Embedded Files
Background information of the projects / tasks which you were involved in.
The project aims to look into creation of device that is more affordable and effective in detecting Parkinson’s disease. The value of such a project is that due to the increasing ageing population in Singapore, the number of PD patients in Singapore are increasing. The treatment these patients have to seek for PD is costly, reaching an average annual cost of SGD11345 per patient, with homecare taking up 76.1%. Thus higher productivity is needed to ease the burden of healthcare, society and patients in Singapore. Besides that Parkinson’s diseases causes freezing of gait (FOG) which is marker reduction of forward progression of the feet despite the intention to walk. This is a common cause of fall and hence it is important to create a device that is able to give patients warning before they are about to experience FOG.
The task that I engaged in was to achieve deeper understanding in Parkinson’s disease and machine learning. And the way it could be done is in the following approach: Firstly we were tasked to complete a literature review related to Parkinson’s disease. Then, we were given individual topics to research on. I was given the task to do research on machine learning and linking it to how it’s used in devices for detection of Parkinson’s disease
Elaboration / record of the activities done
The steps to completing the project are reading of research papers and articles, picking out relevant information, writing literature review and making presentation slides.
For literature review, the professor gave us a total of 31 articles and research paper to read on. After reading, we have to sieve out important information and put it into our literature review.
For individual research, we have to find our own articles to read and understand the topics. After this we collated all this information into a google docs together with the rest of our group mates. We then proceed to compile these information into presentation slides and the presented it to the professor. Finally, the professor gave us some tips to improve on our presentation.
The rationale for such a process is for gaining deeper understanding of Parkinson’s disease and machine learning. The benefits of such a process is that it allowed me to read up more on Parkinson’s disease from trusted sources as these articles were sent by the professor. It also allowed me to gain deeper understanding about Parkinson’s disease and machine learning first which will then allow us to better link both topics together while doing our final presentation slides later on.
The challenges of the project was reading of research papers. These research papers are quite long and technical, hence reading it can get quite tiring at times. We must also be able to identify the relevant information and summarise these information from research paper to include our report.
3 content knowledge / skills learnt
1. Machine learning
Machine learning is an application of artificial intelligence (AI) that provides systems the ability to automatically learn from data, identify patterns and improve from experience without being explicitly programmed. This allows machines to make better decisions in the future based on the examples that we provide
There are 4 methods of machine learning:
a) supervised
Uses labelled set of input data and output data to train model to generate reasonable predictions for the response to new data. The system is able to provide targets for any new input after sufficient training. The learning algorithm can also compare its output with the correct, intended output and find errors in order to modify the model accordingly.
Eg: diagnostics
b) unsupervised
Information used to train is neither labelled nor classified. Unsupervised learning studies how systems can infer a function to describe hidden structures from unlabelled data. The system does not figure right output, but it explores data and can draw inferences from datasets to describe hidden structures from unlabelled data. Eg : big data visualisation
c) Semi supervised
uses a small amount of labelled data to label large amount of unlabelled data. the programmer will cluster similar data using an unsupervised learning algorithm and then use the existing labelled data to label the rest of the unlabelled data. The typical use cases of such type of algorithm have a common property among them – The acquisition of unlabelled data is relatively cheap while labelling the said data is very expensive. The systems that use this method are able to considerably improve learning accuracy
d) Reinforced
a learning method that interacts with its environment by producing actions and discovers errors or rewards. Trial and error search and delayed reward are the most relevant characteristics of reinforcement learning. This method allows machines and software agents to automatically determine the ideal behaviour within a specific context in order to maximize its performance. Simple reward feedback is required for the agent to learn which action is best; this is known as the reinforcement signal. Reinforcement learning is currently the most effective way to hint machine’s creativity.
Eg: game AI
2. Parkinson’s disease
Parkinson's disease (PD) is the second most common neurodegenerative disorder in the world, with an estimated seven to 10 million people worldwide have Parkinson’s disease. The prevalence of the disease ranges from 41 people per 100,000 in the fourth decade of life to more than 1,900 people per 100,000 among those who are 80 and older. The rate of newly diagnosed cases generally increases with age, although it can stabilize in people who are older than 80. An estimated 4 percent of people with Parkinson’s are diagnosed before age 50. Men are 1.5 times more likely to have Parkinson’s than women. The disease affects patients’ quality of life, making social interaction more difficult and worsening their financial condition, due to the medical expenses associated with the disease
Main cause of PD is the loss of nerve cells in the part of the brain called the substantia nigra. Nerve cells in this part of the brain are responsible for producing a chemical called dopamine which acts as a messenger between the parts of the brain and nervous system that help control and coordinate body movements. For PD patients, the amount of dopamine in the brain is reduced and hence patients’ physical movements as well as mental areas like their mood and cognition are affected. This degeneration of nerve cells in the subthalamic area can be attributed to many factors including possible genetic influences, premorbid and nonmotor issues, and a variety of neurologic, cognitive, and psychiatric symptoms
PD affects patients’ physical movements, such as tremors and postural instability, as well as mental states like their mood and cognition. Due to the increasing ageing population in Singapore, the number of PD patients in Singapore are increasing. The treatment these patients have to seek for PD is costly, reaching an average annual cost of SGD11345 per patient.
Symptoms of PD could range from minor inconveniences to lifestyle changing.
PD will cause disorders such as anxiety disorder, depression, apathy and pseudobulbar affect (PBA), also called emotional lability, is a syndrome of emotional dysregulation characterized by spells of crying and/or laughing. Depression increases the risk of secondary impairments such as cognitive impairment and disability or dementia. PD patients have a different symptom profile than depressed patients without PD. This profile includes higher rates of anxiety, pessimism, irrationality, suicide ideation without suicide behavior and less guilt and self-reproach.
The freezing of gait (FOG) is defined as a ’brief, episodic absence or marker reduction of forward progression of the feet despite the intention to walk’ It was found to be the most distressing symptom of PD. It is a common cause of fall, interferes with daily activities, makes people with Parkinson’s lose confidence in walking and significantly impairs quality of life
3. Presentation slide skills
We were given advice and ways to improve on our presentation by the professor
We learnt that to make good slides, we should keep things short and add in more images. By doing so the slides will look more interesting and it is also easier for the audience to read
For presentation wise, besides just memorising script, it is also important to read up beyond what’s presented on that topic as this will aid in q and a sections.
Such skills can be useful in the future when we do projects
2 interesting aspects of your learning
1. Especially when Parkinson’s disease have such a huge impact not only on individual but also the society, doing this project was rewarding as invention of devices that aid in treatment or detection would be a great help in saving lives of people. some of the devices made for detection for PD in the past were bulky and expensive, which is a burden to patient. For example: the PD shoes. It has many wires on the outside of the shoes which may be inconvenient for the patients.
With current technology such as mobile applications, it is not only more convenient but also more affordable.
For example: imotor and mPower apps.
imotor
PD patients frequently develop fluctuations in motor function as a side effect of commonly used anti-parkinsonian medications. These fluctuations, termed “ON” and “OFF” states, are difficult to manage in clinical care and often serve as endpoints in PD clinical trials
Neural Network Construction (NNC) technique was used here to classify data collected by a mobile application (iMotor) into two categories: PD for patients and HV by using a variety of upper limb functional tests
The NNC algorithm discriminated individual PD patients from HVs with 93.11% accuracy and ON vs. OFF state with 76.5% accuracy.
iMotor is a clinically validated digital platform that utilizes the smart-tablet's screen sensing capabilities during upper limb function tests to objectively collect data that allow detection and quantification of neuromotor function
mPower
Early appearing symptoms of Parkinson’s disease include tremor, rigidity, and vocal impairment (dysphonia). Hence speech indicators are important in the identification of PD based on dysphonic signs. Biomarkers derived from human voice can offer in-sight into neurological disorders, such as Parkinson's disease (PD)
The data used for this analysis were collected through mPower, a clinical observational study conducted by Sage Bionetworks using an iPhone app to collect digital biomarkers and health data on participants both with and without PD [10]. To maintain user confidentiality and en-able linking across datasets, each participant was uniquely identified with a distinct healthcode. The method used for collecting the audio data was as smartphone voice activity that recorded participants articulating the /aa/ phoneme for10 seconds
3. With artificial intelligence being a very popular topic of discussion, it was interesting to learn about how AI and machines actually works through the help of machine learning.
1 takeaway for life
One takeaway for life is that although healthcare is a very important part in our lives, it is only with the help of engineering that advanced technologies related to health are created, allowing health professionals to be able to provide better and more effective healthcare services towards us. Not only that, healthcare with the help of technology is also able to bring about more affordable and more accessible healthcare services to everyone.
Page updated
Google Sites
Report abuse