Fractal analysis of dendrites morphology using modified Richardson's and box counting method

Area of Science:

• Neuroscience
• Biophysics
• Quantitative Morphology

Background:

• Fractal analysis is a valuable tool for studying complex natural phenomena.
• Fractal dimension quantifies object complexity, with higher values indicating greater ruggedness or winding structures.
• Applications span biology and medicine, particularly in morphologic studies.

Purpose of the Study:

• To measure the fractal dimension of neuronal dendritic arborization.
• To compare the manual ruler-based method with the computer-based box-counting method.
• To evaluate the impact of image processing (skeletonization) and orientation on measurements.

Main Methods:

• Manual implementation of Richardson's ruler-based fractal dimension measurement.
• Application of the box-counting method for fractal dimension calculation.
• Comparison of results using skeletonized and unskeletonized binary neuron images.

Main Results:

• The box-counting method yielded a significantly larger fractal dimension for unskeletonized images compared to skeletonized ones.
• The box-counting method's results were sensitive to object orientation, unlike the ruler-based method.
• The ruler-based method produced significantly smaller fractal dimension values than the box-counting method.

Conclusions:

• The ruler-based fractal analysis method is less sensitive to image processing and orientation, making it more versatile for dendritic arborization studies.
• While the box-counting method has limitations for this application, the ruler-based method, despite being manual, shows greater utility.
• Development of a computer-based implementation for the ruler-based method is needed for broader neuroscientific adoption.