In India, millions of people are suffering from different respiratory diseases, and globally, billions of people are suffering. The main issue in identifying and resolving these diseases is a lack of awareness and costly treatment. Cough is the first identification feature in any respiratory disease. The cough sounds give important information about the respiratory system and the bacteria included. In this chapter, we will analyze the different cough sounds to detect the problem at early stages on the basis of various sound signals. Here, we discuss standard signal processing and domain-specific features in different cough patterns. These cough patterns include widened airways, narrow airways, fluid-filled air sacs, and stiff lungs. The main objective is to review and analyze cough attributes and observe how they differ in various diseases using machine learning (ML). As per the observation, the duration, intensity, type of cough, and frequency of cough are different in different diseases.