Personal Research Project

Project MATHI

Modular Auditory Therapy for Hallucination Interruption

Shifting Perception. Restoring Auditory Reality.

A neuroacoustic system designed to support individuals experiencing auditory hallucinations through adaptive sound masking and perceptual recalibration.

Created by Suhan D

Student of Chirec International School, Telangana, Hyderabad

The Problem

Understanding Auditory Hallucinations

Auditory hallucinations in schizophrenia are persistent, distressing, and often resistant to treatment. Medications can cause severe side effects and don't help everyone, while advanced therapies like VR, EEG, or neurofeedback remain costly and limited to clinical settings.

Types of Auditory Hallucinations

An estimated 70% of people with schizophrenia experience hallucinated voices, often daily. These voices can be threatening, distressing, and persistent, severely affecting mental health and focus.

Command voicesRunning commentaryMultiple voicesThreatening voices

Existing Interventions

  • Pharmacological: Medications with significant side effects
  • Hardware based: Expensive and inaccessible to many
  • Not user-driven: Lack of agency and flexibility

"Auditory hallucinations are real - and deeply misunderstood."

The Gap in Treatment

Most importantly, none of these approaches offer the patient a sense of real-time control. Today, there is no lightweight, adaptive, and user-managed system that delivers relief at the exact moment hallucinations occur. This absence of autonomy is not just a technical gap—it's a human one.

MATHI was built to close it. Because if the mind is where hallucinations live, then sound might just be how we speak back.

The Solution

What is MATHI?

MATHI is a neuroacoustic system designed to support individuals experiencing auditory hallucinations. By combining adaptive sound masking and perceptual recalibration, it offers a lightweight, non-invasive, and real-time way to interrupt intrusive voices and retrain auditory attention.

"MATHI gives users control over their auditory environment in real-time, creating a sense of agency that's often missing in traditional treatments."

GM

Adaptive Masking (AMT)

Uses real-time microphone input to detect ambient noise and apply tailored sound masking matched to environmental frequency patterns.

Real-time frequency analysis
Dynamic noise generation
Personalized masking profiles

Perceptual Anchoring

Adds natural background sounds like birds or rain to shift focus away from hallucinated voices and toward real-world audio.

Naturalistic sound library
Attention redirection
Grounding techniques

Salience Control

Users can adjust masking intensity based on personal need, simulating control over the "volume" of their hallucinations.

User-controlled intensity
Adaptive gain control
Personalized thresholds

Relief Feedback System

Logs each successful session and automatically adapts the system to reduce intensity over time — simulating perceptual recalibration.

Session tracking
Adaptive algorithms
Progress visualization
The Prototype

MATHI Interface

This is the live MATHI interface — a web-based sound therapy system for hallucination relief.

MATHI Interface

Interface Features

  • Now Playing: Displays active sound + waveform feedback with real-time visualization
  • Salience Control: Lets users adjust masking intensity (1–10) for personalized relief
  • Ambient Mode: Adds calming anchor sounds like rain or birds to ground perception
  • Control Modes: Passive, Manual, or Rotational anchor playback for different situations
  • Sound Profiles: Six tailored sound types including Serenity, Shield & Adaptive AI
  • Log Relief: Tracks session success and adapts over time for improved outcomes

A real-time, adaptive therapeutic system — running entirely in your browser.

Sound Therapy

Sound Profiles

MATHI offers six tailored sound profiles, each powered by a distinct type of colored noise. These profiles are grounded in psychoacoustic science and designed to mask hallucinated voices based on their intensity, frequency, and context. From deep calming tones to full-spectrum blocking, each sound is crafted to give users control over their auditory environment — in real time.

Serenity

Pink Noise

Creates a soft, natural soundscape to reduce low-level hallucinations and calm the mind. Best for quiet environments or moments of rest.

Clarity

Brown Noise

Deep, low-frequency masking that cuts through intrusive thoughts and supports cognitive focus. Ideal during work or high-anxiety episodes.

Dream

White Noise

Full-spectrum noise that blocks a wide range of auditory disturbances. Used for sleep support and high-salience hallucinations at night.

Shield

Mid-Band Filtered Noise (0.5–4 kHz)

Specifically targets and masks speech-like frequencies, offering targeted relief for voice-based hallucinations.

Refresh

Light Pink Noise

Gentle, low-intensity masking for daytime use. Provides comfort without overwhelming the user, perfect for background use.

Adaptive AI

Dynamically Selected (Based on Mic Input)

Uses real-time ambient analysis to choose the best masking profile automatically. Adjusts every 10 seconds based on environmental noise.

Noise Types

White Noise

  • Energy distribution: Equal energy across all frequencies
  • Sounds like: TV static, air conditioning, fan noise
  • Effect: Strong broadband masking; blocks high-salience or sharp voices
  • Used in: Dream

Pink Noise

  • Energy distribution: Equal energy per octave (logarithmic scale)
  • Sounds like: Steady rainfall, wind, ocean waves
  • Effect: Calming and natural; reduces low-level internal chatter
  • Used in: Serenity, Refresh

Brown Noise (a.k.a. Red Noise)

  • Energy distribution: Emphasizes low frequencies; rolls off ~6 dB per octave
  • Sounds like: Distant thunder, rumbling, deep waves
  • Effect: Strong masking power; suppresses deep, resonant hallucinations
  • Used in: Clarity

Mid-Band Filtered Noise

  • Energy distribution: Focused around 500 Hz – 4 kHz (speech frequencies)
  • Sounds like: Soft, speech-level interference
  • Effect: Masks hallucinated voices specifically without overloading other ranges
  • Used in: Shield
Technology

Tech Stack

The technologies behind MATHI's development and implementation create a lightweight, accessible solution that runs in any modern browser.

HTML/CSS/JavaScript

Core structure, styling, and interactivity — lightweight and cross-platform

Web Audio API

Handles real-time sound playback, GainNode control, FFT analysis, and microphone access

Local Storage

Saves session logs, relief data, and adaptive masking preferences directly in the browser

Google Fonts

Clean, modern typography matching MATHI's minimalist aesthetic

Tone.js

Advanced audio abstraction for waveform generation and envelope shaping

p5.js (Upcoming)

Real-time waveform visualizer and ambient audio graphs (upcoming visual feedback)

Thank You

From Suhan D

"Built through weeks of research, design, and experimentation — MATHI is a deeply personal, science-backed solution for those experiencing auditory hallucinations."

I would like to sincerely thank Dr. A. Sheela Ma'am for her support and guidance throughout this research journey.

Made with passion by Suhan D