Lung cancer is the leading cause of mortality worldwide. On an annual basis, lung cancer is responsible for more deaths than all other prevalent types of cancer combined. While there have been great strides made in the field of healthcare, this issue persists. Although most instances are discovered at stages 3 and 4, by which time it's much too late to treat successfully, early detection is key. The mortality rates associated with lung cancer are among the highest of any cancer kind. Consequently, early detection of lung nodules is crucial for improving survival rates. Some CAD systems can detect and categorize these nodules at an early stage. Data collecting, pre-processing, lung segments, nodule identification, false positive reductions, edge detection, and classifications are all now included in CAD systems for lung nodules. The purpose of this work is to develop an effective CAD system for segmenting a CT image of the lung, which will aid radiologists in the early detection and diagnosis of lung cancer. To improve nodule classification, a novel 3-D convolutional neural network (CNN) is employed to segment the CT image. Multiple enhancements were developed to guarantee lightning-fast communication and pinpoint precision in the final output. Specifically, the LUNA16 challenge's LIDC/IDRI database is performed to analyze the design. Our results show that deep learning with focused loss is a superior classification technique, with an accuracy of 97.10%, a sensitivity of 98.00%, and a precision of 97.93%.