IBM: AI-system for assessing the progression of Parkinson's disease

2021: Announcement of a software predicting the progression of Parkinson's disease symptoms

At the end of July 2021, IBM and the Michael J Fox Foundation (MJFF) announced the creation of software that will allow doctors to better understand the biology of Parkinson's disease (PD).

IBM's research arm has created artificial intelligence that can group typical symptoms of Parkinson's disease and accurately determine the progression of those symptoms in a patient, despite whether they are taking medications masking those very symptoms.

IBM unveils software that predicts progression of Parkinson's disease symptoms

This discovery is published in the journal The Lancet Digital Health. IBM Research and MJFF have been collaborating since July 2018 to explore the possibilities of applying machine learning to help physicians better understand the biology behind PD, especially given that it develops differently in different individuals.

While many previous studies have focused on defining characteristics of Parkinson's disease using only baseline information, our method relies on patient data from up to seven years. In addition, our model makes limited a priori assumptions about the pathways of disease progression in the near future, this prognosis is given based on comparison with our previous studies.

As a result, the researchers found that a patient "s condition can vary by several factors, including the ability to perform daily activities; problems related to slow motion, tremor and postural instability; non-motor symptoms such as depression, anxiety, cognitive impairment, and sleep disorders. At the same time, the model could predict when the disease would go into a more serious state.

We found that the findings supported the hypothesis of diverse progression pathways, as indicated by the many trajectories of disease development that we observed. However, the AI model is still able to make accurate predictions. By training the model on one dataset, she was able to successfully predict the progressive state of Parkinson's disease associated with consequences such as dementia and the inability to walk unaided, IBM Research said on August 2, 2021.

Based on these findings, the researchers hope the model can be used to assist physicians in advising patients, as well as to identify those who may benefit from clinical trials.

As next steps, further refinement of the model is planned by incorporating genomic and neuroimaging measurements so that their model can provide an even more detailed characterization of the disease as a whole.^[1][2]

Notes