Qianwen Hu

Qianwen Hu


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 May 2025 in Discover Computing
Verified
Journal Article
Open access
PDF file
QHQianwen Hu
Qianwen Hu
QXQiao Xu
Qiao Xu
JWJingjiao Wu
Jingjiao Wu
LZLi Zhang
Li Zhang
This paper addresses the challenge of how enterprises can achieve optimal economic benefits under the green development model. To tackle this, the study introduces an innovative digital marketing prediction model based on neural networks, specifically utilizing convolutional neural networks (CNNs) to forecast the outcomes of digital marketing strategies within the context of green development. The model integrates linear regression techniques to predict the economic impact of digital marketing efforts on enterprise performance. By incorporating green development principles into the neural network framework, the research explores how sustainability-focused strategies can enhance both marketing effectiveness and profitability. Experimental results demonstrate that the proposed model increases enterprise sales performance by 27.8%, highlighting the substantial economic benefits of combining digital marketing with green development strategies. This paper contributes a novel approach to integrating AI-driven marketing prediction with sustainable business practices, offering valuable insights into the intersection of digital innovation and green growth.

6 February 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
YXYang Xiao
Yang Xiao
QHQianwen Hu
Qianwen Hu
The unique visual characteristics of wood can evoke and regulate affect. This study quantitatively investigates the affective responses to the visual characteristics of wood radial sections based on the dimensions of valence and arousal. First, the visual characteristics of radial sections were quantified, proposing the wood color characteristic parameter (c) and density characteristic parameter (d), and defining their respective ranges. Subsequently, by controlling these visual characteristics, the study explored their influence patterns on affective valence and arousal, thereby constructing an affective response model. Finally, the model was evaluated through an eye-tracking experiment, and prediction results were correlated with eye-tracking metrics. Results indicate that the visual characteristic parameters of radial sections range within 20 < c < 80 (dark brown to light yellow) and 5 < d < 25 (dense to sparse). Lighter wood color (higher c values) and denser grain (lower d values) evoked higher valence and lower arousal, indicating emotions tending toward pleasant relaxation. The proposed model effectively predicted the correlation between sections and affects, achieving a Mean Absolute Error (MAE) of 0.2. Predicted valence and arousal values showed high correlation with pupil data, further validating the model's reliability from a physiological perspective.

12 December 2025 in Psycho-Oncologie
Verified
Journal Article
Open access
PDF file
XSXufang Sun
Xufang Sun
QHQianwen Hu
Qianwen Hu
Breast-cancer treatment often disrupts appearance and identity, leaving roughly one-third of patients in clinically significant distress. This dual-center pilot RCT assessed whether a brief, low-cost mandala-coloring program could alleviate chemotherapy-related distress in women with early-stage breast cancer. Fifty post-surgical patients starting adjuvant chemotherapy were randomized 1:1 to either three 45-minute chair-side coloring sessions delivered over three weeks or an active control involving neutral reading for the same duration; both groups received standard care. Distress (NCCN Distress Thermometer [DT] 0-10) was the primary outcome, with anxiety and depression (HADS-A/D), fatigue (BFI), and global quality of life (EORTC QLQ-C30) as secondary endpoints, all measured at baseline and one week after the final session. Feasibility targets were surpassed: 86% of approached patients enrolled, 96% completed follow-up, and 92% of intervention participants attended every session. Adjusted analyses showed lower postintervention distress in the art-therapy group (mean = 3.4) versus controls (mean = 5.1), giving a −1.7-point difference (p = 0.02, Cohen’s d = 0.85); only 28% of intervention patients, compared with 80% of controls, remained above the DT clinical cut-off (≥4). Anxiety dropped markedly in the coloring arm (8.70 to 6.20) but increased in controls (8.40 to 9.60), producing a −3.3-point adjusted difference (p < 0.001, d ≈ 1.2). Depression and fatigue showed no significant between-group differences, and global quality of life declined in controls but was largely maintained in the art group, although this trend did not reach significance. Structural-equation modelling (χ² = 7.11, CFI = 0.995, RMSEA = 0.036) confirmed a latent “distress” factor defined by DT, HADS-A, and HADS-D; the intervention exerted a significant negative effect on this factor (β = −0.55, p = 0.006), driven chiefly by anxiety reduction, with no direct paths to fatigue or quality of life. In summary, three short mandala-coloring sessions are feasible during chemotherapy and substantially reduce overall distress and anxiety, though depressive symptoms, fatigue, and broad quality-of-life indices were unchanged over this brief interval. Larger, longer trials should verify durability, explore body-image mechanisms, and determine how best to integrate chair-side coloring into routine psycho-oncology practice.