mindLAMP Integration | For Developers

About mindLAMP

mindLAMP (Learn, Assess, Manage, Prevent) is an open-source platform developed by Dr. John Torous and his team at Harvard Medical School/Beth Israel Deaconess Medical Center. It's become a foundational tool for digital psychiatry research and practice.

Why mindLAMP Matters

mindLAMP represents a rare combination in digital mental health: rigorous academic research, open-source principles, and practical clinical utility. It's built by researchers who understand both the potential and the pitfalls of technology in mental health.

Key mindLAMP Capabilities

Active Data Collection

Surveys, ecological momentary assessments (EMA), cognitive tests, and custom activities. Researchers can design their own data collection instruments.

Passive Sensing

Digital phenotyping through smartphone sensors: GPS, accelerometer, screen activity, call/text patterns. Research-grade data collection.

Clinical Dashboard

Visualization of patient data for clinicians. Supports measurement-based care and shared decision-making.

Research Platform

HIPAA-compliant data storage, export capabilities, and integration with research workflows. Supports multi-site studies.

A Complementary Approach

Our tools are designed to complement, not compete with, platforms like mindLAMP. We focus on specific intervention modalities where we have something to contribute:

mindLAMP Strength	MindBridge Complement	Potential Integration
Data collection infrastructure	Physiological intervention tools	Breathing tool as mindLAMP activity
Passive sensing / digital phenotyping	Non-verbal visual interventions	Trigger biophilic tools based on passive data
Survey/EMA platform	Between-session skill practice	Practice tools linked to EMA prompts
Research data management	AI safety frameworks	Safety layer for any AI-enhanced activities
Clinical dashboard	Engagement without manipulation	Ethical design patterns for activities

Technical Integration Possibilities

1. Activities and Interventions

mindLAMP supports custom activities. Our tools could be embedded as activities:

Integration Example: Resonance Breathing

Our breathing tool could be added as a mindLAMP activity, with session data (duration, completion) logged to the research database. This enables:

Correlation of breathing practice with EMA mood ratings
Analysis of optimal timing based on passive data patterns
Research on physiological interventions at scale

2. Just-In-Time Adaptive Interventions (JITAIs)

mindLAMP's passive sensing could trigger our interventions at optimal moments:

Concept: Pattern-Triggered Intervention

mindLAMP detects indicators of elevated stress (reduced movement, disrupted sleep, changed phone usage). The system prompts the user to engage with a breathing exercise or biophilic visual tool—delivering intervention when it's most needed.

Research question: Does timing interventions based on passive data improve outcomes compared to fixed schedules?

3. Data Standards Alignment

For any integration, data formats need to align:

Activity completion: Start time, end time, duration, completion status
User settings: Preferences, accessibility options
Research metadata: Version, configuration, study arm
Privacy considerations: No PII in activity logs

Research Collaboration Possibilities

We're interested in supporting research that advances the field. Potential collaborative studies:

HRV Biofeedback Study

Pilot RCT of resonance breathing tool, with mindLAMP collecting passive data and outcome measures. Primary endpoints: anxiety (GAD-7), HRV coherence.

Biophilic Intervention Study

Evaluate visual immersion tools for stress reduction, using mindLAMP's EMA for momentary mood assessment and passive sensing for behavioral correlates.

JITAI Timing Study

Compare fixed-schedule vs. passive-data-triggered intervention delivery. Does timing based on digital phenotyping improve engagement and outcomes?

Non-Verbal Intervention Acceptability

Qualitative research on user experience with non-verbal tools. Who benefits most? What are barriers to engagement?

Practical Considerations

What We Can Offer

Open-source intervention tools (breathing, biophilic visual, observer cultivation)
Implementation specifications for integration
Research collaboration on physiological intervention studies
AI safety framework documentation and consultation
Literature synthesis and evidence summaries

What We're Seeking

Feedback on tool design from clinical and research perspectives
Collaboration on validation studies
Guidance on integration best practices
Input on additional tools that would be valuable

Our Constraints

Transparency

We're a research initiative with limited resources. We can contribute tools, documentation, and research effort, but we're not positioned to provide enterprise support or assume clinical responsibility. Any deployment would need to be led by appropriately credentialed teams.

Building on Their Work

We want to be clear: we're building on the foundation others have laid. Dr. Torous and colleagues have done the hard work of:

Establishing digital psychiatry as a rigorous research field
Building open-source infrastructure (mindLAMP)
Publishing extensively on both potential and pitfalls
Advocating for ethical standards in digital mental health
Training the next generation of researchers

Our contribution is modest by comparison. We're synthesizing research, developing specific tools, and trying to translate safety insights into practical guidance. The goal is to add to what exists, not replace it.

Our Position

We believe the field benefits from multiple teams working on different pieces of the puzzle. mindLAMP excels at data infrastructure and research platforms. We're focused on specific intervention modalities and safety frameworks. Together, these pieces fit.

Get in Touch

If the mindLAMP team or other researchers are interested in exploring integration or collaboration:

Collaboration Inquiry

We're available to discuss integration possibilities, research collaboration, or other ways we might support your work. Connect through the 22b1 Research Initiative.

← Anti-Patterns to Avoid

Documented failure modes.