Wearable Scratch Detection for Nighttime Flare Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional assessments of atopic dermatitis, such as clinical outcome assessments and patient-reported outcomes, are subjective and episodic, lacking accuracy due to clinician variability and patient compliance issues, and fail to detect scratch events during sleep, which exacerbates pruritus and disrupts sleep, and do not predict future flare events.

Innovation Solution

A wearable sensor device uses accelerometer data and machine learning classifiers to detect scratch events, particularly during sleep, and predicts future itch and flare events, providing objective and timely interventions through computerized decision support tools.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If visual inspection and manual assessment methods are used for scratch detection, then device complexity is reduced, but measurement precision and reliability deteriorate due to subjectivity and human error

Engineering Contradiction:
Improvescratch detection precisionVSAvoiddetection system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces manual visual inspection with an automated image processing system that captures skin images, processes them through algorithms to detect scratches and measure their characteristics, and generates objective assessments. This substitution eliminates human subjectivity while maintaining operational simplicity through software-based analysis.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces an image processing system as an intermediary between the patient's skin and the assessment process. The system captures images, processes them through specialized algorithms, and provides standardized measurements, serving as an objective mediator that eliminates direct human visual assessment while keeping the overall process accessible.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If frequent monitoring is performed to improve flare prediction accuracy, then measurement precision improves, but loss of time and patient burden increase

Engineering Contradiction:
Improveflare prediction accuracyVSAvoidmonitoring time commitment
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary image capture and processing at scheduled intervals to establish baseline measurements and early warning indicators before flares occur. By analyzing trends in scratch depth, width, and number over time, the system predicts potential flares in advance, allowing intervention before full flare development while reducing the need for continuous monitoring.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements monitoring at strategically selected intervals rather than continuous observation. The system captures images at predetermined time points and uses algorithmic analysis to detect trends, providing sufficient data for accurate prediction without requiring excessive time commitment from patients or clinicians.

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If detailed imaging and analysis are performed to improve scratch characterization, then measurement precision improves, but device complexity and processing requirements increase

Engineering Contradiction:
Improvescratch characterization accuracyVSAvoidimage processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides the skin image into multiple regions and analyzes scratch characteristics separately in each region. The image processing algorithm segments scratches from surrounding skin tissue, measures individual scratch parameters (depth, width, length), and aggregates results to provide comprehensive characterization without requiring overly complex global processing.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4167838B1Computerized decision support tool and medical device for scratch detection and flare prediction
Publication Date: 2026.05.20 PFIZER INC
  • EP4167838B1 patent drawingFigure 1
  • EP4167838B1 patent drawingFigure 2
  • EP4167838B1 patent drawingFigure 3A

AI summary

Technology is disclosed for detecting scratch events and predicting flares of pruritus, utilizing motion data sensed from a wearable sensor. Detecting scratch may be done with a two-tier approach by first detecting a hand motion from motion sensed data and then classifying that hand motion as a scratch event using one or more computerized classification models. Embodiments may focus on detecting nighttime scratch by utilizing motion sensed data captured during a user's detected sleep opportunity. Additionally, historical scratch event data may be used to predict a user's itch and flare risk for a future time interval. Decision support tools in the form of computer applications or services may utilize the detected scratch events or predicted itch or flare risk to initiate an action for reducing current itch and/or mitigating future risk, including initiating a treatment protocol that includes therapeutic agent.