medicalai.chief.nnets package

Submodules

medicalai.chief.nnets.covid_net module

medicalai.chief.nnets.covid_net.COVIDNET_Keras(img_input=(224, 224, 3), classes=4)

This is a tensorflow 2.0 network variant for COVID-Net described in Paper “COVID-Net: A Tailored Deep Convolutional Neural Network Design for Detection of COVID-19 Cases from Chest Radiography Images” by Linda Wang et al. Reference: https://github.com/busyyang/COVID-19/

medicalai.chief.nnets.covid_net.PEPXModel(input_tensor, filters, name)

medicalai.chief.nnets.densenet module

medicalai.chief.nnets.densenet.DenseNet121_Model(img_input=(224, 224, 3), classes=3)

Loaded the DenseNet121 network, ensuring the head FC layer sets are left off

• Parameters

  • img_input – optional shape tuple, only to be specified if include_top is False (otherwise the input shape has to be (224, 224, 3) (with ‘channels_last’ data format) or (3, 224, 224) (with ‘channels_first’ data format). It should have exactly 3 inputs channels, and width and height should be no smaller than 32. E.g. (200, 200, 3) would be one valid value.

  • classes – Number of classes to be predicted.

  • Returns – model

medicalai.chief.nnets.inceptionResnet module

medicalai.chief.nnets.inceptionResnet.InceptionResNetV2_Model(img_input=(224, 224, 3), classes=3)

Loaded the InceptionResNetV2 network, ensuring the head FC layer sets are left off

• Parameters

  • img_input – optional shape tuple, only to be specified if include_top is False (otherwise the input shape has to be (224, 224, 3) (with ‘channels_last’ data format) or (3, 224, 224) (with ‘channels_first’ data format). It should have exactly 3 inputs channels, and width and height should be no smaller than 32. E.g. (200, 200, 3) would be one valid value.

  • classes – Number of classes to be predicted.

  • Returns – model

medicalai.chief.nnets.inceptionv3 module

medicalai.chief.nnets.inceptionv3.InceptionV3(img_input=(224, 224, 3), classes=3)

Loaded the InceptionV3 network, ensuring the head FC layer sets are left off

• Parameters

  • img_input – optional shape tuple, only to be specified if include_top is False (otherwise the input shape has to be (224, 224, 3) (with ‘channels_last’ data format) or (3, 224, 224) (with ‘channels_first’ data format). It should have exactly 3 inputs channels, and width and height should be no smaller than 32. E.g. (200, 200, 3) would be one valid value.

  • classes – Number of classes to be predicted.

  • Returns – model

medicalai.chief.nnets.mobilenet module

medicalai.chief.nnets.mobilenet.MobileNet(img_input=(224, 224, 3), classes=3)

Loaded the MobileNet network, ensuring the head FC layer sets are left off

• Parameters

  • img_input – optional shape tuple, only to be specified if include_top is False (otherwise the input shape has to be (224, 224, 3) (with ‘channels_last’ data format) or (3, 224, 224) (with ‘channels_first’ data format). It should have exactly 3 inputs channels, and width and height should be no smaller than 32. E.g. (200, 200, 3) would be one valid value.

  • classes – Number of classes to be predicted.

  • Returns – model

medicalai.chief.nnets.mobilenetv2 module

medicalai.chief.nnets.mobilenetv2.MobileNetV2(img_input=(224, 224, 3), classes=3)

Loaded the MobileNetV2 network, ensuring the head FC layer sets are left off

• Parameters

  • img_input – optional shape tuple, only to be specified if include_top is False (otherwise the input shape has to be (224, 224, 3) (with ‘channels_last’ data format) or (3, 224, 224) (with ‘channels_first’ data format). It should have exactly 3 inputs channels, and width and height should be no smaller than 32. E.g. (200, 200, 3) would be one valid value.

  • classes – Number of classes to be predicted.

  • Returns – model

medicalai.chief.nnets.resnet module

medicalai.chief.nnets.resnet.conv_building_block(input_tensor, kernel_size, filters, stage, block, strides=(2, 2), training=None)

A block that has a conv layer at shortcut.

• Parameters

  • input_tensor – input tensor

  • kernel_size – default 3, the kernel size of middle conv layer at main path

  • filters – list of integers, the filters of 3 conv layer at main path

  • stage – integer, current stage label, used for generating layer names

  • block – current block label, used for generating layer names

  • strides – Strides for the first conv layer in the block.

  • training – Only used if training keras model with Estimator. In other scenarios it is handled automatically.

• Returns

  Output tensor for the block.

Note that from stage 3, the first conv layer at main path is with strides=(2, 2) And the shortcut should have strides=(2, 2) as well

medicalai.chief.nnets.resnet.identity_building_block(input_tensor, kernel_size, filters, stage, block, training=None)

The identity block is the block that has no conv layer at shortcut.

• Parameters

  • input_tensor – input tensor

  • kernel_size – default 3, the kernel size of middle conv layer at main path

  • filters – list of integers, the filters of 3 conv layer at main path

  • stage – integer, current stage label, used for generating layer names

  • block – current block label, used for generating layer names

  • training – Only used if training keras model with Estimator. In other scenarios it is handled automatically.

• Returns

  Output tensor for the block.

medicalai.chief.nnets.resnet.resnet(num_blocks, img_input=None, classes=10, training=None)

Instantiates the ResNet architecture.

• Parameters

  • num_blocks – integer, the number of conv/identity blocks in each block. The ResNet contains 3 blocks with each block containing one conv block followed by (layers_per_block - 1) number of idenity blocks. Each conv/idenity block has 2 convolutional layers. With the input convolutional layer and the pooling layer towards the end, this brings the total size of the network to (6*num_blocks + 2)

  • classes – optional number of classes to classify images into

  • training – Only used if training keras model with Estimator. In other

  • it is handled automatically. (scenarios) –

• Returns

  A Keras model instance.

medicalai.chief.nnets.resnet.resnet110(*, num_blocks=110, img_input=None, classes=10, training=None)

Instantiates the ResNet architecture.

• Parameters

  • num_blocks – integer, the number of conv/identity blocks in each block. The ResNet contains 3 blocks with each block containing one conv block followed by (layers_per_block - 1) number of idenity blocks. Each conv/idenity block has 2 convolutional layers. With the input convolutional layer and the pooling layer towards the end, this brings the total size of the network to (6*num_blocks + 2)

  • classes – optional number of classes to classify images into

  • training – Only used if training keras model with Estimator. In other

  • it is handled automatically. (scenarios) –

• Returns

  A Keras model instance.

medicalai.chief.nnets.resnet.resnet20(*, num_blocks=3, img_input=None, classes=10, training=None)

Instantiates the ResNet architecture.

• Parameters

  • num_blocks – integer, the number of conv/identity blocks in each block. The ResNet contains 3 blocks with each block containing one conv block followed by (layers_per_block - 1) number of idenity blocks. Each conv/idenity block has 2 convolutional layers. With the input convolutional layer and the pooling layer towards the end, this brings the total size of the network to (6*num_blocks + 2)

  • classes – optional number of classes to classify images into

  • training – Only used if training keras model with Estimator. In other

  • it is handled automatically. (scenarios) –

• Returns

  A Keras model instance.

medicalai.chief.nnets.resnet.resnet32(*, num_blocks=5, img_input=None, classes=10, training=None)

Instantiates the ResNet architecture.

• Parameters

  • num_blocks – integer, the number of conv/identity blocks in each block. The ResNet contains 3 blocks with each block containing one conv block followed by (layers_per_block - 1) number of idenity blocks. Each conv/idenity block has 2 convolutional layers. With the input convolutional layer and the pooling layer towards the end, this brings the total size of the network to (6*num_blocks + 2)

  • classes – optional number of classes to classify images into

  • training – Only used if training keras model with Estimator. In other

  • it is handled automatically. (scenarios) –

• Returns

  A Keras model instance.

medicalai.chief.nnets.resnet.resnet56(*, num_blocks=9, img_input=None, classes=10, training=None)

Instantiates the ResNet architecture.

• Parameters

  • num_blocks – integer, the number of conv/identity blocks in each block. The ResNet contains 3 blocks with each block containing one conv block followed by (layers_per_block - 1) number of idenity blocks. Each conv/idenity block has 2 convolutional layers. With the input convolutional layer and the pooling layer towards the end, this brings the total size of the network to (6*num_blocks + 2)

  • classes – optional number of classes to classify images into

  • training – Only used if training keras model with Estimator. In other

  • it is handled automatically. (scenarios) –

• Returns

  A Keras model instance.

medicalai.chief.nnets.resnet.resnet_block(input_tensor, size, kernel_size, filters, stage, conv_strides=(2, 2), training=None)

A block which applies conv followed by multiple identity blocks.

• Parameters

  • input_tensor – input tensor

  • size – integer, number of constituent conv/identity building blocks.

  • conv block is applied once**, ****followed by** (A) –

  • kernel_size – default 3, the kernel size of middle conv layer at main path

  • filters – list of integers, the filters of 3 conv layer at main path

  • stage – integer, current stage label, used for generating layer names

  • conv_strides – Strides for the first conv layer in the block.

  • training – Only used if training keras model with Estimator. In other scenarios it is handled automatically.

• Returns

  Output tensor after applying conv and identity blocks.

medicalai.chief.nnets.vgg16 module

medicalai.chief.nnets.vgg16.VGG16_Model(img_input=(224, 224, 3), classes=3)

Loaded the VGG16 network, ensuring the head FC layer sets are left off

• Parameters

  • img_input – optional shape tuple, only to be specified if include_top is False (otherwise the input shape has to be (224, 224, 3) (with ‘channels_last’ data format) or (3, 224, 224) (with ‘channels_first’ data format). It should have exactly 3 inputs channels, and width and height should be no smaller than 32. E.g. (200, 200, 3) would be one valid value.

  • classes – Number of classes to be predicted.

  • Returns – model

medicalai.chief.nnets.xception module

medicalai.chief.nnets.xception.Xception(img_input=(224, 224, 3), classes=3)

Loaded the Xception network, ensuring the head FC layer sets are left off

• Parameters

  • img_input – optional shape tuple, only to be specified if include_top is False (otherwise the input shape has to be (224, 224, 3) (with ‘channels_last’ data format) or (3, 224, 224) (with ‘channels_first’ data format). It should have exactly 3 inputs channels, and width and height should be no smaller than 32. E.g. (200, 200, 3) would be one valid value.

  • classes – Number of classes to be predicted.

  • Returns – model

Module contents