An image processing device, an image processing method, and an image processing program make it possible to accurately recognize property changes between medical images having different imaging times. A property classification unit classifies each pixel included in a target region of each of a first medical image and a second medical image of which an imaging time is later than an imaging time of the first medical image for the same subject into any one of a plurality of types of properties. A quantitative value derivation unit specifies property pairs each representing a property change between corresponding pixels in at least partial regions within the target regions, and derives a quantitative value in the at least partial regions for each of the property pairs.

A method for performing a computer-aided diagnosis (CAD) for universal lesion detection includes: receiving a medical image; processing the medical image to predict lesion proposals and generating cropped feature maps corresponding to the lesion proposals; for each lesion proposal, applying a plurality of lesion detection classifiers to generate a plurality of lesion detection scores, the plurality of lesion detection classifiers including a whole-body classifier and one or more organ-specific classifiers; for each lesion proposal, applying an organ-gating classifier to generate a plurality of weighting coefficients corresponding to the plurality of lesion detection classifiers; and for each lesion proposal, performing weight gating on the plurality of lesion detection scores with the plurality of weighting coefficients to generate a comprehensive lesion detection score.

A system for determining whether a dataset including a plurality of cross-sectional images includes a predetermined feature is provided, the system including a first AI model to: receive a dataset including a plurality of cross-sectional images as an input, analyse the dataset to identify a subset of cross-sectional images of the dataset capable of including the predetermined feature, and output the subset; a second AI model to: receive a first cross-sectional image from the subset, analyse the first cross-sectional image to determine whether the first cross-sectional image includes the predetermined feature, and output an indication of whether the first cross-sectional image includes the predetermined feature; and a processor configured to: provide the dataset as an input to the first AI model, obtain the output subset from the first AI model, provide the first cross-sectional image from the subset as an input to the second AI model, obtain the output from the second AI model, and based on the output from the second AI model indicating that the first cross-sectional image includes the pre-determined feature, determine that the dataset includes the predetermined feature.

Apparatus, systems, and methods to improve imaging quality control, image processing, identification of findings, and generation of notification at or near a point of care are disclosed and described. An example imaging apparatus includes a processor to at least: process the first image data using a trained learning network to generate a first analysis of the first image data; identify a clinical finding in the first image data based on the first analysis; compare the first analysis to a second analysis, the second analysis generated from second image data obtained in a second image acquisition; and, when comparing identifies a change between the first analysis and the second analysis, generate a notification at the imaging apparatus regarding the clinical finding to trigger a responsive action.

Disclosed are image recognition Artificial Intelligence (AI) training methods for multiple pulsed X-ray source-in-motion tomosynthesis imaging system. Image recognition AI training can be performed three ways: first, using existing acquired chest CT data set with known nodules to generate synthetic tomosynthesis Images, no X-ray radiation applied; second, taking X-ray raw images with anthropomorphic chest phantoms with simulated lung nodules, applying X-ray beam on phantom only; third, acquiring X-ray images using multiple pulsed source-in-motion tomosynthesis images from real patients with real known nodules and without nodules. An X-ray image recognition training network that is configured to receive X-ray training images, automatically determine whether the received images indicate a nodule or lesion condition. After training, image knowledge is updated and stored at knowledge database.

An X-ray imaging system using multiple pulsed X-ray source pairs in-motion to perform highly efficient and ultrafast 3D radiography is presented. The sources move simultaneously on arc trajectory at a constant speed as a group. Each individual source also moves rapidly around its static position in a small distance, but one moves in opposite direction to the other to cancel out linear momentum. Trajectory can also be arranged at a ring structure horizontally. In X-ray source pairs each moves in opposite angular direction to another to cancel out angular momentum. When an individual X-ray source has a speed that equals to group speed but an opposite linear or angular direction, the individual X-ray source is triggered through an external exposure control unit. This allows the source to stay relatively standstill during activation. 3D data can be acquired with wider view in shorter time and image analysis is real-time.

An X-ray imaging system using multiple pulsed X-ray sources to perform highly efficient and ultrafast 3D radiography is presented. There are multiple pulsed X-ray sources mounted on a structure in motion to form an array of sources. The multiple X-ray sources move simultaneously relative to an object on a pre-defined arc track at a constant speed as a group. Electron beam inside each individual X-ray tube is deflected by magnetic or electrical field to move focal spot a small distance. When focal spot of an X-ray tube beam has a speed that is equal to group speed but with opposite moving direction, the X-ray source and X-ray flat panel detector are activated through an external exposure control unit so that source tube stay momentarily standstill equivalently. 3D scan can cover much wider sweep angle in much shorter time and image analysis can also be done in real-time.

An X-ray imaging system using multiple pulsed X-ray sources in motion to perform high efficient and ultrafast 3D radiography using an X-ray flexible curved panel detector is presented. There are multiple pulsed X-ray sources mounted on a structure in motion to form an array of sources. The sources move simultaneously relative to an object on a predefined arc track at a constant speed as a group. Each individual X-ray source can move around its static position at a small distance. When an individual source has a speed equal to group speed, but with opposite moving direction, the individual source and detector are activated. This allows source to stay relatively standstill during activation. The operation results in reduced source travel distance for each individual source. 3D radiography image data can be acquired with much wider sweep angle in much shorter time, and image analysis can also be done in real-time.

A transport system with curved track pair is constructed for multiple pulsed X-ray source-in-motion to perform fast digital tomosynthesis imaging. It includes a curved rigid track pair with predetermined curvature, a primary motor stage car loaded with X-ray sources and wheels loaded with tension or compression springs. The car is driven by primary motor mounted at base frame and an engaged gear mounted at the car. The car can carry heavy loads, travel with high precision and high repeatability at all installation orientations while motion vibration is minimal. It is also scalable to have a larger radius. Track angle span usually can be from about ten degrees to about 170 degrees. During imaging acquisition, X-ray sources can sweep precisely from one location to another. The car has enough clearance to move in its path without rubbing wheels on tracks. Better than 0.2 mm overall spatial precision can be achieved with the digital tomosynthesis imaging.

A C-Arm X-ray imaging system using multiple pulsed X-ray sources in motion to perform efficient and ultrafast 3D radiography is presented. X-ray sources mounted on a structure in motion to form an array. X-ray sources move simultaneously relative to an object on a pre-defined arc track at a constant speed as a group. Each individual source can also move rapidly around its static position in a small distance. When a source has a speed that is equal to group speed but with opposite moving direction, the source at one C-arm end and X-ray flat panel detector at other C-arm end are activated through an external exposure control unit so that source stay momentarily standstill. The C-arm provides 3D X-ray scan imaging over a wide sweep angle and in different position by rotation. The X-ray image can be analyzed by an artificial intelligence module for real-time diagnosis.