Connectivity-based Cortical Parcellation via Contrastive Learning on Spatial-Graph Convolution

Objective.Objective of this work is the development and evaluation of a cortical parcellation framework based on tractography-derived brain structural connectivity.Impact Statement.The proposed framework utilizes novel spatial-graph representation learning methods for solving the task of cortical parcellation,an important medical image analysis and neuroscientific problem.Introduction.The concept of“connectional fingerprint”has motivated many investigations on the connectivity-based cortical parcellation,especially with the technical advancement of diffusion imaging.Previous studies on multiple brain regions have been conducted with promising results.However,performance and applicability of these models are limited by the relatively simple computational scheme and the lack of effective representation of brain imaging data.Methods.We propose the Spatial-graph Convolution Parcellation(SGCP)framework,a two-stage deep learning-based modeling for the graph representation brain imaging.In the first stage,SGCP learns an effective embedding of the input data through a self-supervised contrastive learning scheme with the backbone encoder of a spatial-graph convolution network.In the second stage,SGCP learns a supervised classifier to perform voxel-wise classification for parcellating the desired brain region.Results.SGCP is evaluated on the parcellation task for 5 brain regions in a 15-subject DWI dataset.Performance comparisons between SGCP,traditional parcellation methods,and other deep learning-based methods show that SGCP can achieve superior performance in all the cases.Conclusion.Consistent good performance of the proposed SGCP framework indicates its potential to be used as a general solution for investigating the regional/subregional composition of human brain based on one or more connectivity measurements.

Structure for energy cycle: a unique status of the second law of thermodynamics for living systems

Distinguishing things from beings, or matters from lives, is a fundamental question. Extending E. Schr?dinger's neg-entropy and I. Prigogine's dissipative structure, we propose a chemical kinetic view that the earliest "live" process is embedded essentially in a special interaction between a pair of specific components under a particular, corresponding environmental conditions. The interaction exists as an inter-molecular-force-bond complex(IMFBC) that couples two separate chemical processes: one is the spontaneous formation of the IMFBC driven by a decrease of Gibbs free energy as a dissipative process; while the other is the disassembly of the IMFBC driven thermodynamically by free energy input from the environment. The two chemical processes coupled by the IMFBC originated independently and were considered non-living on Earth, but the IMFBC coupling of the two can be considered as the earliest form of metabolism: the first landmark on the path from things to a being. The dynamic formation and disassembly of the IMFBC, as a composite individual, follows a principle designated as "… structure for energy for structure for energy…", the cycle continues; and for short it will be referred to as "structure for energy cycle". With additional features derived from this starting point, the IMFBC-centered "live" process spontaneously evolved into more complex living organisms with the characteristics currently known.

MR-seq:measuring a single cell's transcriptome repeatedly by RNA-seq

We described a novel single-cell RNA-seq technique called MR-seq(measure a single-cell transcriptome repeatedly),which permits statistically assessing the technical variation and identifying the differentially expressed genes between just two single cells by measuring each single cell twice. We demonstrated that MR-seq gave sensitivity and reproducibility similar to the standard single-cell RNA-seq and increased the positive predicate value. Application of MR-seq to early mouse embryos identified hundreds of candidate intra-embryonic heterogeneous genes among mouse 2-,4-and 8-cell stage embryos. MR-seq should be useful for detecting differentially expressed genes among a small number of cells.