Research And Implementation On Enhancement Of Renal Biopsy Pathological Images

With the increasing aging of Chinese population and the improvement of the social life,the incidence and mortality rate of chronic kidney disease(CKD)have increased annually.Defining the cause of kidney disease and taking early intervention strategies could reduce the incidence and mortality of CKD to some extent.Among them,renal biopsy pathology is an important method to diagnose the cause of CKD.Due to the complexity of the biopsy production procedure,long-time production and the inevitable unsatisfactory production procedure,making the pathological image may have degradation problems such as uneven brightness and poor contrast,which directly affects pathologists diagnosis.In view of the above problems,an enhanced algorithm based on adaptive bilateral logarithm transformation with bandwidth preserving has been proposed in this paper,on account of the analysis of image grayscale or chroma spectrum and the reconstruction of classical logarithm transformation.The proposed method includes three steps.Firstly,the degraded pathological image is transformed into a standardized image by a special intensity transformation called the standard transformation.Then,the adaptive bilateral logarithm transformation with preserved bandwidth is performed according to the average luminance(AL)of the standardized image.Finally,the enhanced renal biopsy pathological image is obtained through a rounding step.For the purpose of further verifing the effectiveness and feasibility of the algorithm,20 degraded renal biopsy pathological images were selected in the experiment.The proposed algorithm is compared with other enhancement algorithms including Brightness Preserving Dynamic Fuzzy Histogram Equalization(BPDFHE)and homomorphic filtering algorithm.Qualitative and quantitative analyses are conducted to evaluate the proposed algorithm.Experimental results show that the overall contrast of the enhanced images of the proposed algorithm are improved subjectively,and the image details are also significantly enhanced.Simultaneously,the average information entropy(IE)and standard deviation(STD)of the objective evaluation index of the algorithm are 7.12 and 33.20,respectively,which higher than the others.In addition,the average running time of the proposed algorithm is relatively short.This proposed algorithm can be used for image enhancement in renal biopsy,and assist the pathologist to make correct diagnosis quickly.It can also be used for preprocessing steps such as cell segmentation and lesion location.