Alzheimer’s, Healthcare and AI
A few years ago, I volunteered at a long term care facility near my home. I was interested in the healthcare field. I wanted to learn more about what it means to take care of someone around the clock at a nursing home facility. There I was, naive and full of energy. I was assigned to the Alzheimer’s floor as a visiting volunteer.
There’s a misconception that Alzheimer’s patients are not lucid.
In fact, the rate of disease progression in Alzheimer’s patients differs widely from patient to patient. Some patients can lead an active lifestyle in the lucid state for years before languishing between the lucid and non-lucid states.
In this nursing facility, on the same floor, I encountered patients who were in their lucid states: lively, energetic and full of life. I also encountered patients who were mostly catatonic and withdrawn into their own world.
For patients at the beginning of their Alzheimer’s journey, the facility offered activities such as board games, card games and sing along concerts.
For patients in the middle of their Alzheimer’s journey who were in and out of their withdrawn states, the facility offered activities such as flower arrangement, manicures, and taking care of stuffed animals.
My most vivid experience was with a patient who I will call Shelly. Shelly was an artist. She did not have children. She loved sitting by the window and bask in the sunshine of the day. When I visited her, she told me about her life as an artist, her fun dating history and why she never married. I visited with her for several weeks before she descended into her darkness. On my last visit with her, I showed her my mother’s artworks. She was grateful to share a moment of joint appreciation for art with me. We hugged when we said goodbye.
A week later, when I visited again, she didn’t recognize me. The nurses told me that she was now in this state almost 24 hours of the day.
I’m not related to Shelly. I only spent a few hours with her in the weeks I visited her.
But, her humanity touched me. My humanity gave her solace in a time that must’ve been extremely fearful for her.
In the U.S., there are 5 million people aged 65 years or older who have Alzheimer’s disease. According to the NIH, Alzheimer’s Disease is the most common cause of dementia among older adults. Dementia is the loss of cognitive functions. There are different types of dementia, each with its own causes: Lewy body dementia, frontotemporal disorder, and vascular dementia.
One of the biggest features of Alzheimer’s disease is that the disease can begin a decade or more before cognitive functions are impaired.
During this preclinical stage, toxic changes are already taking place:
- Abnormal deposits of proteins form amyloid plaques and tau tangles throughout the brain.
- Neurons lose connection with one another and stop functioning.
The damages initially take place in the hippocampus and entorhinal cortex where memories are formed. Then, as more neurons die, the disease affects other regions of the brain.
One of the biggest issues with Alzheimer’s disease is the quality of care. When diagnosed at a late stage, family members may not be able to cope with the rapidly rising responsibility of providing 24-hour care to an elderly loved one. Early diagnosis can potentially allow families to prepare for the financial and life changes that will occur while providing 24-hour care to the elderly loved one.
A better quality of life means that the patient will be able to have a dignified progression into his or her disease. Cost to society as well as to family members can be drastically reduced with a better quality of life.
In recent years, innovative treatments have enabled doctors to delay the progression of the disease in some patients. These treatments must be administered early to effective. Early diagnosis in the pre-clinical period is essential in this case.
Early detection of Alzheimer's disease presents a unique challenge for doctors and researchers. Looking at the problem closely, it’s, in fact, a challenge best suited to be solved by machine learning.
PET scans are used by radiologists to investigate the glucose levels in the brain of patients. In an Alzheimer’s patient’s brain, due to the death of neurons, glucose levels are depleted in regions of an Alzheimer’s patient’s brain. The slow progression of Alzheimer’s disease means that it’s often difficult for radiologists to spot subtle changes in regions of the brain using PET scans. The changes may also be global within different parts of the brain rather than in a restricted area.
It’s as if the radiologists are looking for a needle in a haystack.
Dr. Jae Ho Sohn, MD. MS’s team of scientists trained a machine learning algorithm on 1,921 scans. Then, they tested out the algorithm on patients who are diagnosed with Alzheimer’s. The algorithm correctly identified 92 percent of the patient in the first set and 98% of patients in the second test set. The predictions were made 6 years before the patients received their final diagnosis.
This algorithm still needs more data to have clinical relevance. However, it is a wonderful example of how Artificial Intelligence is changing the way we can diagnose Alzheimer’s disease.
Using Artificial Intelligence to replace human work is not necessarily an efficient use of AI Systems at all. As shown in this example of early diagnosis of Alzheimer's using an AI System, the biggest impact that AI Systems can have in our lives is to use it in areas where human efforts prove to be inadequate.
Who better to look for a needle in a haystack than a machine equipped with the most sensitive of pattern recognition capabilities built for this exact purpose?
Supplementing human work can improve:
- Human’s perception of AI Systems.
- Human’s perceptions of their own work.
- Combining the best features of both humans and AI Systems for a more well-rounded approach to problem-solving.
In this case, the AI System can provide early diagnosis of Alzheimer’s by learning about the PET scans of patients in ways that are superior to the human eye. This provides doctors with more options to treat Alzheimer’s patients. These new treatment options can potentially delay the onset of Alzheimer’s and provide patients and their families with a better quality of life. The radiologist’s job and the doctor’s jobs are not replaced by the AI System. Rather, the AI System is helping the radiologist and the doctor to better serve the patient.
About the Author
Jun Wu is a Content Writer for Technology, AI, Data Science, Psychology, and Parenting. She has a background in programming and statistics. On her spare time, she writes poetry and blogs on her website.