| CPC G06T 7/0012 (2013.01) [A61B 5/7267 (2013.01); G06N 3/04 (2013.01); G16H 30/40 (2018.01); G16H 70/60 (2018.01); G06T 2207/20081 (2013.01); G06T 2207/30041 (2013.01)] | 18 Claims |
|
1. A computer-implemented method for computer aided diagnosis of a medical image, the method comprising:
processing the medical image through a machine learning (ML) model to provide a first feature representation of the medical image, wherein the ML model comprises an input layer, a plurality of parallel convolution layers, and a pooling layer, wherein the first feature representation is an output from the pooling layer, the plurality of parallel convolution layers are configured with different sized convolution filters in which results processed by the different sized convolution filters are concatenated, and a concatenated output of the plurality of parallel convolution layers is inputted into the pooling layer;
generating, by the ML model, a sequence of second feature representations of the medical image, wherein a first second feature representation in the sequence of the second feature representations is generated from the first feature representation through a layer and has a smaller dimension than the first feature representation, each subsequent second feature representation in the sequence of the second feature representations is generated from a respective prior second feature representation through a layer and has a lower dimension than the first respective prior second feature representation; and
generating, by the ML model, an output as a last second feature representation in the sequence of the second feature representations.
|