Doctors need to be skilled diagnosticians; however, future technologies are bound to aid them in determining the diseases their patients suffer from. Artificial intelligence offers vast possibilities within the medical field, promising to increase accuracy and decrease time spent on certain tasks. Lithuanian researchers have made a significant breakthrough. They trained an algorithm to examine fMRI brain scans and identify the early signs of Alzheimer’s Disease with an impressive accuracy of over 99%.
Alzheimer’s disease affects millions worldwide. A propensity for mental decline may show in brain scans well before external symptoms appear. Particularly, mild cognitive impairment (MCI) is the step proceeding Alzheimer’s (although not all MCI leads to Alzheimer’s). Essentially, inside the brain the material may already be changing and atrophying. Doctors can examine an fMRI manually to note this change, but humans are less efficient than machines. As it turns out, they are likely less accurate, too.
The researchers at Kaunas Universities in Lithuania created an algorithm to search for evidence of Alzheimer’s changes in brain scans. They used deep-learning techniques, which means the algorithm was trained, tested, and refined based on a data set of 78,000 scans. Researcher Rytis Maskeliūnas
Early flagging of potential Alzheimer’s disease offers patients more options. While there is not a technical cure for the disease, the more medicine understands about this heartrending prognosis, the more everyone can hope for better health outcomes.
A new algorithm uses artificial intelligence to identify the early signs of Alzheimer’s disease from fMRI brain scans.