Neura
An app to help mental health assessment
An in-home multimodal healthcare system designed to support mental health assessment and intervention for older adults. The system uses AI to track mood, daily activities, and health data, providing evaluations and triggering emergency responses when necessary for proactive support.
Company
Northeastern University (Acedemic Project)
Timeline
11/2024-12/2024
Role
UIUX Designer
Build an AI-driven healthcare platform for older adults, integrating conversational analysis and depression metrics to enhance early detection and enable personalized mental health interventions.
Designed and implemented real-time emergency response mechanisms, improving system dependability and reducing delays in critical health situations.
Team
2 UIUX Designer
1 Devoloper
Key Users
60 and above older adults
Why Mental Health Assessment?
Why Depression
Assessing mental health in older adults is essential due to their increasing population and heightened vulnerability to mental health challenges. Studies reveal that one in five older adults may experience mental health issues, with depression being the most common. Seniors also have the highest suicide rates, emphasizing the urgency of early intervention.
How AI Involved
Monitoring subtle changes in routines, speech patterns, or demeanor, with the assistance of AI, can help detect early warning signs of distress. AI analyzes these patterns alongside physical indicators, such as data from wearable devices, to provide a comprehensive understanding of their mental health stages. By leveraging AI's ability to process and identify early symptoms, we can offer timely support, enhance well-being, and prevent severe outcomes more effectively.
Involved Technology
Neura leverages smart devices to collect health data such as heart rate, blood pressure, and sleep cycles, integrating it with conversational insights from smart home devices like Alexa or Google Home to detect signs of depression. By embedding the Geriatric Depression Scale (GDS) into everyday conversations, Neura provides natural, unobtrusive mental health assessments, especially for users hesitant to consult specialists. Additionally, the app offers features like health reports and an AI chatbot, enabling users to track their well-being and access personalized support seamlessly.
Competitor analysis
StoryBoard
User Story for AI-Driven In-Home Mental Health Monitoring System
The AI-assisted in-home healthcare system is designed for older adults facing unnoticed or untreated mental health challenges. For example, Mr. Smith, an elderly man, begins showing signs of depression, such as withdrawing from activities and talking to his Alexa device for companionship. The AI system passively collects data from his smart devices, monitoring emotional and biometric indicators to detect early signs of distress.
When the AI identifies concerning patterns, such as depressive language and increased stress levels, it alerts Mr. Smith’s primary care doctor. Despite the warning, Mr. Smith dismisses treatment, attributing his sadness to the loss of his wife and feeling isolated as his son is preoccupied with work and family.
As his condition worsens, the AI detects high-risk thoughts and an urgent crisis. It promptly notifies Mr. Smith’s son and healthcare provider, enabling timely intervention. This proactive system continuously monitors mental health, offering timely support and ensuring safety for vulnerable older adults like Mr. Smith.
Voiceflow Prototype
-
Relieving Activities for health and well-being (Oewel et al.)
-
Music playing, mindfulness meditation …
-
-
Relevant Usage
-
Search & Music
-
Health information seeking (symptoms, treatments)
-
-
Combination with psychological diagnosis
-
Depression/Anxiety → ineffective responses
-
-
VA offers more secure disclosure & interaction with care partners
-
Older adults
-
Caregivers: track reports and summaries
-
Voice Interface
-
Design Considerations
-
Behavioral Activation Events
-
DC1: Enable tailored AI responses for various user needs
-
DC2: Fit psychological screening questions into the proper context
-
DC3: Ensure natural and relieving conversation flows
AI response
Model: Claude 3 - Haiku
Prompt structure:
-
Role: You are a healthcare professional, support your user in the best way you can.
-
User Input: Answer this question: {user_input} in one or two sentences.
-
Task Specific: Please give feedback to {task} that they are engaging in a friendly and encouraging manner.
-
Limitation: Don't mention that you are a healthcare professional.
Metric:
-
Geriatric Depression Scale (GDS)
-
A depression screening tool for older adults (Yesavage et al.)
-
Used extensively with older adults
-
-
Natural Language Conversation
-
Make screening questions sound natural/human-like
-
Prompt structure: friendly, supportive but not monitoring
-
-
Create variances
-
-
Fit to Scenarios
-
Applicable to other psychological metrics
workflow for neura
Retired older adults who feel lonely at home living by themselves
We created a conversational agent called Neura. It is a function inside a virtual assistant (ex: Alexa)
This workflow chart outlines how Neura supports mental health for older adults living alone. Users interact with a virtual assistant, like Alexa, which integrates depression screening into conversations. Based on collected data, Neura generates reports with mood trends, behavioral insights, and relieving plans, accessible on mobile, desktop, or Alexa. Caregivers can review and share these reports, with automatic alerts triggered for severe depression scores, ensuring timely support and intervention.
Smart Watch Interfaces
This smartwatch interface detects and displays users' emotional states, analyzing heart rate, blood pressure, stress levels, and daily conversations. With 12 variations across four moods—happy, neutral, sad, and angry—each screen offers supportive messages, encouraging further interaction with the Neura app. This design fosters user engagement while promoting mental well-being.
App Interfaces
-
Emotion Tracking: The app monitors and visualizes users' emotional trends based on data from wearables and user inputs, helping them understand their mental state over time.
-
Conversational Analysis: Through integration with smart home devices like Alexa or Google Home, Neura identifies patterns and keywords in daily conversations, detecting potential early signs of depression.
-
AI Chatbot: An interactive assistant designed to offer real-time support, conduct assessments based on the Geriatric Depression Scale (GDS), and provide natural conversational experiences tailored to the user's needs.
-
Resource Recommendations: Personalized suggestions for activities such as yoga, mindfulness exercises, music, and articles aim to improve emotional health and resilience. The app includes options for both structured programs and self-paced resources.
-
Health Reports: Users can generate detailed health reports for review by family members or healthcare providers. In cases of critical mental health levels, reports are sent automatically to ensure timely intervention.
Key features:
All third-party data collected is used exclusively to assess the user's depression level. The app gathers this data to create detailed health reports that users and their families can review. Users have the option to share these reports with their healthcare providers. However, if the depression level is deemed critical, the app automatically sends the report to the provider to alert them, ensuring timely intervention and preventing potential crises.
Web Interfaces
The Neura web interface offers a comprehensive mental health report tailored for older adults. It integrates physical health metrics, such as blood pressure and heart rate, with behavioral insights derived from conversations, highlighting positive and negative patterns along with key phrases that reflect the user’s day.
A "Send" button allows users to share their reports at their discretion, ensuring autonomy. In cases where the depression level reaches a severe threshold, the system automatically alerts caregivers or healthcare providers. This thoughtful design balances proactive mental health support with user privacy and control.
