The most recent statistics suggest that almost 500,000 Canadians are facing dementia, with 76,000 more being diagnosed annually. Further, dementia unevenly effects women, with two-thirds of Canadians living with dementia being women. Unfortunately, the effectiveness of non-pharmacological treatment for dementia remains low due to the one-size-fits-all, generalized treatment activities currently prescribed to dementia patients across the nation. Being able to track the progression of dementia in a patient and recommend personally tailored treatments means increasing the effectiveness of said treatments, improving outcomes and, ultimately, the lives of those effected.
Dementions is a device that tracks the progression of dementia in a patient and suggests personalized, non-pharmacological activities and treatments. While wearing an electroencephalogram (EEG), a device that measures the brain’s electrical activity, users are walked through various tasks that determine their dementia’s progression. Tasks are divided into categories of dementia symptoms such as memory impairment or behavioral instability and, when the tasks are finished, the patient and caretaker are given optimized activities as well as insights into dementia’s development.
Impacting more than 55 million individuals across the world, dementia. Women are disproportionately impacted, accounting for two-thirds of Canadians living with dementia. Despite this prevalence, the effectiveness of non-pharmacological treatments remains low, largely due to the generalized, one-size-fits-all nature of current therapy regimens. This lack of personalized treatment often results in suboptimal outcomes, leaving patients and caregivers struggling to manage symptoms effectively. Even modest improvements in outcomes could lead to profound societal benefits in terms of improved well-being for hundreds of thousands of families.
The problem is clear: there is a critical need for tools that enable the accurate tracking of dementia progression and the delivery of personalized, non-pharmacological treatments. Dementions addresses this gap by using an electroencephalogram (EEG) to assess patients' neurological responses while they complete tasks tailored to key dementia symptoms, such as memory impairment or behavioral instability. By analyzing this data, Dementions provides actionable insights into disease progression and recommends optimized, evidence-based activities. This personalized approach aims to improve treatment adherence, enhance patient outcomes, and ultimately transform how dementia care is delivered in Canada and globally.
After consulting caretakers working on the front lines of dementia treatment we came to learn that status quo of handling dementia patients in care homes is underscored by human-centred analysis. Due to a lack of existing technological solutions on market, caretakers do not regularly track the development of dementia as a result of the time it would take to conduct daily analyses. Furthermore, caretakers we spoke to outlined the ineffectiveness of this strategy is tracking progression, as even experienced caregivers may develop tunnel vision as opposed to developing an objective understanding. This translates into a hit-and-miss approach towards personalization of activities. Further research and communication with clinical psychologists further highlighted the lack of objective and technology-based solutions to dementia treatment.
Currently, the best tool for systemically and thoroughly assessing mental status is a Mini Mental State Examination (or MMSE for short). The MMSE is an eleven question measure that tests five areas of cognitive function: orientation, registration, attention and calculation, recall, and language. The maximum score is 30. A score of 23 or lower is indicative of cognitive impairment. The MMSE can screen patients with cognitive impairment from those without it, and can categorise dementia into stages based solely off cognitive impairment, it doesn’t take into account other important factors in understanding stage progression. However, the tool is not able to pin point the case for changes in cognitive function. In addition, the instrument relies heavily on verbal response and reading and writing. Therefore, patients that are hearing and visually impaired, intubated, or those with other communication disorders may perform poorly even when cognitively intact.
Wishing to leverage EEG technology in determining dementia stage development, we came across a dataset that mapped the MMSE scores of dementia patients with their EEG signal data and built a model that achieved an accuracy of 93%. We used OpenBCI’s Ganglion Headset and conducted user tests with dementia patients, who reported that the headset was uncomfortable and intimidating. Clinical psychiatrists we spoke to reported that large wearables would scare away patients and would be adopted less than a lower-impact wearable would.
We decided to pivot to using data gathered from the Muse 2 headset to overcome this problem. The model’s accuracy using the Muse is 86%. However, when going to re-conduct user tests, we realized the model’s accuracy decreased, likely due to the decreased signal quality coming from the Muse. To solve this, we built in multiple layers of redundancy. Doing this ensures our analysis is spread over multiple domains effected by the disease and reduces the effect of a slightly less accurate model.
In order to obtain an efficient and comprehensive categorization of dementia stage on a daily basis, Dementions standardizes the various scales of progression into a holistic five-component scale that gives a more in depth and time effective analysis of patients that will enable interactive and specialized care and treatment to patients. Our scale is 1 - 100 meter determining the severity of dementia across five critical components: physical, behavioral, emotional, speech, and memory. Each component is evaluated independently, with scores reflecting the degree of impairment. This metric allows for a nuanced understanding of a patient’s condition, moving beyond general assessments to provide a detailed profile that can guide personalized care plans. By incorporating real-time data from wearable devices, caregiver reports, and AI-driven analytics, the scale ensures a dynamic and precise evaluation process. This approach not only enhances diagnostic accuracy but also facilitates the design of targeted interventions tailored to the specific needs of patients, ultimately improving their quality of life and supporting caregivers in delivering the most effective care.
Dementia profoundly impacts multiple dimensions of a patient’s life, and effective diagnosis requires a holistic approach that addresses these areas. Below is a structured exploration of the physical, behavioral, emotional, speech, and memory dimensions of dementia, including symptoms, diagnostic methods, innovative approaches, activity recommendations, and evidence-based results.
The physical dimension of dementia primarily involves a decline in motor coordination and visuospatial abilities. As the disease progresses, patients often experience difficulty with tasks such as dressing, eating, and navigating familiar environments. These impairments not only compromise their independence but also elevate the risk of falls and injuries, posing serious health risks (World Health Organization, 2021). The loss of fine motor control further exacerbates challenges in daily living, as patients may struggle with tasks requiring precision, such as buttoning a shirt or using utensils. These physical symptoms significantly affect the quality of life and contribute to a sense of helplessness.
The Clock Drawing Test (CDT) is a widely used tool for assessing visuospatial and motor skills. It requires patients to draw a clock displaying a specific time, and clinicians evaluate the accuracy of the spatial arrangement and number placement. While the CDT is effective in identifying cognitive and motor deficits, traditional scoring methods rely on subjective interpretation, leading to variability in assessments (Shulman, 2000). This subjectivity can result in inconsistent diagnosis and treatment planning.
To address the limitations of traditional methods, Dementions leverages a Convolutional Neural Network (CNN) for clock analysis. CNNs are highly effective in image recognition tasks, capable of detecting nuanced patterns in spatial arrangements and line accuracy that human evaluators may overlook (LeCun et al., 2015). This objective approach ensures consistency and precision, allowing for repeatable and accurate assessments. By automating the analysis, the software eliminates human bias and provides a more detailed understanding of the patient's motor and visuospatial impairments. We have also integrated a gesture imitation task that has users make their thumb and middle finger touch. Then another CNN analyses the image and gives the patient a score based on how fast they are able to complete the task.