Task interaction network for prostate cancer diagnosis

A method for training a task-interaction network (TI-Net) to jointly perform a segmentation task to locate prostate cancer areas and a classification task to assess aggressiveness of lesions based on a multi parametric-magnetic resonance imaging (mp-MRI) scan image in terms of a plurality of defined categories, the method comprising:

• preparing a training dataset of mp-MRI images, each mp-MRI image having a labelled category and including a first type MRI slice and a second type MRI slice;
• for each mp-MRI image:
• registering the first type MRI slice with the second type MRI slice via non-rigid registration based on mutual information of the first and second type MRI slices;
• normalizing intensity of the registered first and second type MRI slices with zero mean and unit variance;
• center-cropping and resizing the normalized first and second type MRI slices in an axial plane to reduce noisy from irrelevant information;
• spatially aligning and concatenating the cropped and resized first and second type MRI slices;
• feeding the aligned and concatenated first and second type MRI slices to the TI-Net;
• training a backbone network in the TI-Net to extract an initial discriminative feature representation from the aligned and concatenated first and second type MRI slices;
• training an auxiliary segmentation branch in the TI-Net to generate an initial probability map and predict an initial lesion mask based on the initial discriminative feature representation;
• training a classification branch in the TI-Net to determine a plurality of category prototypes corresponding to the plurality of defined categories respectively and predict a lesion aggressiveness based on the initial discriminative feature representation, the plurality of category prototypes and the initial lesion mask;
• and
• training a main segmentation branch in the TI-Net to predict a lesion location based on the initial discriminative feature representation, the plurality of category prototypes and the predicted lesion aggressiveness.

A method using a task-interaction network (TI-Net) for jointly performing a segmentation task to locate prostate cancer areas and a classification task to assess aggressiveness of lesions based on a multi parametric-magnetic resonance imaging (mp-MRI) scan image in terms of a plurality of defined categories, the TI-Net including a backbone network, an auxiliary segmentation branch, a classification branch and a main segmentation branch, the method comprising:

• extracting, by the backbone network, an initial discriminative feature representation from an aligned and concatenated first and second type MRI slices of the mp-MRI scan image;
• generating, by a probability mapping module in the auxiliary segmentation branch, an initial probability map based on the initial discriminative feature representation;
• performing, by an auxiliary segmentation module in the auxiliary segmentation branch, a softmax operation on the initial probability map to obtain an initial lesion mask;
• generating, by a lesion awareness module in the classification branch, a refined discriminative feature representation based on the initial lesion mask and the initial discriminative feature representation;
• determining, by a prototyping module in the classification branch, a plurality of category prototypes corresponding to the plurality of defined categories respectively;
• predicting, by a classification module in the classification branch, a lesion aggressiveness based on the refined discriminative feature representation and the plurality of category prototypes;
• generating, by a category allocation module in the main segmentation branch, a hybrid feature representation based on the initial discriminative feature representation, the plurality of category prototypes and the predicted lesion aggressiveness;
• and
• predicting, by a main segmentation module in the main segmentation branch, a lesion location based on the hybrid feature representation.