Research on a PTSD Risk Assessment Model Using Multi-Modal Data Fusion
Abstract
:1. Introduction
- Application and Extension of Zero-Inflated Poisson Models: Given the distributional characteristics of the PCL total score—namely, its discreteness and excessive zeros—we employ the Zero-Inflated Poisson regression framework. We construct three variants: a general ZIP model and a multi-modal ZIP model incorporating EEG-derived features.
- Development of Multi-Modal Deep Learning Models for EEG-Based PTSD Prediction: We propose and implement three multi-modal deep neural networks—Hybrid LSTM-FCNN, Hybrid RNN-FCNN, and Hybrid CNN-FCNN—which combine EEG time series data and psychological questionnaire data. These models are capable of automatically extracting high-level predictive features from EEG signals via LSTM, RNN, or CNN layers, and integrating them with questionnaire features through fully connected networks.
- Systematic Comparison of Modeling Strategies: We conduct extensive comparative experiments to evaluate the performance of the proposed deep learning models against (a) each other, (b) single-modality neural network models without EEG input, and (c) the traditional ZIP regression models. The results confirm that the Hybrid LSTM-FCNN model consistently achieves superior accuracy and generalization performance.
2. Literature Review
3. Methodology
3.1. Data Description
3.1.1. Questionnaire Data
3.1.2. EEG Data
3.2. Feature Extraction
3.2.1. Questionnaire Feature Selection
- Information Gain Calculation: To identify the most relevant questionnaire features for predicting PTSD risk, information gain (IG) was calculated for 26 variables using Python. The total PCL score was used as the response variable. Variables with IG values above the median threshold (0.0718) were retained, resulting in 13 selected features. This method enables global feature selection, capturing variables that contribute most significantly to the overall predictive model. However, IG does not reflect category-specific importance, limiting its application to local or class-specific feature selection.The algorithmic process for feature selection based on information gain is as follows:
- Calculate the information entropy of the original data: .
- Select a feature and classify the data according to its feature values. Then, calculate the information entropy for each category and compute the weighted sum to determine the information entropy of this classification method: .
- Calculate the information gain for the selected feature: .
- Repeat steps 2 and 3 to compute the information gain for all features, retaining those with the highest information gain.
- Random Forest Re-Selection: To refine the feature set obtained from the information gain method, Random Forest (RF) was employed for further feature selection, as shown in Algorithm 1. RF estimates feature importance based on the average reduction in impurity across all decision trees. The 13 variables previously selected were input into a Random Forest model implemented in Python. According to the results, the total ASD score and ASD alertness were the most important features, with the former alone contributing nearly half of the total importance. In contrast, ASD diagnosis and trauma scene emotional reaction showed minimal contributions. The top 11 features, accounting for cumulative importance of 0.996422, were retained for subsequent analysis.
Algorithm 1: Random Forest Feature Selection Algorithm |
1: Initialize , total trees. 2: for to do 3: Generate bootstrap sample from 4: Train decision tree on 5: Identify Out-Of-Bag (OOB) samples 6: Compute baseline accuracy of on 7: for each feature to do 8: Create perturbed OOB dataset by permuting feature values 9: Compute perturbed accuracy of on 10: Calculate accuracy decrease: 11: end for 12: end for 13: for each feature to do 14: Compute importance: 15: end for 16: Rank features by (descending order) |
3.2.2. EEG Data Feature Extraction
3.3. Modeling Framework
3.3.1. Zero-Inflated Poisson Regression Model
3.3.2. Deep Neural Network Architectures
Algorithm 2: Multi-Modal Data Fusion Based on Deep Learning Models |
1: Import time series and cross-sectional datasets 2: Split cross-sectional data into discrete/continuous types 3: Apply one-hot encoding to discrete data 4: Standardize continuous features 5: Combine processed discrete/continuous features 6: Separate PCL scores, Apply Box–Cox transformation [42] 7: Perform PCA dimensionality reduction on EEG data 8: Split dataset into training/testing sets (4:1 ratio) 9: Construct: - Time series models: LSTM, RNN, CNN - Cross-sectional data input model 10: Merge outputs via fully connected layer 11: Train model using fit method + EarlyStopping callback 12: Predict on test set 13: Evaluate model performance |
3.4. Multi-Modal Fusion Strategy
3.5. Model Training and Evaluation
4. Results
4.1. Results of the Zero-Inflated Poisson Model
4.2. Output of Deep Neural Network Models
4.2.1. Comparison of Multi-Modal Data Deep Learning Models
4.2.2. Ablation Experiment
4.2.3. Output Distribution of Different Deep Learning Models
4.2.4. Test of Statistical Significance of Model Performance
4.3. Results Comparison
5. Discussion
6. Conclusions
Supplementary Materials
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Kocer, D.; Farriols, N.; Cifre, I.; Nomen, M.; Lalande, M.; Calvelo, A. PTSD Among Refugee Rescue Workers: Effects of Compassion Satisfaction and Fatigue on Burnout. J. Loss Trauma 2023, 29, 421–437. [Google Scholar] [CrossRef]
- Bremner, J.D.; Hoffman, M.; Afzal, N.; Cheema, F.A.; Novik, O.; Ashraf, A.; Brummer, M.; Nazeer, A.; Goldberg, J.; Vaccarino, V. The environment contributes more than genetics to smaller hippocampal volume in Posttraumatic Stress Disorder (PTSD). J. Psychiatr. Res. 2021, 137, 579–588. [Google Scholar] [CrossRef] [PubMed]
- Chatzea, V.E.; Sifaki-Pistolla, D.; Vlachaki, S.A.; Melidoniotis, E.; Pistolla, G. PTSD, burnout, and well-being among rescue workers: Seeking to understand the impact of the European refugee crisis on rescuers. Psychiatry Res. 2018, 262, 446–451. [Google Scholar] [CrossRef] [PubMed]
- Perrin, M.A.; DiGrande, L.; Wheeler, K.; Thorpe, L.; Farfel, M.; Brackbill, R. Differences in PTSD Prevalence and Associated Risk Factors Among World Trade Center Disaster Rescue and Recovery Workers. Am. J. Psychiatry 2007, 164, 1385–1394. [Google Scholar] [CrossRef]
- Luo, T.G.; Sun, Y.; Chen, Q.; Chen, J.; Jiang, R.H.; Li, S.Z.; Sai, X.Y. Preliminary establishment of an early susceptibility screening model for post-traumatic stress disorder in rescue workers. Chin. J. Emerg. Resusc. Disaster Med. 2020, 15, 739–742+745. [Google Scholar]
- Zhou, S.L.; Jin, Y.H.; Xia, X.H.; Lu, C. Investigation of post-traumatic stress disorder symptoms among medical rescue workers affected by the Tianjin port explosion. Tianjin Nurs. 2018, 26, 184–187. [Google Scholar]
- Zhou, Z. A Study on the Symptomatology of PTSD in Chinese Fire Rescue Personnel Based on DSM-5; Institute of Psychology, Chinese Academy of Sciences, University of Chinese Academy of Sciences: Beijing, China, 2021. [Google Scholar]
- Berger, W.; Coutinho, E.S.F.; Figueira, I.; Marques-Portella, C.; Luz, M.P.; Neylan, T.C.; Marmar, C.R.; Mendlowicz, M.V. Rescuers at risk: A systematic review and meta-regression analysis of the worldwide current prevalence and correlates of PTSD in rescue workers. Soc. Psychiatry Psychiatr. Epidemiol. 2012, 47, 1001–1011. [Google Scholar] [CrossRef]
- Hyland, P.; Shevlin, M.; Brewin, C.R.; Cloitre, M.; Karatzias, T. Validation of post-traumatic stress disorder (PTSD) and complex PTSD using the International Trauma Questionnaire. Acta Psychiatr. Scand. 2017, 136, 313–322. [Google Scholar] [CrossRef]
- Shalev, A.Y.; Gevonden, M.; Ratanatharathorn, A.; Laska, E.; van der Mei, W.F.; Qi, W.; Lowe, S.; Lai, B.S.; Bryant, R.A.; Delahanty, D.; et al. Estimating the risk of PTSD in recent trauma survivors: Results of the International Consortium to Predict PTSD (ICPP). World Psychiatry 2019, 18, 77–87. [Google Scholar] [CrossRef]
- Wshah, S.; Skalka, C.; Price, M. Predicting posttraumatic stress disorder risk: A machine learning approach. JMIR Ment. Health 2019, 6, e13946. [Google Scholar] [CrossRef]
- Dupont, T.; Kentish-Barnes, N.; Pochard, F.; Azoulay, E. Prediction of post-traumatic stress disorder in family members of ICU patients: A machine learning approach. Intensive Care Med. 2024, 50, 114–124. [Google Scholar] [CrossRef] [PubMed]
- Shahzad, M.N.; Ali, H. Deep learning-based diagnosis of PTSD using 3D-CNN and resting-state fMRI data. Psychiatry Res. Neuroimaging 2024, 343, 111845. [Google Scholar] [CrossRef] [PubMed]
- Wang, J.; Ouyang, H.; Jiao, R.; Liu, W. The application of machine learning techniques in posttraumatic stress disorder: A systematic review and meta-analysis. NPJ Digit. Med. 2024, 7, 121. [Google Scholar] [CrossRef] [PubMed]
- Ge, R.; Zhou, M.; Luo, Y.; Meng, Q.; Mai, G.; Wang, G.; Zhou, F. McTwo: A two-step feature selection algorithm based on maximal information coefficient. BMC Bioinform. 2016, 17, 142. [Google Scholar] [CrossRef]
- Liu, X.; Zhou, L.; Luo, Y. Pruning deep neural network models via minimax concave penalty regression. Appl. Sci. 2024, 14, 3669. [Google Scholar] [CrossRef]
- Li, W.; Li, H.; Luo, Y. Dynamic and static enhanced BIRCH for functional data clustering. IEEE Access 2023, 11, 111448–111465. [Google Scholar] [CrossRef]
- Scherer, S.; Stratou, G.; Gratch, J.; Morency, L.-P. Investigating voice quality as a speaker-independent indicator of depression and PTSD. Proc. Interspeech 2013, 2013, 847–851. [Google Scholar] [CrossRef]
- Vilenchik, D.; Cwikel, J.; Ezra, Y.; Hausdorff, T.; Lazarov, M.; Sergienko, R.; Abramovitz, R.; Schmidt, I.; Perez, A. Method matters: Enhancing voice-based depression detection with a new data collection framework. Depress. Anxiety 2025, 2025, 4839334. [Google Scholar] [CrossRef]
- Anketell, C.; Dorahy, M.J.; Shannon, M.; Elder, R.; Hamilton, G.; Corry, M.; MacSherry, A.; Curran, D.; O’Rawe, B. An exploratory analysis of voice hearing in chronic PTSD: Potential associated mechanisms. J. Trauma Dissociation 2010, 11, 93–107. [Google Scholar] [CrossRef]
- Xu, R.; Mei, G.; Zhang, G.; Gao, P.; Judkins, T.; Cannizzaro, M.; Li, J. A voice-based automated system for PTSD screening and monitoring. Stud. Health Technol. Inform. 2012, 173, 552–558. [Google Scholar] [CrossRef]
- Gootzeit, J.; Markon, K. Factors of PTSD: Differential specificity and external correlates. Clin. Psychol. Rev. 2011, 31, 993–1003. [Google Scholar] [CrossRef] [PubMed]
- Chen, H.J.; Guo, Y.; Ke, J.; Qiu, J.; Zhang, L.; Xu, Q.; Zhong, Y.; Lu, G.M.; Qin, H.; Qi, R.; et al. Characterizing typhoon-related posttraumatic stress disorder based on multimodal fusion of structural, diffusion, and functional magnetic resonance imaging. Neuroscience 2024, 537, 141–150. [Google Scholar] [CrossRef] [PubMed]
- Liu, B.; Liu, J.; Wang, G.; Huang, K.; Li, F.; Zheng, Y.; Luo, Y.; Zhou, F. A novel electrocardiogram parameterization algorithm and its application in myocardial infarction detection. Comput. Biol. Med. 2015, 61, 178–184. [Google Scholar] [CrossRef] [PubMed]
- Gao, R.; Zhang, G.; Wang, X.; Wang, X.M.; Yu, T. Application of a multimodal neuroimaging data management system in functional neurosurgery. Chin. J. Mod. Neurol. 2024, 24, 525–531. [Google Scholar]
- Zhao, Y.; Jin, X.; Xiao, H.; Wang, Y. Application of a multimodal model combining computer vision and structured data in the referral of diabetic retinopathy. Chin. J. Digit. Med. 2024, 19, 29–35. [Google Scholar]
- Cai, C.F.; Li, J.; Jiao, Y.P.; Wang, X.X.; Guo, G.H.; Xu, J. Progress and challenges of deep learning-based multimodal data fusion methods in oncology. Data Comput. Front. 2024, 6, 3–14. [Google Scholar]
- Azam, M.A.; Khan, K.B.; Salahuddin, S.; Rehman, E.; Khan, S.A.; Khan, M.A.; Kadry, S.; Gandomi, A.H. A review on multimodal medical image fusion: Compendious analysis of medical modalities, multimodal databases, fusion techniques and quality metrics. Comput. Biol. Med. 2022, 144, 105253. [Google Scholar] [CrossRef]
- Zhou, M.; Luo, Y.; Sun, G.; Mai, G.; Zhou, F. Constraint programming based biomarker optimization. BioMed Res. Int. 2015, 2015, 910515. [Google Scholar] [CrossRef]
- Li, W.; Hu, C.; Luo, Y. A deep learning approach with extensive sentiment analysis for quantitative investment. Electronics 2023, 12, 3960. [Google Scholar] [CrossRef]
- Krüger-Gottschalk, A.; Knaevelsrud, C.; Rau, H.; Dyer, A.; Schäfer, I.; Schellong, J.; Ehring, T. The German version of the Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5): Psychometric properties and diagnostic utility. BMC Psychiatry 2017, 17, 379. [Google Scholar] [CrossRef]
- Keane, T.M.; Rubin, A.; Lachowicz, M.; Brief, D.; Enggasser, J.L.; Roy, M.; Hermos, J.; Helmuth, E.; Rosenbloom, D. Temporal stability of DSM–5 posttraumatic stress disorder criteria in a problem-drinking sample. Psychol. Assess. 2014, 26, 1138–1145. [Google Scholar] [CrossRef] [PubMed]
- Dong, X.Y.; Tao, G.Y.; Zhang, L.; Zhao, J.; Liu, J.N. Psychological crisis management for frontline nursing staff during emergencies. Chin. Med. Ethics 2021, 34, 173–177. [Google Scholar]
- Forkus, S.R.; Raudales, A.M.; Rafiuddin, H.S.; Weiss, N.H.; Messman, B.A.; Contractor, A.A. The Posttraumatic Stress Disorder (PTSD) Checklist for DSM–5: A systematic review of existing psychometric evidence. Clin. Psychol. Sci. Pract. 2023, 30, 110–121. [Google Scholar] [CrossRef] [PubMed]
- Cheng, L.; Li, D.; Li, X.; Yu, S. The optimal wavelet basis function selection in feature extraction of motor imagery electroencephalogram based on wavelet packet transformation. IEEE Access 2019, 7, 174465–174481. [Google Scholar] [CrossRef]
- Stewart, W.K.; Pretty, G.C.; Shaw, M.G.; Chase, J.G. Creating smooth SI. B-spline basis function representations of insulin sensitivity. Biomed. Signal Process. Control 2018, 44, 270–278. [Google Scholar] [CrossRef]
- Zhong, B.; Wang, P.F.; Wang, Y.Q.; Wang, X.L. A review of EEG data analysis techniques based on deep learning. J. Zhejiang Univ. Eng. Sci. 2024, 58, 879–890. [Google Scholar]
- Lambert, D. Zero-inflated Poisson regression, with an application to defects in manufacturing. Technometrics 1992, 34, 1–14. [Google Scholar] [CrossRef]
- Zhao, F.; Zhang, C.; Geng, B. Deep multimodal data fusion. ACM Comput. Surv. 2024, 56, 216. [Google Scholar] [CrossRef]
- Wang, M.H.; Fang, H.J.; Gong, H.X.; Luo, J.L. Feature extraction and recognition of EEG signals using a multi-scale hybrid convolutional network. J. Huaqiao Univ. (Nat. Sci.) 2023, 44, 628–635. [Google Scholar]
- Makhdoomi, S.M.; Rakhra, M.; Singh, D.; Singh, A. Artificial-intelligence based prediction of post-traumatic stress disorder (PTSD) using EEG reports. In Proceedings of the 2022 5th International Conference on Contemporary Computing and Informatics (IC3I), Uttar Pradesh, India, 14–16 December 2022; pp. 1073–1077. [Google Scholar] [CrossRef]
- Yingqian, L.; Yundong, T. Functional coefficient cointegration models with Box–Cox transformation. Econ. Lett. 2024, 234, 111472. [Google Scholar] [CrossRef]
- Smith, D.L.; Held, P. Moving toward precision PTSD treatment: Predicting veterans’ intensive PTSD treatment response using continuously updating machine learning models. Psychol. Med. 2023, 53, 5500–5509. [Google Scholar] [CrossRef]
Variables | Amount | Minimum Value | Maximum Value | Mean Value | Standard Deviation | Skewness | Kurtosis |
---|---|---|---|---|---|---|---|
Age | 903 | 15 | 56 | 24.21 | 3.929 | 1.779 | 6.33 |
Height | 903 | 158 | 200 | 174.19 | 5.360 | 0.303 | 0.61 |
Weight | 903 | 40 | 130 | 71.11 | 12.740 | 2.275 | 6.98 |
ASD score | 903 | 19 | 91 | 21.56 | 6.466 | 4.306 | 26.2 |
ASD isolation | 903 | 5 | 21 | 5.69 | 1.863 | 3.643 | 16.1 |
ASD re-experience | 903 | 5 | 25 | 5.57 | 1.763 | 4.769 | 30.6 |
ASD avoidance | 903 | 4 | 20 | 4.53 | 1.6 | 3.87 | 19.1 |
ASD alertness | 903 | 5 | 25 | 5.78 | 1.905 | 4.011 | 22.8 |
Mentaltoughness | 903 | 10 | 50 | 39.48 | 14.791 | −1.231 | −0.01 |
BMI | 903 | 13 | 45.35 | 23.44 | 4.18 | 2.940 | 10.6 |
PCL score | 903 | 0 | 72 | 1.40 | 5.320 | 6.776 | 61.3 |
Variables | Description |
---|---|
Gender | 1 = Male; 2 = Female |
Education level | 1 = College or above; 2 = High school; 3 = Technical secondary or below |
Marital status | 1 = Single; 2 = Married; 3 = Other |
Household monthly income per capita | 1 = <3000; 2 = 3000–4999; 3 = 5000–8000; 4 = >8000 |
Recent accident experience | 1 = No; 2 = Yes |
Family or close friend accident | 1 = No; 2 = Yes |
Witnessed severe injury | 1 = No; 2 = Yes |
Seen dead body during rescue | 1 = No; 2 = Yes |
Trauma scene emotional reaction | 1 = No; 2 = Yes |
Mental illness history | 1 = No; 2 = Yes |
Medication use in the last month | 1 = No; 2 = Yes |
Smoking history | 1 = No; 2 = Yes |
Smoking status | 1 = Non-smoker with no secondhand smoke; 2 = Non-smoker with passive smoking; 3 = Active smoker |
Alcohol consumption | 1 = No; 2 = Yes |
ASD diagnosis | 0 = Negative; 1 = Positive |
Genetic history | 1 = Negative; 2 = Positive |
Model | Hybrid LSTM-FCNN | Hybrid RNN-FCNN | Hybrid CNN-FCNN | FCNN |
---|---|---|---|---|
Trait | ||||
LSTM layers | 2 | |||
Dropout rate applied to LSTM layers | 0.5 | |||
RNN layers | 2 | |||
Dropout rate applied to RNN layers | 0.5 | |||
Number of convolutional layers | 2 | |||
Number of pooling layers | 2 | |||
Number of neurons in the first fully connected layer | 32 | 32 | 32 | 32 |
Dropout rate of the first fully connected layer | 0.5 | 0.5 | 0.5 | 0.5 |
Number of neurons in the second fully connected layer | 16 | 16 | 16 | 16 |
Dropout rate of the second fully connected layer | 0.5 | 0.5 | 0.5 | 0.5 |
Training epochs | 600 | 600 | 600 | 600 |
Batch size | 32 | 32 | 32 | 32 |
Optimizer | Adam | Adam | Adam | Adam |
Variables | Coefficient |
---|---|
Age | −0.02988 |
Weight | −1.68882 |
Height | 1.33177 |
ASD score | 0.29176 |
BMI | 5.13504 |
Mental toughness | 0.03725 |
Household monthly income per capita | −0.21238 |
ASD re-experience | −0.69806 |
Variables | Coefficient |
---|---|
ASD isolation | −3.133 |
ASD avoidance | −15.532 |
ASD alertness | −4.126 |
PC1 | 1.047 |
PC2 | 5.304 |
PC3 | 2.363 |
PC4 | 7.578 |
PC5 | −11.018 |
Deep Learning Model | MSE-Train | MSE-Test |
---|---|---|
Hybrid LSTM-FCNN | 0.0593406 | 0.1460319 |
Hybrid RNN-FCNN | 0.0477106 | 0.1650831 |
Hybrid CNN-FCNN | 0.0789325 | 0.1571726 |
Data Type | Model | MSE-Train | MSE-Test |
---|---|---|---|
Multi-modal | Hybrid LSTM-FCNN | 0.0593406 | 0.1460319 |
Hybrid RNN-FCNN | 0.0477106 | 0.1650831 | |
Hybrid CNN-FCNN | 0.0789325 | 0.1571726 | |
Ablation | CSDNN | 0.1302956 | 0.1830982 |
TS-only | 0.1738542 | 0.3044681 | |
RNN-FCNN wo/Cross | 0.0518096 | 0.1696373 |
Model Comparison | t Value | p Value |
---|---|---|
RNN-FCNN vs. CSDNN | −2.785 | 0.0496 |
CNN-FCNN vs. LSTM-FCNN | 2.891 | 0.0445 |
CNN-FCNN vs. CSDNN | −3.878 | 0.0179 |
LSTM-FCNN vs. CSDNN | −4.780 | 0.0088 |
CSDNN vs. RNN-FCNN wo/Cross | 2.287 | 0.0842 |
Data Type | Model | MSE-Train | MSE-Test |
---|---|---|---|
Single data | ZIP | 1.118069 | 1.275401 |
CSDNN | 0.1302956 | 0.1830982 | |
Multi-modal data | ZIP | 0.2814173 | 0.2971681 |
Hybrid LSTM-FCNN | 0.0593406 | 0.1460319 | |
Hybrid RNN-FCNN | 0.0477106 | 0.1650831 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Luo, Y.; Shang, Y.; Zhu, D.; Zhang, T.; Hu, C. Research on a PTSD Risk Assessment Model Using Multi-Modal Data Fusion. Mathematics 2025, 13, 1901. https://doi.org/10.3390/math13111901
Luo Y, Shang Y, Zhu D, Zhang T, Hu C. Research on a PTSD Risk Assessment Model Using Multi-Modal Data Fusion. Mathematics. 2025; 13(11):1901. https://doi.org/10.3390/math13111901
Chicago/Turabian StyleLuo, Youxi, Yucui Shang, Dongfeng Zhu, Tian Zhang, and Chaozhu Hu. 2025. "Research on a PTSD Risk Assessment Model Using Multi-Modal Data Fusion" Mathematics 13, no. 11: 1901. https://doi.org/10.3390/math13111901
APA StyleLuo, Y., Shang, Y., Zhu, D., Zhang, T., & Hu, C. (2025). Research on a PTSD Risk Assessment Model Using Multi-Modal Data Fusion. Mathematics, 13(11), 1901. https://doi.org/10.3390/math13111901