A Review on Automatic Lung Lesion Detection from Various Imaging Techniques

Lung cancer is a horrible disease that may take human life. It is the most analyzed disease on the planet that may be considered as life-threatening disease. If it can be diagnosed earlier then treatment is a solution that may save human life. There are different imaging strategies through which this disease can be diagnosed and treated accordingly. However, a Computed Tomography image commonly known as a CT scan image is the better option for diagnosing disease with a better level of accuracy as compared to other imaging techniques such as X-ray, Ultrasound, and many more. If a disease can be diagnosed automatically then it saves medical professional time as well as human life. A routine checkup can become easier and it can be processed in less time. The intention of this paper is to review various previously implemented systems related to the automatic diagnosis of lung cancer. There are so many researches that have been done that are based on CNN, DNN, Edge Detection, and various image processing or machine learning techniques. The objective of this work is to define the limitations and drawbacks of various systems that are lacking somewhere.