A DICCCOL-based K-nearest landmark detection method for identifying common and consistent 3-hinge gyral folding landmarks

3-hinge gyral folding pattern is one of the most important brain architectures in brain imaging analysis field, which provides an anatomical profile to better understand the working mechanism of the brain. It is designed to further describe the gyri/sulci pattern of the cortical folding and to identify the significant regions on the gyral regions. Due to the huge variability of individual brains, it is tough to identify the consistent 3-hinge regions across the subjects, as well as the 3-hinge gyral folding patterns across different subjects. In order to study the consistency of the 3-hinge regions, in this paper, we propose a DICCCOL-based K-nearest landmark detection method for identifying the common and consistent 3-hinge gyral folding landmarks. Given the superiority of our previous DICCCOL system which provides dense and individualized common and consistent cortical landmarks across the subjects, it is therefore utilized in this work to identify the 3-hinge gyral landmarks. We successfully identified 79 common and consistent 3-hinge landmarks across individual brains, which exhibited fundamental differences with those inconsistent 3-hinge landmarks. Specifically, those consistent 3-hinges are more related to the functions of movement, sensation, memory and decision, while those inconsistent ones are more related to language, audio processing and visual. The comparison studies between DICCCOL system and the widely used brain landmark atlas further confirmed the superiority of the proposed DICCCOL-based method in identifying 3-hinge landmarks. Our work provided a novel and uniform platform to investigate the brain functions.