Data-Driven Clustering of Plantar Thermal Patterns in Healthy Individuals: An Insole-Based Approach to Foot Health...

Area of Science:

• Biomechanics
• Wearable Technology
• Thermal Imaging

Background:

• Plantar foot temperature monitoring is vital for foot health, especially for individuals with diabetes.
• Traditional thermography in controlled settings doesn't reflect real-world, shod foot conditions during activity.
• Wearable insole sensors offer continuous, dynamic temperature monitoring in everyday situations.

Purpose of the Study:

• To explore normative thermal patterns in shod feet using insole sensors and a data-driven clustering approach.
• To compare in-shoe temperature patterns with traditional thermographic findings.
• To assess the potential of in-shoe sensors for real-world foot health monitoring.

Main Methods:

• Collected temperature data from 27 healthy participants using insoles with 21 sensors.
• Applied data-driven clustering algorithms (k-means, fuzzy c-means, OPTICS, hierarchical clustering) to analyze thermal patterns.
• Identified thermal patterns without pre-defined foot regions or models.

Main Results:

• Six primary thermal patterns were identified, with the 'butterfly pattern' (elevated medial arch) being most common (51.5%).
• Clustering algorithms showed high similarity, with variations mainly due to granularity.
• In-shoe data reflected some normative patterns seen in thermography but captured dynamic, real-world distributions.

Conclusions:

• In-shoe temperature monitoring provides a more accurate representation of foot thermal behavior during daily activities compared to traditional methods.
• The proposed system captures dynamic thermal variations, offering enhanced insights into foot health in ambulatory conditions.
• This approach holds potential for improved monitoring and early detection of foot complications, particularly in at-risk populations.