Patient-specific deep learning image denoising methods and systems

A non-transitory computer-readable storage medium including instructions which, when executed, cause at least one processor to at least:

• process a first patient image of a first patient to add a first noise to the first patient image to form a noisy image input;
• thicken the first patient image from a thin slice image volume to a thick slice image volume to reduce noise in the first patient image before the first noise is to be added to the first patient image,
• train the first deep learning network using the noisy image input as input to identify the first noise;
• and
• when the first deep learning network is trained to identify the first noise, deploy the first deep learning network as a second deep learning network model to be applied to a second patient image of the first patient to identify a second noise in the second patient image, wherein the second noise identified by the second deep learning network model is to be removed from the second patient image to form a denoised patient image to be output.

A computer-implemented method of image denoising comprising:

• processing, using at least one processor, a first patient image of a first patient to add a first noise to the first patient image to form a noisy image input;
• thickening the first patient image from a thin slice image volume to a thick slice image volume to reduce noise in the first patient image before the first noise is to be added to the first patient image,
• training, using the at least one processor, the first deep learning network using the noisy image input as input to identify the first noise;
• and
• when the first deep learning network is trained to identify the first noise, deploying, using the at least one processor, the first deep learning network as a second deep learning network model to be applied to a second patient image of the first patient to identify a second noise in the second patient image, wherein the second noise identified by the second deep learning network model is to be removed from the second patient image to form a denoised patient image to be output.