Automated Follicular Unit Classification via Digital Imaging

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current hair transplantation methods are inefficient in classifying and identifying follicular units, particularly in determining the number of hairs and end points, which hampers precise harvesting and transplantation, leading to potential damage and suboptimal aesthetic results.

Innovation Solution

An automated system using digital imaging and processing techniques to classify follicular units by acquiring images, segmenting them, calculating contours and outline profiles, determining defects, and tracking follicular units to accurately identify the number of hairs and end points, thereby facilitating precise transplantation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual classification methods are used to identify follicular units, then the process is simple to perform, but the precision and accuracy of classification is insufficient

Engineering Contradiction:
Improveclassification accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces manual visual inspection and mechanical classification with an automated digital imaging and image processing system. The system captures images of the scalp surface, processes them through algorithms to identify follicular unit boundaries and hair counts, and automatically classifies FUs without manual intervention, thereby improving precision while accepting increased system complexity.

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

Solution Approach 2:

The patent creates a digital copy (image) of the scalp surface and processes this copy to classify follicular units. By working with image data rather than directly manipulating physical tissue, the system achieves higher classification accuracy through detailed pixel-level analysis while keeping the actual surgical procedure separate.

Inventive Principle:
Principle #26Copying

2Measurement precision

If automated digital imaging system is implemented to classify follicular units, then the precision and accuracy of classification improves, but the device complexity increases

Engineering Contradiction:
Improvefollicular unit identification accuracyVSAvoidimaging and processing system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the image processing task into distinct functional modules: image acquisition, preprocessing (noise reduction, contrast enhancement), follicular unit boundary detection, hair strand counting, and classification. This segmentation allows each module to be optimized independently and facilitates systematic debugging and maintenance, managing the overall system complexity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent designs the digital imaging system to perform multiple functions: capturing scalp images, processing images to enhance features, identifying follicular unit boundaries, counting hairs within each unit, determining spatial orientation, and classifying FUs. This multi-functionality consolidates what would otherwise require multiple separate devices into a single integrated system.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Reliability

If traditional harvesting methods are used without precise follicular unit classification, then the harvesting process is faster, but the risk of hair transection and damage increases

Engineering Contradiction:
Improvehair graft integrityVSAvoidharvesting efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent performs follicular unit classification and identification before the actual harvesting procedure. By pre-mapping the scalp surface, counting hairs in each FU, and determining their spatial orientation in advance, the system provides a detailed guide for the harvesting process. This preliminary classification allows surgeons to harvest with higher confidence and precision, reducing the risk of transection while maintaining efficiency through pre-planning.

Inventive Principle:
Principle #10Preliminary action

4Loss of time

If manual follicular unit identification is performed, then the equipment required is simple, but the time required for classification and planning increases

Engineering Contradiction:
Improveclassification timeVSAvoidautomated system complexity
Core Design Contradiction:
Loss of timeVSDevice complexity

Solution Approach 1:

The patent implements continuous image processing and analysis that operates automatically once the imaging system is activated. The system continuously processes images to identify and classify follicular units without requiring intermittent manual intervention, thereby significantly reducing the time required for classification and planning compared to manual methods.

Inventive Principle:
Principle #20Continuity of useful action

Data Source

PatentEP2054839B1System and method for classifying follicular units
Publication Date: 2019.09.25 RESTORATION ROBOTICS INC
  • EP2054839B1 patent drawingFigure 1
  • EP2054839B1 patent drawingFigure 2
  • EP2054839B1 patent drawingFigure 3

AI summary

A system and method for classifying follicular units based on the number of hairs in a follicular unit of interest comprises acquiring an image of a body surface having a follicular unit of interest, processing the image to calculate a contour of the follicular unit and an outline profile which disregards concavities in the contour, and determining the number of defects in the outline profile to determine the number of hairs in the follicular unit. The system and method may also adjust for hairs which converge beneath the skin and for images which appear as a single wide hair but which are actually multiple hairs. In another aspect, a system and method for determining the end points of a follicular unit comprises generating a skeleton of a segmented image and identifying the end points from the skeletonized image.