Article: Mindfulness state recognition and psychological resilience enhancement model based on CNN-Mamba and causal intervention Journal: International Journal of Reasoning-based Intelligent Systems (IJRIS) 2026 Vol.18 No.20 pp.16 - 27 Abstract: Current college students face multiple psychological pressures, and psychological resilience serves as a crucial protective factor. This paper proposes a mindfulness recognition and psychological resilience enhancement model that integrates CNN-Mamba with causal reasoning. Based on experiments involving multimodal data from 240 college students, the mindfulness state recognition model achieved an accuracy rate of 89.72%, which is 17.27% higher than that of the structural causal models (SVM) (p < 0.01) and 6.15% higher than that of the long short-term memory (LSTM) (p < 0.01). After an 8-week personalised intervention, the psychological resilience of the experimental group increased by 24.36% (p < 0.01), which is approximately twice that of the intervention group (12.39%). Ablation experiments confirmed that removing the multimodal fusion and causal reasoning modules decreased the recognition accuracy by 13.41% and 5.46% (p < 0.01), respectively. This study provides a new paradigm for mental health intervention, addressing the shortcomings of traditional methods that are highly subjective and homogenising. Inderscience Publishers - linking academia, business and industry through research

Title: Mindfulness state recognition and psychological resilience enhancement model based on CNN-Mamba and causal intervention

Authors: Wenjuan Xi

Addresses: School of Medicine, Jiangxi Institute of Technology, Nanchang, 330098, China

Abstract: Current college students face multiple psychological pressures, and psychological resilience serves as a crucial protective factor. This paper proposes a mindfulness recognition and psychological resilience enhancement model that integrates CNN-Mamba with causal reasoning. Based on experiments involving multimodal data from 240 college students, the mindfulness state recognition model achieved an accuracy rate of 89.72%, which is 17.27% higher than that of the structural causal models (SVM) (p < 0.01) and 6.15% higher than that of the long short-term memory (LSTM) (p < 0.01). After an 8-week personalised intervention, the psychological resilience of the experimental group increased by 24.36% (p < 0.01), which is approximately twice that of the intervention group (12.39%). Ablation experiments confirmed that removing the multimodal fusion and causal reasoning modules decreased the recognition accuracy by 13.41% and 5.46% (p < 0.01), respectively. This study provides a new paradigm for mental health intervention, addressing the shortcomings of traditional methods that are highly subjective and homogenising.

Keywords: mindfulness intervention; psychological resilience; CNN-Mamba; personalised intervention.

DOI: 10.1504/IJRIS.2026.155777

International Journal of Reasoning-based Intelligent Systems, 2026 Vol.18 No.20, pp.16 - 27

Received: 26 Mar 2026
Accepted: 05 Jun 2026
Published online: 13 Aug 2026 *

Title: Mindfulness state recognition and psychological resilience enhancement model based on CNN-Mamba and causal intervention

Keep up-to-date