System and method for analyzing medical images to detect and classify a medical condition using machine-learning and a case pertinent radiology atlas
A system for analyzing medical images to detect and classify a medical condition, the system includes an input for receiving a medical image, a convolutional neural network coupled to the input and configured to analyze the medical image to generate a prediction including a probability of the presence of the medical condition in the medical image and an atlas creation module coupled to the convolutional neural network and configured to generate an atlas comprising a set of image features and a set of training images. Each image feature is assigned with at least one training image associated with the medical condition. The system further includes a prediction basis selection module coupled to the convolutional neural network and the atlas and configured to create a prediction basis for the prediction generated by the convolutional neural network.
1. A system for analyzing medical images to detect and classify a medical condition, the system comprising:
an input for receiving a medical image;
a convolutional neural network coupled to the input and configured to analyze the medical image to generate a prediction including a probability of the presence of the medical condition in the medical image;
an atlas creation module coupled to the convolutional neural network and configured to generate an atlas comprising a set of image features and a set of training images, wherein each image feature is assigned with at least one training image associated with the medical condition;
a prediction basis selection module coupled to the convolutional neural network and the atlas, the prediction basis selection module configured to create a prediction basis configured to provide an explanation for the prediction generated by the convolutional neural network, the prediction basis created by selecting at least one image feature and at least one training image assigned to the selected image feature from the atlas; and
a display coupled to the convolutional neural network and the prediction basis selection module, the display configured to display at least the prediction and the prediction basis.
2. The system according to claim 1 , wherein the medical condition has a plurality of subtypes.
3. The system according to claim 2 , wherein the convolutional neural network is configured to generate a prediction indicating a probability of the presence of each of the plurality of subtypes of the medical condition and each image feature in the atlas is assigned with at least one training image associated with at least one of the plurality of subtypes of the medical condition.
4. The system according to claim 2 , wherein the medical condition is intracranial hemorrhage (ICH) and the plurality of subtypes includes intraparenchymal hemorrhage (IPH), intraventricular hemorrhage (IVH), subdural hemorrhage (SDH), epidural hemorrhage (EDH) and subarachnoid hemorrhage (SAH).
5. The system according to claim 1 , wherein the convolutional neural network is a deep convolutional neural network.
6. The system according to claim 1 , wherein the convolutional neural network comprises a plurality of deep convolutional neural networks.
7. The system according to claim 1 , wherein the medical condition is intracranial hemorrhage (ICH).
8. The system according to claim 1 , further comprising a preprocessing module coupled to the input and the convolutional neural network, the preprocessing module configured to perform a multi-window conversion on the medical image received at the input.
9. The system according to claim 1 , further comprising a preprocessing module coupled to the input and the convolutional neural network, the preprocessing module configured to perform slice interpolation.
10. The method according to claim 1 , wherein the convolutional neural network comprises a plurality of deep convolutional neural networks.
11. The method according to claim 1 , wherein the medical condition is intracranial hemorrhage (ICH).
12. A method for analyzing medical images to detect and classify a medical condition, the method comprising:
receiving a medical image;
analyzing the medical image using a convolutional neural network to generate a prediction including a probability of the presence of the medical condition in the medical image;
creating a prediction basis configured to provide an explanation for the prediction generated by the convolutional neural network by selecting at least one image feature from an atlas comprising a set of image features and a set of training images, wherein each image feature is assigned with at least one training image associated with the medical condition; and
displaying at least the prediction and the prediction basis on a display.
13. The method according to claim 12 , wherein the medical condition has a plurality of subtypes.
14. The method according to claim 13 , wherein the convolutional neural network is configured to generate a prediction indicating a probability of the presence of each of the plurality of subtypes of the medical condition and each image feature in the atlas is assigned with at least one training image associated with at least one of the plurality of subtypes of the medical condition.
15. The method according to claim 13 , wherein the medical condition is intracranial hemorrhage (ICH) and the plurality of subtypes includes intraparenchymal hemorrhage (IPH), intraventricular hemorrhage (IVH), subdural hemorrhage (SDH), epidural hemorrhage (EDH) and subarachnoid hemorrhage (SAH).
16. The method according to claim 12 , wherein the convolutional neural network is a deep convolutional neural network.
17. The method according to claim 12 , further comprising a performing a multi-window conversion on the medical image received at the input.
18. The system according to claim 12 , further comprising performing slice interpolation.