Image feature classification
A method and system for image feature classification using a NN-based learning algorithms to make a decision about a feature in a medical image or image part. In particular, embodiments may make use of a phase of a multi-phasic image to improve classification accuracy. For instance, embodiments may combine different phases of multiphasic images as training data.
1 . A computer-implemented method for classifying image features of multi-phasic medical images, the computer-implemented method comprising:
obtaining a medical image and a phase identifier associated with the medical image, the phase identifier representing a phase of the medical image;
combining the medical image and the phase identifier to generate modified image data;
providing the modified image data as an input of a layer of a Neural Network (NN)-based learning algorithm configured to classify image features; and
obtaining, for the modified image data, an image feature classification result from the NN-based learning algorithm;
wherein combining the medical image and the phase identifier comprises:
processing the medical image with an image feature generation algorithm to generate a tensor representing one or more features of the medical image; and
combining the phase identifier with the tensor to generate, as the modified image data, a modified tensor, and
wherein the phase identifier comprises a numerical value.
2 . The computer-implemented method of claim 1 , wherein the medical image comprises an image of at least a portion of a liver having a lesion, and wherein the NN-based learning algorithm comprises a Liver Imaging Reporting and Data System feature classifier trained to make a decision about a classification of an image feature.
3 . The computer-implemented method of claim 1 , wherein the NN-based learning algorithm is trained using a training algorithm configured to receive an array of training inputs and known outputs, wherein the training inputs comprise medical image, medical image segments or medical image features, and wherein the known outputs comprise image feature classifications.
4 . The computer-implemented method of claim 1 , wherein providing the modified image data to the layer of the NN-based learning algorithm comprises:
providing the modified tensor to a hidden layer trained to make a decision about a classification of an image feature.
5 . The computer-implemented method of claim 4 , wherein the hidden layer of the NN-based learning algorithm includes a convolution layer or a fully connection layer.
6 . The computer-implemented method of claim 5 , wherein modifying data of the medical image comprises concatenating the phase identifier to the medical image as an extract channel.
7 . The computer-implemented method of claim 5 , wherein modifying data of the medical image comprises at least one of:
adding the numerical value of the phase identifier to pixel values of the medical image; and
multiplying pixel values of the medical image by the numerical value of the phase identifier.
8 . The computer-implemented method of claim 1 , wherein combining the medical image and the phase identifier comprises:
modifying data of the medical image based on the phase identifier to generate the modified image data.
9 . The computer-implemented method of claim 1 , wherein obtaining, for the modified image data, the image feature classification result from the NN-based learning algorithm comprises:
obtaining a plurality of image feature classification results from the NN-based learning algorithm; and
switching between the plurality of image feature classification results based on the phase identifier so as to select the image feature classification result.
10 . A non-transitory computer readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to:
obtain a medical image and a phase identifier associated with the medical image, the phase identifier representing a phase of the medical image;
combine the medical image and the phase identifier to generate modified image data;
provide the modified image data as an input of a layer of a Neural Network (NN)-based learning algorithm configured to classify image features; and
obtain, for the modified image data, an image feature classification result from the NN-based learning algorithm;
wherein, to combine the medical image and the phase identifier, the instructions, when executed by the one or more processors, further cause the one or more processors to:
process the medical image with an image feature generation algorithm to generate a tensor representing one or more features of the medical image; and
combine the phase identifier with the tensor to generate, as the modified image data, a modified tensor, and
wherein the phase identifier comprises a numerical value.
11 . The non-transitory computer readable medium of claim 10 , wherein the NN-based learning algorithm is trained using a training algorithm configured to receive an array of training inputs and known outputs, wherein the training inputs comprise medical image, medical image segments or medical image features, and wherein the known outputs comprise image feature classifications.
12 . The non-transitory computer readable medium of claim 10 , wherein providing the modified image data to a layer of the NN-based learning algorithm comprises:
providing the modified tensor to a hidden layer of the NN-based learning algorithm trained to make a decision about a classification of an image feature.
13 . The non-transitory computer readable medium of claim 10 , wherein combining the medical image and the phase identifier comprises:
modifying data of the medical image based on the phase identifier to generate the modified image data.
14 . The non-transitory computer readable medium of claim 10 , wherein modifying data of the medical image comprises at least one of:
adding the numerical value of the phase identifier to pixel values of the medical image; and
multiplying pixel values of the medical image by the numerical value of the phase identifier.
15 . The non-transitory computer readable medium of claim 10 , wherein obtaining, for the modified image data, the image feature classification result from the NN-based learning algorithm comprises:
obtaining a plurality of image feature classification results from the NN-based learning algorithm; and
switching between the plurality of image feature classification results based on the phase identifier so as to select the image feature classification result.
16 . A system for classifying image features of multi-phasic medical images, the system comprising:
a processor configured to:
obtain a medical image and a phase identifier associated with the medical image, the phase identifier representing a phase of the image,
combine the medical image and the phase identifier to generate modified image data,
provide the modified image data as an input of a layer of a Neural Network (NN)-based learning algorithm configured to classify image features, and
obtain, for the modified image data, an image feature classification result from the NN-based learning algorithm,
wherein, to combine the medical image and the phase identifier, the processor is further configured to:
process the medical image with an image feature generation algorithm to generate a tensor representing one or more features of the medical image; and
combine the phase identifier with the tensor to generate, as the modified image tensor, a modified tensor, and
wherein the phase identifier comprises a numerical value.
17 . The system of claim 16 , wherein the NN-based learning algorithm is trained using a training algorithm configured to receive an array of training inputs and known outputs, wherein the training inputs comprise medical image, medical image segments or medical image features, and wherein the known outputs comprise image feature classifications.
18 . The system of claim 16 , wherein providing the modified image data to a layer of the NN-based learning algorithm comprises:
providing the modified tensor to a hidden layer of the NN-based learning algorithm trained to make a decision about a classification of an image feature.
19 . The system of claim 16 , wherein combining the medical image and the phase identifier comprises:
modifying data of the medical image based on the phase identifier to generate the modified image data.
20 . The system of claim 16 , further comprising:
a user interface configured to receive, from the processor, and display the image feature classification result.