An endoscope system includes an endoscope that acquires an image of a subject and a processor. The processor acquires position and posture information of the endoscope, estimates a three-dimensional shape of the subject based on the image and the position and posture information, generates, based on the position information, a cross section crossing a trajectory on which the endoscope passed, and determines an erroneous estimation portion based on a shape characteristic of the three-dimensional shape on the cross section.

An endoscope system includes an endoscope configured to pick up an image of an inside of a subject to acquire an image, and a processor including hardware. The processor estimates a three-dimensional position of a target portion within the subject based on an image, sets for the three-dimensional position a reliability corresponding to a predetermined parameter related to the endoscope system determined when the image is acquired, and selects, when a plurality of three-dimensional positions for the target portion exist, a predetermined three-dimensional position among the plurality of three-dimensional positions depending on the reliability and generates a three-dimensional shape image.

Systems and methods for processing electronic images from a medical device comprise receiving a first image frame and a second image frame from a medical device, and determining a region of interest by subtracting the first image frame from the second image frame, the region of interest corresponding to a visual obstruction in the first image frame and/or second image frame. Image processing may be applied to the first image frame and/or second image frame based on a comparison between a first area of the first image frame corresponding to the region of interest and a second area of the second image frame corresponding to the region of interest, and the first image frame and/or second image frame may be provided for display to a user.

Provided are systems and methods for analyzing sonograms of abdomen to identify presence or absence of free fluid in the abdomen. The systems and methods are useful for performing or assisting with point of care diagnosis of presence of internal bleeding in trauma patients without requiring input from a radiologist or another clinician trained to analyze sonograms for presence or absence of free fluid. Also provided are methods and system for training a medical imaging system for analyzing sonograms of abdomen to identify presence or absence of free fluid in the abdomen.

Systems and methods for processing electronic images from a medical device comprise receiving a first image frame and a second image frame from a medical device, and determining a region of interest by subtracting the first image frame from the second image frame, the region of interest corresponding to a visual obstruction in the first image frame and/or second image frame. Image processing may be applied to the first image frame and/or second image frame based on a comparison between a first area of the first image frame corresponding to the region of interest and a second area of the second image frame corresponding to the region of interest, and the first image frame and/or second image frame may be provided for display to a user.

Presented herein are systems and methods that provide for improved 3D segmentation of nuclear medicine images using an artificial intelligence-based deep learning approach. For example, in certain embodiments, the machine learning module receives both an anatomical image (e.g., a CT image) and a functional image (e.g., a PET or SPECT image) as input, and generates, as output, a segmentation mask that identifies one or more particular target tissue regions of interest. The two images are interpreted by the machine learning module as separate channels representative of the same volume. Following segmentation, additional analysis can be performed (e.g., hotspot detection/risk assessment within the identified region of interest).

Systems and methods related to combing multiple images are disclosed. An example method of combining multiple images obtained by an endoscope includes obtaining a first input image formed from a first plurality of pixels, wherein a first pixel of the plurality of pixels includes a first characteristic having a first value. The method also includes obtaining a second input image formed from a second plurality of pixels, wherein a second pixel of the second plurality of pixels includes a second characteristic having a second value. The method also includes subtracting the first value from the second value to generate a motion metric and generating a weighted metric map of the second input image using the motion metric.

A system for blob detection using deep learning is disclosed. The system may include a non-transitory computer-readable storage medium configured to store a plurality of instructions thereon which, when executed by a processor, cause the system to train a U-Net and generate a probability map including a plurality of centroids of a plurality of corresponding blobs, derive two distance maps with bounded probabilities, apply Difference of Gaussian (DoG) with an adaptive scale constrained by the two distance maps with the bounded probabilities, and apply Hessian analysis and perform a blob segmentation.

The present disclosure relates to stylizing a medical image. In accordance with certain embodiments, a method includes generating a medical image, segmenting the medical image into a first region and a second region, applying a first style to the first region and a different second style to the second region thereby generating a stylized medical image, and displaying the stylized medical image.

Systems and methods for tissue specific time gain compensation of an ultrasound image are provided. The method comprises acquiring an ultrasound image of a subject and displaying the ultrasound image over a console. The method further comprises selecting by a user a region within the ultrasound image that requires time gain compensation. The method further comprises carrying out time gain compensation of the user selected region of the ultrasound image. The method further comprises identifying a region having a similar texture to the user selected region and carrying out time gain compensation of the user selected region by an artificial intelligence (AI) based deep learning module.