Machine Learning-Based Classification of Jhana Advanced Concentrative Absorption Meditation (ACAM-J) using 7T fMRI
arXiv.org February 13, 2026 Puneet Kumar, W. F. Yang, Alakhsimar Singh et al.
Functional MRI brain scans can distinguish the advanced meditation state known as Jhana (ACAM-J) from ordinary consciousness with modest accuracy. Using regional homogeneity (ReHo) maps from 20 advanced meditators, machine learning classifiers reached 66.82% accuracy. The prefrontal and anterior cingulate brain regions were most important for the classification, consistent with their roles in attention and metacognition. The results suggest that machine learning can feasibly classify advanced meditation states, supporting future work on neuromodulation and mechanistic models of such states.