Computer vision in medicine
Горбина М. А., Левина О. С. Computer vision in medicine // Молодой ученый. 2014. №9. С. 131-133. URL https://moluch.ru/archive/68/11724/ (дата обращения: 22.01.2018).
Electronics is a field of engineering and applied physics dealing with the design and application of electronic circuits. The operation of circuits depends on the flow of electrons for generation, transmission, reception and storage of information.
Today it is difficult to imagine our life without electronics. It surrounds us everywhere. Electronic devices are widely used in scientific research and industrial designing; they control the work of plants and power stations, calculate the trajectories of space ships and help people to discover new phenomena of nature. Automatization of production processes and studies on living organisms became possible due to electronics .
A lot of scientific discoveries are occurring each year. Most of them change our lives. Big steps forward were made in medicine, too. Health centers are able to treat a lot of diseases. There are a lot of qualified experts in our country. High technologies help to make a diagnosis. Medical technologies have become a central part of human experiences including birth, family relations, work life, ageing and death, no matter if one is ill, disabled or healthy.
At first, find out what computer vision is in general. Computer vision is a relatively new field of medicine. Computer Vision is a science that includes methods for acquiring, processing, analyzing and understanding images and, in general, high-dimensional data from the real world in order to produce numerical or symbolic information, e.g., in the forms of decisions. A subject matter in the development of this field has been to copy the abilities of human vision by electronically perceiving and understanding an image.
Computer Vision plays a very important role in the field of medicine. This area is characterized by the extraction of information from image data for the purpose of making a medical diagnosis of a patient. Generally, image data is in the form of microscopy images, X-ray images, ultrasonic images, and tomography images. An model of information which can be extracted from such image data is detection of tumors, arteriosclerosis or other malign changes. It can also be measurements of organ dimensions, blood flow, etc. This application area also supports medical research by providing new information, e.g., about the structure of the brain, or about the quality of medical treatments .
There are various motivations for using digital image processing methods in medicine:
- new modalities and multimodal analysis: foremost comes certainly the possibility of exploring new imaging modalities, leading to new anatomical or functional insights;
- further, image analysis will support the combined evaluation of data from different modalities;
- morphometry: the use of computerized techniques allows better precision and repeatability, with, as a consequence, improved objectivity of measurement of morphometric parameters like size, area, volume, circumference, etc.;
- improved interpretation: the sensitiveness of those new imaging modalities coupled with the power of recent visualization techniques enable more refined diagnosis than with using conventional exploratory methods;
- more accurate prediction: a consequence is the ability of providing more finely tuned medical treatment, for example lower doses in radiation therapy or more accurate positioning in head surgery;
- process automation: many medical operations can benefit from the reliability provided by automatic processing, from the screening of biological specimen to vision guided surgery;
- understanding of volume data: recognition of structures from volume data is not a spontaneous visual task and will benefit from computerized processing and visualization.
Using computers was one of the most important technological changes in the 20th century medicine. They took a central place in medical care from the 1950s. Computerized machines in hospitals monitored patients continuously. Imaging techniques such as MRI or PET were possible because faster computers could reconstruct images of the body. More diagnostic tests were developed because automated laboratory equipment performed tests quicker and more precisely.
MRI, or magnetic resonance imaging, uses strong magnetic fields to change the spin of atoms in our bodies. Radio signals detect these tiny changes. MRI computers process this information and construct images of soft tissues inside the body, from the brain to blood vessels. The advantage of MRI is its virtually harmless to the patient because, unlike CT scanners, MRI does not generate radiation .
The spinning atom effect is known as nuclear magnetic resonance (NMR). It was first observed during the late 1930s, but medical applications were not found for the NMR technique until the 1970s. It was renamed MRI because ‘nuclear’ was off-putting for patients. Over 60 million MRI examinations are now carried out annually .
PET, or positron emission tomography, is an imaging technology. Unlike CT or MRI, PET does not image structures inside the body. Instead, PET machines produce images of body functions in 3D. These include blood flow and parts of the brain responsible for specific mental processes. Patients swallow or are injected with safe radioactive materials called ‘tracers’, and are then placed inside the PET machine while detectors scan and track the tracers around the body. Computers process this information into images showing body function.
However, PET is not used in medicine as often as other scanning technique. PET technique is complex and expensive. To use a PET device, hospitals need enormous machines called cyclotrons to produce radioactive tracers on site.
And some words about SPECT. SPECT — Single Photon Emission Computed Tomography is a 3D nuclear medicine imaging technique that provides qualitative and quantitative physiologic measurements. As a complement to current practices, SPECT or CT is beneficial for earlier stage testing of pharmacokinetics, pharmacodynamics, disease and tissue targeting as well as later stage regulatory, biodistribution and drug metabolism studies .
Key advantages over pre-existing methods include:
- non-invasive imaging of fine kinetics can be examined allowing establishment of a kinetic curve by tissue or even by tissue compartment (heterogeneity);
- 3-dimensionality with tissue localization;
- clinical translatability;
- simultaneous response to questions in different regions of a biologic and, therefore, address different questions simultaneously and more cost-effectively;
- tracer multiplexing — multiple isotopes of different energies can be used to address multiple simultaneous questions.
Another area of computer vision is a virtual reality (VR). What is it and what does it do?
VR is being applied to a wide range of medical areas, including remote and local surgery, surgery planning, medical education and training, treatment of phobias and other causes of psychological distress, skill training, and pain reduction. It is also used for the visualization of large-scale medical records, and in the architectural planning of medical facilities .
VR for surgery involves applications of interactive immersive computer technologies to help perform, plan and simulate surgical procedures. In performance, the VR guides the surgeon, sometimes with a robot to execute the procedure under the surgeon's control. In other words, VR is used to give the surgeon 3D interactive views of areas within the patient. Planning is carried out preoperatively to find the best approach to surgery, involving minimum damage. Simulation is mostly used in training, using patient data often registered with anatomical information from an atlas. It may be used for routine training or to focus on particularly difficult cases and new surgical techniques .
VR can most immediately and successfully be applied today for physical and mental health and rehabilitation. This is partly because of the technical demands, particularly in terms of detailed image and interactivity. These systems often imitate the physical environment, a world of rooms, doors, buildings, etc., many of which are simple shapes and much easier to model that the irregular and contoured surfaces of internal organs.
Imaging has become an essential component in many fields of bio-medical research and clinical practice. Biologists learn cells and generate 3D microscopy data sets, virologists generate 3D reconstructions of viruses from micrographs, radiologists identify and quantify tumors from MRI and CT scans, and neuroscientists detect regional metabolic brain activity from PET and functional MRI scans. Analysis of these diverse types of images requires sophisticated computerized quantification and visualization tools.
An important goal of researchers in the field of computer vision is to create an computerized system that can equal or surpass the capacity of the human brain to recognize faces.
In this matter Computer Vision is developing very slowly. At first it was rather primitive. PC recognizes only the abstract form of the object, then the motion, etc. But research in this technology is very intense and now it is possible with the help of computer vision to recognize the facial details of a particular person among the many images of data. It finds application in many areas of human life. Computer vision was created by humans, and consequently it is tied closely to the research of human vision. Some medical specialists believe that human vision is similar to an extremely complex version of computer vision. Human vision, like computer vision, is bound by rules — a whole lot of them. It is used to mimic and simulate the behavior of biological optics. Now it has not been invented the perfect device the human eye which could be replaced by a blind person.
It is important for the future. But to replace the human eyes are very difficult. The progress in this area would make possible blind people to see and this is new horizons in medicine and in people's lives.
1. Newcombe R., Izadi S., Hilliges O., Molyneaux D., Kim D., Davison A., Kohli P., Shotton J., Hodges S., and Fitzgibbon A.. Kinectfusion: Real-time dense surface mapping and tracking. In IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2011.
2. Matsuo T., Fukushima N., Ishibashi: Weighted Joint Bilateral Filter with Slope Depth Compensation Filter for Depth Map Refinement. 300–309, VISAPP 2013, International Conference on Computer Vision Theory and Application.
3. Parker R. C. 1 Kinect Depth Inpainting and Filtering URL: http://www.radfordparker.com/projects.html.
4. Радовель, В. А. Английский язык. Основы компьютерной грамотности. / В. А. Радовель. — Ростов-на-Дону: Феникс, 2009. — 224 с.