An Artificial Intelligence-Based Platform for Medical Diagnosis

Medical diagnosis using artificial intelligence models uses methods of information collection via the Internet of Things (IoT) for the categorization of diseases. This may be attributed to a number different issues, such as inefficient auxiliary frameworks, high expenses associated with acquiring datasets, or difficulties encountered while constructing classifiers. The present paper analyses the recent developments in AI for medical diagnosis. The research uses a Deep Convolutional Neural Network (DCNN) to categorise medical conditions. In this paper, a model for categorising medical diagnosis has been introduced by making it easier for the network to adapt to new medical data. This article estimates the degree to which the classification falls into the right medical diagnosis category and separates the training sample into normal, critical, and suggestion samples using a dynamic threshold to solve the diagnostic issue. The recommended strategy categorises input to help the convolutional neural network learn more. This research modifies the convolutional neural network design to accommodate for input variety and data evolution temporal dynamics. This adjustment is made with the goal of achieving the five physiological data properties of bio signals (heart rate, blood pressure, EEG, ECG, and oxygen level). Because of the realised CNN optimisation algorithm model, the prediction effect has been increased, and the accuracy rate has been found to be 92.8% in medical diagnosis, which is a reasonably good performance in machine learning algorithms.