DemWatch

by Akash Patel

Medical Devices & Digital Health

DemWatch and Companion App
“Helping both the patient and physician effectively manage dementia”

Background
Dementia is a term used to encompass a variety of conditions that all share the similarity of a diminished ability to think, reason, and recall events. Although over 5 million adults throughout the United States suffer from some form of dementia, treatments in the field remain severely underdeveloped. The problems with current dementia care lie in the mental instability of patients due to their loss of motor ability and the difficulty for physicians to assess the degree of memory loss. Patients frequently enter a “threat state” of mind in which their inability to recollect information causes them to exhibit aggressive tendencies. When dementia patients are emotionally agitated, they are more likely to have mental instability, either in the form of anxiety or depression. Over 40% of dementia patients exhibit signs of depression, which usually leads to delirium, a condition in which the patient’s memory declines twice as fast. The stress and anxiety cause the patient to enter the “threat state” and their ability to recollect information worsens exponentially, causing their condition and emotions to spiral out of control. To combat this, physicians use drugs that numb patients by suppressing their emotional fluctuations. However, since the nature of dementia varies greatly and is difficult to quantify, physicians rely on guesswork when prescribing drug treatments. As a result, over 73.5% of dementia patients rely on antidepressants or antipsychotics that have been incorrectly prescribed leading to increased memory loss, sedation, and even drug dependence in patients over 60. Even worse, half of the time, drug interventions fail to control the patient’s emotions. Alongside the ineffective drug treatments, non-drug interventions are also being explored. In 163 studies conducted, of over 23,000 dementia patients, non-pharmaceutical interventions were more effective in calming down agitated patients than the drug interventions. However, the interventions were rudimentary including taking the patient outside and listening to music, which lacks applicability and scope in helping patients.

Method
DemWatch and Companion App is a wearable watch and application that utilizes AI, machine learning, and surveys to soothe dementia patients when they are in an agitated state and then uses the gathered data to assist the physician in managing the patient. The watch’s primary function is to soothe the patient when they are agitated through a preloaded slideshow of soothing videos compiled by the caretaker and family. Since soothing sounds and videos have been shown to calm dementia patients, a chatbot with a soothing voice will be included with the slideshow to calm the patient. The watch will monitor the patient’s vitals and mood, playing the slideshow until the patient is relaxed again. In order to do so, the watch will use two main AI. First, is the emotion-tracking AI built upon Affectivia’s AI, that will use a camera built on the watch, in order to read the physiological microdata on the patient’s face that expresses their emotions. To get the patient to face the camera, the watch will produce noises and lights until the emotion tracking AI gets a reading on the patient’s mood. If the assessment is that the patient is agitated, then the second AI will also begin assessing the patient. The second AI, built upon Healthstat’s wearable blood pressure machine, uses a built-in sensor on the back of the watch crown. The AI sensor will take the patient’s blood pressure and vital signs, because when someone is stressed or anxious their blood pressure rapidly increases. The AI will take note of this and in conjunction with the emotion-tracking AI it will be used to assess if the patient is in a “threat state” or not. If the patient is in the “threat state,” the watch will play the slideshow with the chatbot overlay, as well as record the duration of the episode until the patient calms down. To assess if the patient is calm, the emotion-tracking AI and the vital tracking AI, will produce a holistic score on a scale of 1-10 (1 being happy and 10 being extremely aggressive) in order to assess the patient’s mental state. The slideshow and soothing sounds will turn off once the patient’s score reaches below the threshold denoting a calm mood, and the timer recording the duration of the “episode” will stop. If ineffective in calming the patient, the information will be sent to the developer to aid in redeveloping the watch. The companion app will have a set of questions regarding the frequency of the patient’s “episodes” and the duration, the survey will be sent to the caretaker and family. The app will compile the responses and the data from the watch, and send it to the developer and physician. Personal information will be confidential, and the information gathered will be used to improve the watch and app. The information will be used by the physician to understand the severity of the patient’s condition which will help them make judgements on whether to pursue drug interventions and if so to what extent. This will improve the quality of care for the patient and prevent the risks of overprescription leading to a worsened condition. The app and watch will both effectively soothe the patient, preventing them from becoming stressed and aid the physician. The innovation will be beta tested by a test group of dementia patients of diverse origins in order to ensure that it works for all types of patients. After the widespread release, the recorded data from patients will be used to improve the watch’s emotion tracking and survey questions. The innovation will be a medical device that can be recommended by doctors to their patients for use.

Conclusion
Increasing the applicability and scope of non-drug interventions will greatly improve care for dementia patients and decrease the number of people with severe dementia across the board. By effectively suppressing a dementia patient’s anxiety while also aiding their physician in prescribing drugs, the progression of dementia will slow and the patients’ quality of life will improve. DemWatch and Companion App uses AI, machine learning, and surveys in order to help dementia patients relax when they are agitated and record the frequency of their “episodes,” then cross reference the data via the survey, in order to help the physician understand the degree of the patient’s condition and prevent delirium. The watch and app have the potential to revolutionize research in non-drug treatments for dementia patients by bringing AI to dementia care, and improve outcomes across the board in the future. The use of the DemWatch and Companion App is bound to decrease the progression rate of dementia in many patients and improve their quality of life.