Microarray technology has increased exponentially in the past few years. Such technology encompasses various applications related to drug design, tumor detection, clinical diagnosis and treatment, and environmental health research. However, the segmentation of the spots in microarray image is still a problem due to variations of spots qualities such as spots shapes and sizes for example. Hence, we have introduced a new microarray spot segmentation algorithm based on Integro-Differential Operator (IDO). Because it includes a spot detection step, this IDO can be used to find any spot regardless of its size and shape within the microarray image. The IDO algorithm is applied on the cDNA microarray images to improve the accuracy and the efficiency of the spots segmentation process. As the IDO is able to segment each spot based on the fact that the illumination difference between the inside and the outside of the pixels in spot edge circle is maximum. We have also tested our algorithm on the Stanford Microarray Database (SMD). The numerical results show that, the IDO is a powerful technique to improve the overall spots segmentation on the microarray image and to also give better results in spot segmentation and analysis process.

在过去几年中，微阵列技术呈指数增长。此类技术涵盖了与药物设计，肿瘤检测，临床诊断和治疗以及环境健康研究相关的各种应用。然而，由于斑点质量的变化，例如斑点形状和大小，微阵列图像中斑点的分割仍然是一个问题。因此，我们引入了一种新的基于积分微分算子（IDO）的微阵列点分割算法。因为它包括斑点检测步骤，所以此IDO可以用来查找任何斑点，无论其在微阵列图像中的大小和形状如何。IDO算法应用于cDNA微阵列图像，以提高斑点分割过程的准确性和效率。由于IDO能够基于光斑边缘圆中像素内部和外部之间的照度差异最大这一事实对每个光斑进行分割。我们还在斯坦福大学微阵列数据库（SMD）上测试了我们的算法。数值结果表明，IDO是一种强大的技术，可以改善微阵列图像上的整体斑点分割，并且在斑点分割和分析过程中也能提供更好的结果。