Professor Ingfried Halldorsen speaks enthusiastically and passionately about the project she and her research team at the Mohn Center for Medical Visualization (mmivShe has been working with him for five years.
At Haukeland Hospital, a highly qualified team consisting of mathematicians, developers and other health workers has opened up new frontiers for the investigation of brain diseases.
Artificial intelligence is also used to gain a new understanding of diseases.
“Our original mission was to use artificial intelligence to analyze gynecological cancer image data,” says Halldorsen.
machine interpretation
Scan: A radiologist takes MRI images of the body, where a cancerous tumor has been found in the uterus. Photo: Ole Enes Ebbesen/TV 2
Researchers at MMIV have already made significant progress using artificial intelligence (AI) to find and map round carcinoid tumors in the human body.
With the advances of artificial intelligence in the past year (and months) — also demonstrated by services like ChatGPT — researchers believe that AI can provide increased knowledge about brain diseases, and contribute in the long term to their prevention and treatment.
More specifically, the big disease mysteries Alzheimer’s diseaseAnd ALSAnd multiple sclerosis (MS) And Parkinson’s Disease.
The use of advanced radiographic imaging techniques allows for a detailed representation of the brain and its functions. It has been shown that small changes in the brain can say something about a patient’s risk of developing brain diseases in the future, such as dementia, says Haldorsen.
AI tools could eventually be used to contribute to the interpretation of the images, and the data could be compared to other patient data of interest to disease progression.
In this way, AI can detect small changes in the brain that a doctor may not be able to detect with the naked eye.
“brain age”
Disease detection: Brain researchers are getting increasingly better tools for examining different parts of the body — including the brain. Photo: Penelope Alida Larsen / TV 2
New studies have shown that you can estimate the “age of a person’s brain,” Haldorsen says, based on advanced radiographs.
In patients with brain diseases, it will be possible to see in the images that their brain is aging faster than in healthy patients.
According to her, AI tools will eventually be able to help estimate brain age based on brain images.
– Maybe a little scary idea, says the professor. Explaining how this information can be used to ensure the best possible treatment for patients is yet to be clarified, and more knowledge is needed on this.
– It’s 50/50
Savings: – I think that artificial intelligence will be very important, and it will lead to spending less resources on sometimes boring tasks, says Professor Ingfried Halldorsen. Photo: Ole Enes Ebbesen/TV 2
She also believes that interpretation with the help of artificial intelligence for example in the long term, MRI images will be able to help with neurodegenerative diseases such as dementia, amyotrophic lateral sclerosis and Parkinson’s disease.
– There are complex links between the results of the images and the pathological condition of the patient, but the images will be able to be analyzed and compared in large data banks, and with image data from collaborating institutions. There are a lot of people working on this right now, Haldorsen says.
One of them is senior researcher Frank Reimer at the Mohne Medical Imaging Center (MMIV).
He is a physicist, and has worked for a number of years with cancer and neurological diseases both at the University of Cambridge and now at the MMIV Center alongside Haldorsen and Neuro-SysMed in Bergen.
– Do you think artificial intelligence can help us solve mysteries surrounding these diseases in the future?
Bergen: Frank Reimer is a brain specialist, researching the use of artificial intelligence to solve mysteries surrounding brain diseases such as Alzheimer’s, amyotrophic lateral sclerosis, multiple sclerosis, brain cancer, and Parkinson’s. Photo: Penelope Alida Larsen / TV 2
– I think it will help us have more time to do other things. Right now, AI probably can’t help us treat MS and Parkinson’s disease, but it can help people get a faster diagnosis.
– This will make it easier for doctors to follow patients, according to Reimer.
disease mysteries
Even the most experienced researchers still do not know why the aforementioned brain and nerve diseases occur in the body.
It is common in disorders such as ALS and MS that nerves are damaged, and patients’ muscle tone is impaired.
Like Alzheimer’s and Parkinson’s, one can observe what happens to the body and brain as the disease progresses, but why it occurs is partly a mystery.
– They searched for a long time, spent a lot of money, analyzed genetic data. However, you still know a lot of what you don’t. For cancer, you can see some characteristics, such as the fact that it can appear due to cell damage.
Moderator: Frank Rimmer says he has students trying to improve scanned images, comparing algorithms from old and new machine learning. Photo: Penelope Alida Larsen / TV 2
– Reimer says that neurological diseases are something completely different.
This is how algorithms help
Although AI is still at an early stage in the healthcare system, Rimmer says they’re already getting good help with specific tasks.
– Scanning machines have built-in algorithms from suppliers that can scale up and improve images, while scanning takes less time. Rimmer says they also remove noise from people.
After the scans, the MMIV Center uses its own algorithms to interpret and read the images, so that, for example, spots on the brain can be detected.
This is one of the signs of disease activity in MS.
– You have to be careful
Learning: Brain specialist Frank Rimmer instructs several students in the use of artificial intelligence in the healthcare system. Photo: Penelope Alida Larsen / TV 2
The first self-developed software was created four years ago, and in line with the developments of artificial intelligence, the algorithms have also become much better and the accuracy of measurements has increased.
– What benefits can this bring?
– Hospitals save money and resources, while patients get a faster diagnosis and a better quality of life. Rimmer says it will be easier for doctors to keep track of patients.
At the same time, the researcher was clear that one must have a cautious expectation of what artificial intelligence can achieve today.
IMAGE: A researcher at the MMIV Center shows a scanned image of a person’s spine. Now they use artificial intelligence to find wrongdoing. Photo: Ole Enes Ebbesen/TV 2
It is also clear that regular research is still the most important thing.
– We can trust technology, but not one hundred percent. A breakthrough in medical research may be just around the corner, but it may also take some time. We have to be part of the development in AI, but at the same time we have to be realistic, Rimmer says.
This is how artificial intelligence works in milliseconds
“Explorer. Unapologetic entrepreneur. Alcohol fanatic. Certified writer. Wannabe tv evangelist. Twitter fanatic. Student. Web scholar. Travel buff.”
