medical diagnosis process using neural networks

Upload: samantha-roop

Post on 14-Apr-2018

221 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/29/2019 Medical diagnosis process using neural networks

    1/24

    Artificial Neural Networks In

    Medical Diagnosis

  • 7/29/2019 Medical diagnosis process using neural networks

    2/24

    Definition

    Artificial Neural network is an information

    processing system composed of highly

    interconnected processing elements

    (neurons) working in union to solve

    specific problems.

  • 7/29/2019 Medical diagnosis process using neural networks

    3/24

    The general structure of artificial neural

    networks consists of neurons interconnected to

    form three types of layers:

    Input Layer (Receives input from outer world)

    One or more Hidden Layers (Process input)

    Output Layer (Gives output to outer world)

  • 7/29/2019 Medical diagnosis process using neural networks

    4/24

    Biological Inspiration

    ANN is Highly interconnected set ofneurons.

    Inspired from biological neurons.

    Output is produced by applying activationfunction on the input.

  • 7/29/2019 Medical diagnosis process using neural networks

    5/24

    Activation functions

    Linear Function

    Logistic Function

    Linear (x) = x

    Logistic (x) = 1/(1+exp(-x))

  • 7/29/2019 Medical diagnosis process using neural networks

    6/24

    Sign Function

    Step (x) = 1 if x >= t, else 0

    Step Function

    Sign (x) = +1 if x >= 0, else -1

  • 7/29/2019 Medical diagnosis process using neural networks

    7/24

    Transformation

    How input is transformed to outputs:

  • 7/29/2019 Medical diagnosis process using neural networks

    8/24

    Architecture

    Neurons are linked together in layers to

    form a specific architecture.

    Transforms inputs into meaningful outputs.

    Artificial Neural network can be broadly of

    two types:

    Feed Forward Architecture

    Recurrent Architecture

  • 7/29/2019 Medical diagnosis process using neural networks

    9/24

    Feed Forward Neural Networks

    The information ispropagated from theinputs to the outputs

    Computations of Non

    linear functions from ninput variables.

    There is NO cyclebetween outputs andinputs.

    x1 x2 xn..

    1st hidden

    layer

    2nd hidden

    layer

    Output layer

  • 7/29/2019 Medical diagnosis process using neural networks

    10/24

    Recurrent Neural Networks

    Can have arbitrarytopologies.

    Training is more

    difficult. There can be cycles

    between outputs andinputs.

    x1 x2

    1

    0

    10

    1

    0

    0

    0

  • 7/29/2019 Medical diagnosis process using neural networks

    11/24

    Training

    Every neuron in ANN has a weight

    associated with it.

    Training takes place by giving different

    instance inputs.

    Weights are adjusted until the desired

    results are got as output.

  • 7/29/2019 Medical diagnosis process using neural networks

    12/24

    Training can be performed in two modes: Supervised

    Unsupervised

    Some training algorithms of common useare: BPN, MLP, RBF, LVQ etc.

    Training is continued until a desired

    accuracy is achieved or error comes to aminimum level.

  • 7/29/2019 Medical diagnosis process using neural networks

    13/24

    Neural Networks in Medical Field

    Disease Diagnosis

    Prediction of chances of a disease

    Robotic Handling of Surgery Measuring Effect of new Medicines

    Continuous Patient Monitoring

  • 7/29/2019 Medical diagnosis process using neural networks

    14/24

    Collect Samples

    Take both normal and abnormal

    samples of images or data from tests.

    More the number of samples, more will

    be accuracy.

    Type of image depends upon the nature

    of research.

  • 7/29/2019 Medical diagnosis process using neural networks

    15/24

    Preprocessing

    Preprocessing is necessary for renderingthe image fit for analysis.

    If one is using data then image processing

    steps are not required. Various methods used by different

    researchers are:

    Enhancement Filtering

    Compression etc..

  • 7/29/2019 Medical diagnosis process using neural networks

    16/24

    Extract Features

    Extract features of the image that are

    necessary for diagnosis.

    Difference between left and right eye

    features will be used for diagnosis.

    In case of data, important parameters are

    selected.

  • 7/29/2019 Medical diagnosis process using neural networks

    17/24

    Normalize

    A data set is created by normalizing thefeatures and actual diagnosis results.

    S. No. Features Result

  • 7/29/2019 Medical diagnosis process using neural networks

    18/24

    Training And Testing Set

    Dataset can be randomly divided into

    mutually exclusive training and testing

    sets.

    Training setis usually larger than the

    testing set.

    Size depends on number of available

    samples.

  • 7/29/2019 Medical diagnosis process using neural networks

    19/24

    K-Fold Cross Validation

    Another method for selecting training and

    testing sets.

    Samples are divided into K mutually

    exclusive subsets of equal size.

    K-1 subsets are used for training and one

    subset is used for testing.

    The procedure is repeated K times taking

    different subset for testing.

  • 7/29/2019 Medical diagnosis process using neural networks

    20/24

    Design of Neural Network

    Number ofinputs and outputs.

    Number ofhidden layers.

    Number ofneurons in each hidden layer. Interconnections between different

    layers.

    Training algorithm to be used. Initial weights.

  • 7/29/2019 Medical diagnosis process using neural networks

    21/24

    Training and Testing

    Input the training set along with the

    results.

    Training is performed until weights

    continue to change.

    After training, test set is introduced but

    the results are not fed along with.

    Now the neural network has to perform

    similar operations and produce diagnosis.

  • 7/29/2019 Medical diagnosis process using neural networks

    22/24

    System Evaluation

    Methods such as confusion matrix can

    be used to evaluate system performance.

  • 7/29/2019 Medical diagnosis process using neural networks

    23/24

    Evaluation Parameters

    Sensitivity=

    Specificity=

    Accuracy=

  • 7/29/2019 Medical diagnosis process using neural networks

    24/24

    THANKS