System and method for detecting recurrence of a disease
A method for determining a recurrence of a disease in a patient includes generating a medical image of an organ of the patient and then extracting an invasive edge around an area of interest in the medical image. A plurality of radiomics features is obtained from the invasive edge and the recurrence of the disease is determined based on the plurality of radiomics features.
1 . A method for determining a recurrence of a disease in a patient, the method comprising:
generating a medical image of an organ of the patient;
extracting an invasive edge around an area of interest in the medical image;
obtaining a plurality of radiomics features from the invasive edge; and
determining the recurrence of the disease based on the plurality of radiomics features,
wherein the disease defines a breast cancer,
wherein determining the recurrence of the disease comprises:
ranking the plurality of radiomics features based on their ability to predict the recurrence of the cancer using a machine learning classifier; and
providing high ranking radiomics features to a deep neural network for classifying the medical image as a disease recurrent medical image or a disease non-recurrent medical image, and
wherein the deep neural network is trained on radiomics features that have been selected to predict the recurrence of the disease based on a plurality of training data pairs comprising noisy and pristine medical images.
2 . The method of claim 1 , wherein the medical image comprises an X-ray image, a computed tomography (CT) image, a magnetic resonance image (MRI) or an ultrasound image.
3 . The method of claim 1 , wherein the area of interest includes a tumor region.
4 . The method of claim 1 , wherein extracting the invasive edge comprises segmenting the area of interest from the medical image.
5 . The method of claim 4 , wherein the invasive edge includes an edge surrounding the area of interest having a width.
6 . The method of claim 5 , wherein the width of the invasive edge is dependent on the nature of the disease and a resolution of the medical image.
7 . The method of claim 5 , wherein the width of the invasive edge is in the range of 0.5 mm to 1 cm.
8 . The method of claim 1 , wherein the plurality of radiomics features comprises shape, texture, intensity, and gradient magnitude of the invasive edge.
9 . The method of claim 1 , wherein the disease comprises triple negative breast cancer, receptor positive breast cancer, or HER-2 positive breast cancer.
10 . The method of claim 1 further comprising extracting an area in the tumor region along with the invasive edge around the tumor region and determining the recurrence of the disease based on the radiomics features in both the invasive edge and the tumor region area.
11 . A system comprising:
a memory storing a machine learning model including a classifier and a deep learning network;
a display device; and
a processor communicably coupled to the memory and configured to:
receive a medical image of an organ of the patient;
rank a plurality of radiomics features based on their ability to predict the recurrence of the cancer using the classifier;
providing high ranking radiomics features of an invasive edge around an area of interest in the medical image to the deep learning network;
classify the medical image using the deep neural network as a disease recurrent image or a disease non-recurrent image based on the high ranking radiomics features; and
display the disease recurrent image via the display device,
wherein the disease defines a breast cancer, and
wherein the deep neural network is trained on radiomics features that have been selected to predict the recurrence of the disease based on a plurality of training data pairs comprising noisy and pristine medical images.
12 . The system of claim 11 , wherein the processor is further configured to extract the invasive edge by segmenting the area of interest from the medical image.
13 . The system of claim 11 , wherein the invasive edge includes an edge surrounding the area of interest having a width.
14 . The system of claim 13 , wherein the width of the invasive edge is in the range of 0.5 mm to 1 cm.
15 . The system of claim 11 , wherein the plurality of radiomics features comprises shape, texture, intensity, and gradient magnitude of the invasive edge.
16 . The system of claim 11 , wherein the disease comprises triple negative breast cancer, receptor positive breast cancer, or HER-2 positive breast cancer.