مجلة الدراسات العليا

Permanent URI for this communityhttps://repository.neelain.edu.sd/handle/123456789/1519

Browse

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Detection and Classification of Anemia using Artificial Neural Networks: Comparison of Three Models
    (Graduate College, 2017-10-01) Esra Mohammed Ibrahim, Mohamed A. A. Elmaleeh; Daalia Mahmoud
    This paper presents different classifier algorithms neural network to diagnose and classify anemia and compare these current classifier with feed forward back propagation, Elman network and Non-linear Auto-Regressive exogenous model . The results obtained from a range of models while conduction the experiments. The proposed network is trained by using the data received from clinical laboratory test results for 230 patients. The network has nine inputs (age, gender, RBC, HGB, HCT, MCV, MCH, and MCHC, WBCs) and an output. The simulation results obtained for different patients show that the proposed artificial neural network detects the disease very fast, and precise. Therefore this network can be implemented to automatically present the anemia patients’ reports in the clinical laboratory. The proposed technique can be implemented in hardware with minimal cost.
  • Thumbnail Image
    Item
    Differential Anemia Predation via Feed Forward Neural network
    (Graduate College, 2017-10-01) Esra Mohammed Ibrahim, Mohammed A. A. Elmaleeh
    Anemia is a one of the most common diseases which extensively scattered infamy of the least developing countries. The current methods of screening Anemia is carried out through complete blood tests received from clinical laboratory using manual approach for analyzing patient lead to high percentage of false positive test result of complete blood cell count (CBC) which can be reduced by employing intelligent Artificial Feed forward Neural Networks(FFNN).The proposed system was based on FFNN multi-layer perception (MLP) algorithm with 9 inputs, 50 neurons, and 1 hidden layer. It was used as the most powerful differential diagnosis instrument with accuracy as 88.0% for learning data, 87.5% for validation and test, and 86.5 % for overall networks. The FFNN detects anemia disease with high accuracy . Keywords: Anemia, complete Blood count, Feed Forward Neural Networks.