Artificial intelligence-based leukemia screening and diagnosis

Many blood disorders are caused by a deficiency or excess of various blood factors. For example, if a person does not have enough red blood cells, they will have anemia. Also, when the number of white blood cells is lower than normal, a person will have leukopenia. Blood is a living tissue that is made up of four main parts: plasma, red blood cells, white blood cells, and platelets. Plasma is the liquid part of the blood. This part contains water, protein, and salt. Blood diseases cause disorders in any of these parts. Some diseases may also affect several parts of the blood. Most of these blood diseases are related to genetic disorders. But sometimes other diseases, the use of certain medications, and a diet with very few nutrients can cause such disorders. This article shows an innovative use of artificial neural networks in medicine. The proposed technique involves training a multilayer perceptron (a type of artificial neural network) with a BP learning algorithm to recognize patterns and predict five blood disorders from blood test results from the H۱ device. Blood test parameters and physician diagnoses of ۴۵۰ patients from Taleghani Hospital in Kermanshah, Iran, are used in a supervised learning method to update the network parameters. The method is implemented to diagnose the following disorders and cancers: megaloblastic anemia, thalassemia, idiopathic thrombocytopenic purpura (ITP), chronic myelogenous leukemia, and lymphoproliferative disease. Many blood disorders are caused by a deficiency or excess of various blood factors. For example, if a person does not have enough red blood cells, they will have anemia. Also, when the number of white blood cells is lower than normal, a person will have leukopenia.

کلیدواژه ها:

Artificial intelligence ، leukemia ، blood disorders ، perceptron ، neuron ، neural network