Lung Nodule Segmentation Using Machine Learning and Deep Learning Techniques

Global lung cancer mortality is growing. This supports early cancer screenings. CT lung nodule segmentation is complicated and affects medical research, surgical planning, and diagnostic decision support. All are complex issues with important applications. Machines and humans struggle to split non solitary nodules with uncertain boundaries. Since segmentation has distinct limits, single nodules are easier to divide. Several researchers have proposed CT-based lung evaluation algorithms. Growing imaging datasets and the need to swiftly and precisely define normal and diseased lung lobes are the reasons. Multi-process lung segmentation methods with manual empirical parameter modifications are common. First lung slice and nodule segmentation using ML and DL is essential for cancer detection. This detects cancer at various stages. Deep learning techniques have improved healthcare image analysis. There are few deep learning approaches like ResNet 50,101, VGG16, Autoencoders, U-Net with modifications, and graph convolutional networks to classify lung nodules, COVID-19, and pneumonia. This chapter includes a summary of datasets that are open to the public and are the primary resources utilized by scholars working in this area. A direct look into the field of diagnosing lung disorders is what we hope to achieve with the information provided in this chapter.