Segmentation of bones for CT images, trauma cases
Automatic and semi automatic segmentation of CT trauma cases is a highly challenging topic. The thing is that abnormality of trauma cases doesn't allow to use a priori patterns (say, active shape model, ASM) and also the radiological response of tissues in trauma is changing. Сurrently we’re trying to combine different imaging techniques to extract bones and their fragments with acceptable accuracy. We’re experimenting now with a combination of segmentation techniques based on intensity thresholding, Gaussian mixture model, and edge extraction, as well as usage of morphological operations.
Method aimed at detecting patients’ body parts with no doctor involved
This method is about body parts’ detection and accurate automotive positioning of fractured bone models inside the patient with no doctor present. This method is being developed as a replacement of the existing gesture and voice interface management of 3D models in space. This is a combination of inner capabilities of HoloLens, Сomputer Vision technology and Procrustean Analysis. It gives us an opportunity to move the model towards the object rotating around its axis or moving in space.
3D reconstruction of the patient bone on the basis of a feq X-rays and the database of benchmark human bone models
Our system will make it possible to perform a 3D reconstruction of any patients’ bone based on a few X-rays and the database of benchmark human bone models relevant to a corresponding gender or age group.This technology can be efficiently implemented when working with easy bone fractures (no need for CT) , or when there’s no opportunity to perform CT (in developing countries).
AR in Cardiology is an indisputable component for successful development, improving the efficiency of performing cardiac surgeries. Cardiac procedures are planned on the basis of CT with contrast and 3D reconstruction of the coronary arteries. Then a doctor performs the procedure using angiography or 2D imaging. Next, the doctor memorizes the image in three dimensions and tries to mentally compare it with the 2D image that he or she sees.
AR offers a way to find blood vessels under the skin accurately and quickly by applying a 3D model of blood vessels to the patient during surgery. Our system can be used by surgeons on patients undergoing a reconstructive lower or upper limb surgery. With the HoloLens Headset the surgeon looks at the patient’s leg and sees the inside part of it. The surgeons sees not only the bones but blood vessels and can exactly identify where the targets are located.
Spinal needle injection procedures are used for anesthesia and analgesia, such as lumbar epidurals. These procedures require careful placement of a needle, both to ensure effective therapy delivery and to avoid damaging sensitive tissue such as the spinal cord. An important step in such procedures is the accurate identification of the vertebral levels.
Our AR system will be able to identify lumbar vertebral levels to assist in the process of spinal needle insertion for epidural anesthesia delivery.
Collecting of evidence-based data related to the use of the system in medical practice
As a result of close cooperation with practising surgeons, the system has built a patients’ case database continuing to expand and accumulate relevant evidence-based data. System’s algorithms performance is improved to to the very level when the system can become a part of the built-in software of a surgical robot.