(Concise science article summary)
Dementia is often considered to be a singular disease, but research has shown that it is often rooted in multiple interconnected brain diseases. A new blood test has been developed by using medical records from more than 3,000 patients from two clinics at Washington University in St. Louis. An AI algorithm was used to identify the proteins that best signal whether a form of Dementia or Parkinson’s might be present within the brain. Eventually, fifteen proteins were identified as the most useful for the blood test. The researchers compared the AI model’s output with data from a separate group of patients at the Banner Sun Health Research Institute in Arizona.
(Ethical analysis)
Ethical values are rooted in universal respect for access to the necessities of life. Accomplishing such access to survival and human flourishing requires a collaborative social existence that values honesty, transparency, and support for independent reasoning as an essential part of a person’s quest for health, happiness, and survival. Healthcare testing takes place within the context of social and biological necessity and hence requires honesty and transparency about the methodological origins of a given medical test.
AI systems are trained with massive amounts of data as well as statistical models that allow for noticing obscure patterns within massive amounts of data. For this reason, AI is an increasingly useful tool within scientific medical research. However, AI systems’ outputs are biased to the extent that there is a lack of magnitude and meaningful diversity within the training data involved in generating the outputs. The result is that AI generated outputs can be highly inaccurate. A sample size of patients that is slightly greater than 3,000 may not be a sufficient sample size for creating a universally applicable blood test for identifying dementia and interconnected brain conditions. Furthermore, those 3,000 + patients may not sufficiently represent the genetic diversity of the patients who would hypothetically use this blood test in the future.
This does not mean that the AI generated outputs within this study are not incredibly useful. However, honesty and transparency are ethically required when it comes to informing patients and their families about the role of AI within medical research that created the medical tests. Respect for the humanity of others requires respect for one of the most fundamentally essential functions of human nature, which is the application of informed reason for the purpose of advancing health, happiness, and survival. For this reason, it is necessary to inform patients and their families about the potential inaccuracies involved in using AI to develop medical tests. Informing patients and their families need not involve cultivating an excessive bias against the use of AI within medical research.
(citation) A blood test for dementia may tell you if you have more than one type Science News By Michael Howerton
Citations
Y. Xu et al. GPND-AI NULISA: A 15-Protein AI classifier for diagnosis and co-pathology profiling across neurodegenerative diseases. Alzheimer’s & Dementia. Vol. 22, May 2026, e71420. doi: 10.1002/alz.71420