Applying Data Science in Computer Vision: Detection of Malignant and Benign Cancer Tumors

Tumors are abnormal growths of tissue that can either be malignant or benign. Malignant tumors indicate cancer, while benign tumors are generally minimally harmful. The difference between the two is not evidence at first glance, and many tests must be conducted to identify it. Tumors need to be determined at an early stage, considering that treatments are most effective during the early stages of cancer when it has still not yet infected a lot of cells. Computer Vision is a technology that enables computers to simulate the “perceiving the real world” phenomena. With it, computers can create analyses and conclusions based on different images provided. With computer vision, a good data set is essential. Data Science makes use of different preprocessing techniques that aid in analyzing data. Applying different preprocessing methods to a data set allows the creation of high-quality data that can be used as a training data set for computer vision models. There are various computer vision models; the models that will be discussed are CapsNet, VGG16, ResNet50, and GoogLeNet. Concurrently, their accuracy and performance would be compared to each other. Furthermore, the importance of data science relative to computer vision will also be discussed. The increase in accuracy of having preprocessed high-quality data as a data set for computer vision models will also be measured.