing does not require prior knowledge and does not affect the subsequent extraction of texture features. Texture parameters are extracted as image features and calculated quantitatively. Combined with some classification algorithms, such as SVM, back propagation (BP) neural network, etc., texture parameters can distinguish different biological tissues or different pathological conditions within the same biological tissues 
. When texture features based on GLCM and DTCWT are quantitatively analyzed, normal regions and tumor regions in an X-ray ILPCI image can be distinguished, and different stages of hepatic tumors can be classified by the SVM technique.