Early detection of dementia

pledged of$20,000 goal

10Days Left

Campaign Ends on January 31, 2022

Project Description

Brief intro: NeuroHarb novelty is based on AI-powered technology to trace and capture various speech modifications to reduce the ambiguity of traditional language tests and support diagnostic efforts of cognitive impairment and disorders by clinicians.

Why it matters:

Cognitive decline in older age, leading to dementia, is a growing problem in industrialized and industrializing societies

2015 : 47 million people
2050 : app.141 million people

Cognitive decline is exhausting economically and socially:

2015 : global cost --$818 billion
App 85% of costs are related to family and social, rather than medical care

Harmful effects are not only associated with advanced forms of cognitive decline as mild cognitive impairment can significantly affect the quality of life of both patients and their caretakers. Overall, around 10-20% of the elderly have mild cognitive impairment. Dementia is developing years before the first symptoms become obvious (up to 15-20 years). Due to the alarming growth of the number of people with dementia worldwide, early detection of cognitive decline and impairment is one of the key issues. Progressive techniques like neuroimaging or genetic measurement show good results, but can’t be widely spread due to invasiveness and high cost.

Cognitive disorders tend to emerge sporadically difficult to detect throughout infrequent checks. Besides the challenges, self-reported questionnaires or data sets collected in lab settings impose, an additional challenge is an access to data that truly reflects the mental state of a patient. The range of variations in scores might overlap with those participants, who do not experience any decline in cognitive functions making them not highly reliable in capturing individual differences. Recent studies have suggested that various speech modifications can be one of the earliest signs of cognitive impairment before other signs become obvious. Traditional language tests are not highly reliable in this context since they tend to provide quite ambiguous results AI-powered techniques

can recognize various language modifications
can capture early signs of cognitive decline
can help to glean an understanding of the true mental state of a patient
can detect symptoms that might potentially be harbingers of various cognitive impairments

Updates

The Campaign FAQs

What is the current stage?

MVP stage, have been working on it myself - including meta-analysis, a concept set up, and coding. My biggest struggle is proprietary data - I had an interest from Mexico to run a pilot but it did not happen.

What is the vision?

Dementia is the 1st step, there is more to it, more details, data, and applications including sensors and movement patterns tracking. The ultimate desire is not only detection but an affordable 'tool' e.g. in a form of interactive exercise to mitigate potential risks based on the results of trials via utilization of neuroplasticity and extension of cognitive reserve. We often overlook simpler solutions and try to mask the symptoms and not prevent potential risks. Also, I am determined to improve training vocabularies due to its bias (majority of white males of middle age and class) to improve overall performance as well as share with the scientific community. Ideally, I want it to be an extension to voice-activated services like alexa or siri and alert people who agree to have this possibility (MCI, dementia and other ).

Monetary Contribution

One-on-one chat to talk about entrepreneurship, neuroscience, tech or data science. This is a place where you can ask any questions.

0 Supporters

Select this reward

50.00 USD

Digital coffee / tea / drink

One-on-one chat to talk about entrepreneurship, neuroscience, tech or data science. This is a place where you can ask any questions.

20 Left

0 Supporters

Select this reward

150.00 USD

Data science / tech mentorship call

I am a data science nomad and AI enthusiast with a strong vision of data usage for social good. I have degrees in economics, bid data/machine learning and applied neuroscience (on a way). Further, I have extensive experience in AI research, data science and AI in the area of cognitive health.

This call is a perfect place to ask questions, I am happy to guide, share my experience, answer questions and suggest the best path based on my experience.

20 Left

0 Supporters

Select this reward

400.00 USD

Data science consulting and R&D

Are you considering implementing data science into your business practice or improving your current strategy? Are you wondering if your data are being used inefficiently? Or are you searching for new unique approaches to solving your business problems?

I will

perform scorecard performance based on your needs to identify strengths, gaps, potential untapped opportunities, and weaknesses
advice on the best plan of action with detailed technical and supporting documentation including staff training at necessity or development of the educational package
advise on best implementation strategy
identify your R&D needs and develop state-of-art approaches or proprietary methodologies based on recent developments in the field

10 Left

0 Supporters

Select this reward

650.00 USD

Machine learning and AI

Are you wondering what might happen in the future or would you like to predict the outcome of your campaign or which customer will churn? Are you wondering what type of customers uses your products or services?

I will

evaluate your existing data and advise on best practices of data storage and pre-processing
perform data analysis and output key metrics and characteristics
develop machine learning model depending on your needs and supporting technical documentation to avoid the effect of a ‘black box’

5 Left

0 Supporters

Select this reward

650.00 USD

Natural language processing

Are you swamped with a massive amount of text data and wondering how to use it? Are you curious about how your product or brand is perceived?

I will

advise on smarter data analysis and categorization of text data
analyze the sentiment of customer’s reviews or comments
identify influencers and social networks of users
perform topic modeling to identify which topics are the most popular among unclassified data or users
cluster users based on their similarity or perform similarity analysis of the text
identify customer’s profiles even without knowing their major characteristics like age or gender (proprietary methodology)

5 Left

0 Supporters

Select this reward

About the Entrepreneur

by NeuroHarb

Paris,

Created 1 Campaign

https://www.f6s.com/neuroharb

Health + Science

Rewards

Monetary Contribution

One-on-one chat to talk about entrepreneurship, neuroscience, tech or data science. This is a place where you can ask any questions.

0 Supporters

Select this reward

50.00 USD

Digital coffee / tea / drink

One-on-one chat to talk about entrepreneurship, neuroscience, tech or data science. This is a place where you can ask any questions.

20 Left

0 Supporters

Select this reward

150.00 USD

Data science / tech mentorship call

This call is a perfect place to ask questions, I am happy to guide, share my experience, answer questions and suggest the best path based on my experience.

20 Left

0 Supporters

Select this reward

400.00 USD

Data science consulting and R&D

I will

perform scorecard performance based on your needs to identify strengths, gaps, potential untapped opportunities, and weaknesses
advice on the best plan of action with detailed technical and supporting documentation including staff training at necessity or development of the educational package
advise on best implementation strategy
identify your R&D needs and develop state-of-art approaches or proprietary methodologies based on recent developments in the field

10 Left

0 Supporters

Select this reward

650.00 USD

Machine learning and AI

I will

evaluate your existing data and advise on best practices of data storage and pre-processing
perform data analysis and output key metrics and characteristics
develop machine learning model depending on your needs and supporting technical documentation to avoid the effect of a ‘black box’

5 Left

0 Supporters

Select this reward

650.00 USD

Natural language processing

Are you swamped with a massive amount of text data and wondering how to use it? Are you curious about how your product or brand is perceived?

I will

advise on smarter data analysis and categorization of text data
analyze the sentiment of customer’s reviews or comments
identify influencers and social networks of users
perform topic modeling to identify which topics are the most popular among unclassified data or users
cluster users based on their similarity or perform similarity analysis of the text
identify customer’s profiles even without knowing their major characteristics like age or gender (proprietary methodology)

5 Left

0 Supporters

Select this reward