A Neuroprosthetic Solution for Parkinson’s Disease: Decoding the Human Mind
Abstract:
Current Parkinson’s Disease (PD) treatments range from dopamine medicines to surgical stimulation solutions to the brain, such as Deep Brain Stimulation (DBS). In addition, the appropriate diagnosis of early Parkinson’s treatments determines each patient’s unique treatment plan. Regardless of the accuracy of the diagnosis and treatment, recent solutions solely garner Parkinson’s symptoms’ improvements. Also, several factors, including age, blood pressure, etc. impact the relief level of such treatment plans. A future solution for PD that includes a minimally-invasive apparatus using a brain-machine or computer interface (BMI), along with a custom hardware that runs a machine learning (ML) or artificial intelligence (AI) algorithm to decode neuron activity in the premotor & motor cortex region would better treat PD.
What is Parkison’s Disease?:
PD is a degressive brain disorder that affects basic body movements like walking, limb usage, speech, etc. According to Parkinson’s.org, an estimated “nearly one million will be living with PD in the U.S. by 2020, which is more than the combined number of people diagnosed with multiple sclerosis, muscular dystrophy and Lou Gehrig's disease (or Amyotrophic Lateral Sclerosis)...” (Parkinson’s.org 2020). Also, PD causes changes to both mental and behavioral aspects, thus resulting in depression, anxiety, and memory difficulty beyond the typical Parkinson’s symptoms of shaking (tremors), stiffness, and difficulty with walking, balance, and coordination, (NIH.gov 2018). Developing treatments for PD could lead to improvements in both the focused and other neurodegenerative diseases and illnesses with PD-like symptoms. Thus, a larger number of individuals benefit from PD research.
Given PD’s statistical relevance in today’s society, it is of greater necessity to provide resources and money towards PD-related research. For instance, NIH.gov (NIH.gov 2018) cites age as an increased risk factor for Parkinson’s. A study completed by Jama Neurology article (JAMA Neurology 1987) explores the rise in PD over 20 years. The article refers to the following in regards to smoking, age, and PD: “... ‘Between then (1996) and 2009, the smoking rate among American men went down from 67 percent to 23.5 percent… incidence rates of parkinsonism in men increased from 38.9 per 100,000 person-years between 1976 and 1985 to 55.9 between 1996 and 2005…”. The evidence suggests a cause-and-effect relationship between smoking, health improvement, and the prolongation of life. Life prolongation due to less smokers leads to a higher incidence of PD. Therefore, newer solutions for PD need to address such potential improvements within society.
Another such example is depicted in an article regarding the rise of PD, in which the author states that the most widely used surgical treatment in recent years for PD, deep brain stimulation (DBS), is often too costly or not available in certain regions of the US (JAMA Neurology 2018). Less effective alternatives, such as the dopamine hormone (Levo-Dopa), are then used alone. So, wealth and financial status also act as limiting factors in PD treatments.
Current PD Solutions:
PD is characterized by the loss of dopamine producing neurons in the Basal Ganglia, and treatments of PD utilize a combination of dopamine-enhancing medication paired with surgical treatment options. Neurons in the Basal Ganglia require the neurotransmitter dopamine to activate other neurons, so the loss of dopaminergic neurons slows down this communication and results in slow movement, stiffness, and the loss of balance.
To initially treat the diminished levels of dopamine, the most widely used medicinal solution for Parkinson’s was developed in the 1960s, called Levodopa or L-dopa. L-dopa is a chemical that is converted into dopamine. The initial version of the treatment caused several side effects, including nausea and vomiting. To reduce these side effects, scientists and researchers combined levodopa with several other chemical compounds, also prolonging levodopa effects. For instance, Carbidopa is added to levodopa to prevent it from converting to dopamine outside the brain, preventing nausea and vomiting. Another example of PD chemical-based treatments is dopamine agonists. In contrast to dopamine-based medicines, dopamine agonists only mimic the dopamine effects on the brain and are not as effective as levodopa-infused treatments. Over the years, however, the benefits and effects of chemicals and levodopa treatments wear off (Parkinson's News Today 2020). This results in the intake of higher doses of levodopa and, eventually, drug indused dyskinesia. For such advanced stages, surgical treatments are available, the most commonly used and effective FDA-approved treatment being DBS. In order to understand the mechanism behind DBS, one needs to understand the anatomy of the Basal Ganglia in relation to the motor cortex region in the brain. Before sending signals out to the peripheral nervous system, the motor cortex sends motion related information to neurons in the Basal Ganglia of the brain, which has two pathways to regulate motion. These are the direct pathway, which produces excitatory feedback to the cortex that causes desired muscles to function, and the indirect pathway, producing inhibitory signals to suppress movement in undesired muscles; both pathways require dopamine for their actions (Neijm.org 2006). In severe PD cases, due to reduced dopamine, the neurons in the indirect pathway stops functioning; this results in too much inhibition on both desired and undesired muscles, making the intended motion slow and jerky.
In DBS surgical treatment, surgeons place two 12-gauge electrodes (one for each hemisphere) into the indirect pathway. An electric current generator is implanted near the collarbone area to send electrical pulses to the brain via the electrodes, causing the inhibitory neurons in the indirect pathway to start working in the basal ganglia (Neijm.org 2006). This process decreases Parkinson's disease symptoms, mainly tremors, but does not treat or reduce degeneration of the brain. There are also various risks pertaining to DBS treatment, including infections, seizure, strokes, or brain hemorrhage (Science Daily 2012). As the DBS electrodes are placed using 12-gauge wire from the top of the brain and are several inches long, reaching into the basal ganglia, the DBS solution is not appropriate for all patients, especially those with health conditions, like high blood pressure or very old age.
Figure 1. Direct and Indirect Pathways of the Basal Ganglia
G. (2013, October). Basal Ganglia [Digital image]. Retrieved June, 2020, from https://s3.amazonaws.com/static.wd7.us/1/17/Basal-Ganglia.gif
A Possible Solution to Parkison’s Disease:
A two-piece neuroprosthetic for Parkinson’s Disease that uses both an implant for brain machine interface (BMI) to capture neural activity and an external hardware to run a novel machine learning (ML) algorithm that targets PD may work as a solution to PD in the future, improving on the concept of DBS electrodes.
There are millions of neurons in the pre-motor & motor cortex regions that affect different types of body movements. For treating PD tremor symptoms, several of these neural groups need to be monitored and controlled.The microelectrodes will monitor individual neurons by sending or receiving information to a ML hardware wirelessly. Each electrode will capture specific neuron activity in analog and convert it to digitized firing patterns, as shown in figure 2., to decode different body movements. Signals from each channel are digitized in real-time to identify action potentials (spikes or neuron firings) using an on-chip detection circuit (in the implant).
Figure 2 Digitized signal for representing neurons firing output for the 32 channels Musk E. (2019). Figure 8 [Digital Image]. Retrieved June, 2020, from
https://www.biorxiv.org/content/10.1101/703801v4.full.pdf
This solution adds such a unique Machine learning (ML) algorithm, a method that has not been tested with previous BMI trials in tech companies, to process the signal patterns, decode the kind of motion being attempted, and prepare the microelectrodes to either stimulate or block the motion-specific neurons. To best implement the solution, the ML algorithm can be designed to look at neural-triggered patterns for PD-specific movements; this will be based on patterns learned off of motor neurons’ firing patterns in clinical trials of healthy patients.
The proposed solution improves on current PD solutions, such as DBS, as it is safer and results in less or no tissue damage. As opposed to the larger and deeply-invasive DBS electrodes, the micro electrodes are placed at 2 to 3 mm depth in the cortex. In addition, although the solution cannot deter or directly relieve the degenerative effects of PD, it has the potential to take over the motion modulating function of the neurons as the disease progresses. Overall, the function of the algorithm would be to either complement or take over the motor functions of the Basal Ganglia.
Figure 3. An Overall View of Proposed PD Treatment
Conclusion:
As scientists search for improved solutions to Parkison’s disease, it will be important for them to look at the intersection of machine learning and AI as a possible avenue for treatments. Nonetheless, scientists must anticipate possible obstacles in implementing such solutions. Treatments need to address that the exact location of electrode placement may vary from person to person in such treatments. Also, solutions need to analyze the need to develop better compression and communication techniques to receive, process, and send the huge amounts of information between the ML device and brain implant, within milliseconds. Thus, future technological developments should help limit the prevalence of Parkinson's disease in society, seeking better quality of life during older ages for several individuals.
Bibliography
1. February 1, 2018 Issue of JAMA Neurology. (n.d.). Retrieved from https://jamanetwork.com/journals/jamaneurology/issue/75/2
2. Deuschl, G., Al., E., For the German Parkinson Study Group, University, A., D. F. Gudbjartsson and Others, A. L. Wyllie and Others, & L. A. Jackson and Others. (2006, November 23). A Randomized Trial of Deep-Brain Stimulation for Parkinson's Disease: NEJM. Retrieved September 06, 2020, from
https://www.nejm.org/doi/full/10.1056/nejmoa060281
3. Is Parkinson's Disease on the Rise? (n.d.). Retrieved September 06, 2020, from https://www.parkinson.org/blog/science-news/science-article/parkinsons-prevalence
4. JJohn A. Baron, M. (1987, November 01). Parkinson's Disease and Smoking. Retrieved September 06, 2020, from
https://jamanetwork.com/journals/jamaneurology/article-abstract/586933
5. Less-invasive method of brain stimulation helps patients with Parkinson's disease. (2012, October 16). Retrieved September 06, 2020, from
https://www.sciencedaily.com/releases/2012/10/121016103413.htm
6. Marras, C., Beck, J., Bower, J., Roberts, E., Ritz, B., Ross, G., . . . Parkinson’s Foundation P4 Group. (2018, July 10). Prevalence of Parkinson's disease across North America. Retrieved September 06, 2020, from
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6039505/
7. PhD, S. (2020, April 03). Dopamine Infused Directly to Brain Eased Symptoms in Monkeys. Retrieved September 06, 2020, from
https://parkinsonsnewstoday.com/2020/03/30/dopamine-infused-directly-to-brain-eased-s ymptoms-in-monkeys/