通过研究染色体自动分析技术在评估儿童急性淋巴细胞白血病预后中的应用,提出了一种基于数学形态学的染色体自动计数算法。算法首先对输入的染色体图像进行了预处理,接着用标号法给图像中的每个对象赋予惟一的标号。为了消除图像中噪声对统计结果的影响,按照设定的规则将噪声对象删除。对于舍有粘连染色体的对象,通过腐蚀运算计算出实际染色体个数,并根据结果修正标号值,修正结果即为图像中的染色体数目。实验证明,该算法可以自动统计出图像中的染色体数目。
An automatic chromosome counting algorithm based on mathematical morphology is presented after the application of the chromosome automatic analysis technique is studied to evaluate the prognosis for childhood acute lymphoblastic leukemia. Firstly, the pre-processing is used to handle input images. Then, the label algorithm is utilized to assign an exclusive label to each object in the image. To eliminate the influence of noise, the labelled noise objects are deleted based on the given qualification. For the objects with conglutinate chromosomes, the number of chromosomes contained in them is computed by the erodition algorithm. And the value of the label is adjusted according to the computational results. The adjusted result is regarded as the number of chromosomes in the image. The experiment proves that the algorithm can automatically count the number of chromosomes in an image.