ARchy everyone! In the last article, we explored the connection between AR and AI, and their areas of application. One of the areas we looked into was medicine, but we did not talk about it in much detail. In this article, we will have a closer look at the role of AR and AI in medical imaging. We will also talk about VR and its application in medicine.
VR in medical imaging:
The first thing that comes to mind when we think about VR in medical imaging is the training of students, for example, surgery simulation.
This is an obvious implementation of VR in medicine. However, it can go far beyond that. This technology allows even the most experienced surgeons to hone their skills or compensate for the lack of experience during surgery. Wait, that’s not all! Who would have thought that VR helps patients with neurological disorders, such as senile dementia? A great example would be a program called Virtual Relief. Through virtual reality patients with dementia are encouraged to use their cognitive functions in a safe and relaxed environment, thus practicing their planning, organizational and prioritization skills.
Another good example is a Swiss startup MindMaze that helps people restore
their motor skills after a stroke. They use virtual reality to map the patient’s movements onto 3D avatars, thus reactivating damaged neural pathways and
activating new ones.
Virtual reality can also help to treat severe paranoia. Oxford University study found that allowing patients to face their fears in virtual reality leads to significant reductions in anxiety. The study claims that patients are more willing to try out new psychological techniques knowing that the situation is unreal, thus reducing the effects of paranoia in real life.
VR is also used before surgery or other medical procedures. A patient puts on VR glasses with calming content, and this has proven to help relieve stress, fear or anxiety he might feel.
As you can see, VR is already in demand, and its role in medical imaging is
AR in medical imaging:
What about augmented reality? In some ways, its functionality might be
similar to virtual reality (surgeries, training of students and doctors).
Let’s have a look at three companies that have achieved impressive
results in AR:
1. Brain Power
2. Medsights Tech
Brain Power is a Massachusetts-based company founded in 2013. It focuses on the practical application of neurobiology achievements using the latest wearable technologies, such as Google Glass. The company creates software that converts wearable devices into neuro-assistive devices. These devices are used for teaching life skills to children and adults with autism. The suite of applications called “Empowered Brain” is designed to help children with their social skills, language, and positive behavior. The software contains powerful data collection and analysis tools that allow you to customize feedback for each person.
Medsights Tech is a company that is working to make x-ray vision real. It developed software to test the feasibility of using AR to create accurate three- dimensional tumor reconstructions. Sophisticated image reconstruction technology basically gives surgeons x-ray vision — without any radiation and in real time. It has been designed to be intuitive for practitioners of various fields, including technicians, surgeons, and other medical professionals. It has also been tested for multiple body parts, such as skin, head, neck and gastrointestinal tract.
Some people are afraid of blood tests. They fear that the nurse will not find the vein on the first try and drawing a blood sample will be a painful procedure. This is where AccuVein comes to the rescue. It uses AR in a portable scanner that projects an image onto the skin and shows nurses and doctors where the veins are in the patients’ body.
This is just a tiny part of where, how and by whom AR is used in medical
Leave claps in the comments if you like this article! More detailed review on AR and VR in medicine is going to be out soon.
AI in medical imaging:
As we already know, computer vision runs the show. This time we will
change our vector to a convolutional neural network for biomedical image
segmentation — U-net.
U-net not only defines the class of the whole image, but it also segments its
areas by class. It creates a mask that divides the image into several classes.
The network is trained end-to-end on a small number of images. As a
result, it works very quickly. Segmentation of a 512×512 image takes less than a second on a modern GPU.
The stochastic gradient boosting method is usually used to train the network. Many real pictures and masks corresponding to each picture are used for training. The mask indicates which class the corresponding pixel in the image belongs to. For example, if we have two classes, a tumor, and background, it is natural to mark the corresponding tumor pixels on the mask as “1” and the background as “0”. The image coming to the input of the neural network is firstly compressed and then expanded to the size of the mask.
An example of how this technology is used in the segmentation of blood
vessels in the retina.
Now you understand what part AI, AR, VR play in medicine. These technologies make life easier for both patients and doctors and open up new opportunities for people suffering from serious diseases.