Automatic Diagnosis of Lung Diseases (Pneumonia, Cancer) with given Reliabilities on the Basis of an Irradiation Images of Patients

What is it about?

The article proposes algorithms for the automatic diagnosis of human lung diseases pneumonia and cancer, based on images obtained by radiation irradiation, which allow us to make decisions with the necessary reliability, that is, to restrict the probabilities of making possible errors to a pre-planned level. Since the information obtained from the observation is random, Wald’s sequential analysis method and Constrained Bayesian Method (CBM) of statistical hypothesis testing are used for making a decision, which allow us to restrict both types of possible errors. Both methods have been investigated using statistical simulation and real data, which fully confirmed the correctness of theoretical reasoning and the ability to make decisions with the required reliability using artificial intelligence. The advantage of CBM compared to Wald’s method is shown, which is expressed in the relative scarcity of observation results needed to make a decision with the same reliability. The possibility of implementing the proposed method in modern computerized X-ray equipment due to its simplicity and promptness of decision-making is also shown.

Featured Image

Why is it important?

A method of automatic diagnosis of pneumonia and lung cancer with computerized X-ray equipment is proposed, which requires very little memory and time to make a decision. At the same time, both types of possible errors can be limited to predetermined levels with guarantee. The method is based on the method of sequential analysis of Bayesian type of statistical hypothesis testing. The results of the experimental investigation, both on modeled and real data, showed the ease of implementation, high reliability and accuracy of the proposed method of automatic diagnosis. In our opinion, the implementation of the mentioned method in serially produced relevant equipment will significantly increase the quality of the diagnosis, which in turn will play a decisive role in the final recovery of the patient. It should be noted here that the implementation of this method in serially produced modern relevant equipment, provided with microprocessor equipment, requires insignificant time and material costs. If necessary, we can provide the relevant computer program implemented on MATLAB to the interested party.