Software Optimization in Ultrasound Imaging Technique Using Improved Deep Belief Learning Network on the Internet of Medical Things Platform

The Internet of Medical Things (IoMT) is the array of medical instruments and related technologies that link Information Technology (IT) systems in health care through online smart interface networks. Presently, Medical images are considered an essential source of a patient's clinical record and may be utilized to analyze infectious disease along with their diagnostic procedure, operations, and prediction. In cancer diagnostics, at an early stage, it is hard to classify using present diagnostics scanning methods such as Computed tomography (CT) and magnetic resonance imaging (MRI). In this research paper, an Improved Deep Belief Learning Network (IDBLN) is introduced in the software application of ultrasound imaging techniques for the effective feature analysis and classification of various cancer images at the early stages. IDBLN contains an improved automatic feature selection framework for a useful classification and detection of cancer images during scanning. An Enhanced Optimized Crow Search algorithm (EOCS) is suggested for classification to boost cancer detection efficiency during image scanning. Finally, outcomes are analyzed based on accuracy, sensitivity, and minimum error rate, proving the proposed framework's performance.