Multi organ tumor detection using deep learning

Computed tomography (CT) or magnetic resonance imaging (MRI) images of various organs are frequently used to identify tumors or other masses using machine learning. The primary objectives of deep learning in MRI have always been segmenting and categorizing reconstructed magnitude images. In order to identify malignancies early and to provide crucial details like location and size with a high level of accuracy, this research offers a deep-learning model that learns to divide various organs and tumors on datasets into their appropriate categories. The outcomes show that the proposed model performs effectively when applied to an untested brain MRI dataset. In order to confirm the generated transfer-learned model’s flexibility, general capacity, and dependability for real-time application in the future, it will also be evaluated using the CT scans of a different machine. UNet is used to carry out effective automated tumor detection. Python is used to carry out the simulation. The results show an accuracy range of 95-97% with the input ni files. Future real-time diagnosis applications can use the suggested architecture because of its strength, adaptability, generalizability, and high accuracy. Hence, the suggested deep-learning architecture can aid in early tumour diagnosis for medical professionals.