Diagnostic Classification Models Research Articles

Applying support vector machines to a diagnostic classification model for polytomous attributes in small-sample contexts.

Over several years, the evaluation of polytomous attributes in small-sample settings has posed a challenge to the application of cognitive diagnosis models. To enhance classification precision, the support vector machine (SVM) was introduced for estimating polytomous attribution, given its proven feasibility for dichotomous cases. Two simulation studies and an empirical study assessed the impact of various factors on SVM classification performance, including training sample size, attribute structures, guessing/slipping levels, number of attributes, number of attribute levels, and number of items. The results indicated that SVM outperformed the pG-DINA model in classification accuracy under dependent attribute structures and small sample sizes. SVM performance improved with an increased number of items but declined with higher guessing/slipping levels, more attributes, and more attribute levels. Empirical data further validated the application and advantages of SVMs.

Estimating the Reliability of Skill Transitions in Longitudinal Diagnostic Classification Models

Diagnostic classification models (DCMs) have grown in popularity as stakeholders increasingly desire actionable information related to students’ skill competencies. Longitudinal DCMs offer a psychometric framework for providing estimates of students’ proficiency status transitions over time. For both cross-sectional and longitudinal DCMs, it is important that researchers estimate and report reliability so stakeholders and end-users can evaluate the trustworthiness of results. Over the past decade, researchers have developed and applied various metrics for reliability in the DCM framework. This study extends these metrics onto the longitudinal DCM context and consists of three parts: (a) the theory and development of the new longitudinal DCM reliability metrics, (b) a simulation study to examine the performance of the developed metrics and establish thresholds, and (c) an empirical data analysis to illustrate an application of the developed metrics. This paper concludes with a discussion of our recommendations for applying the developed metrics.

Improving Instructional Decision‐Making Using Diagnostic Classification Models

AbstractIn recent years, educators, administrators, policymakers, and measurement experts have called for assessments that support educators in making better instructional decisions. One promising approach to measurement to support instructional decision‐making is diagnostic classification models (DCMs). DCMs are flexible psychometric models that facilitate fine‐grained reporting on skills that students have mastered. In this article, we describe how DCMs can be leveraged to support better decision‐making. We first provide a high‐level overview of DCMs. We then describe different methods for reporting results from DCM‐based assessments that support decision‐making for different stakeholder groups. We close with a discussion of considerations for implementing DCMs in an operational setting, including how they can inform decision‐making at state and local levels, and share future directions for research.

Evaluating a National Traditional Chinese Medicine Examination via Cognitive Diagnostic Approaches

The current research utilized diagnostic classification models (DCMs), an advanced psychometric theory, to evaluate the examination’s quality using psychometric methods for a more precise and comprehensive understanding of health professionals’ competence. Data was gathered from 16,310 fourth-year Traditional Chinese Medicine undergraduates who completed the Standardized Competence Test for Traditional Chinese Medicine Undergraduates (SCTTCMU) comprising 300 multiple-choice items. The study examined the fundamental assumptions, model-data fit, and cognitive diagnostic theory models’ item and test properties. The generalized deterministic input, noisy, “and” gate model applied in this research demonstrated a strong alignment with the real response data, meeting all the necessary assumptions. Cognitive diagnostic analysis indicated that all items exhibited satisfactory psychometric characteristics, and the reported scores offered insights into candidates’ proficiency in cognitive skills. It is expected that the advent of modern psychometric technology will contribute to the improvement of refined diagnostic information for health professional candidates. Furthermore, this research holds the potential to significantly enhance sustainability in healthcare practices, knowledge, economics, resource use, and community resilience.

A Comparison of Mixed and Partial Membership Diagnostic Classification Models with Multidimensional Item Response Models

Diagnostic classification models (DCM) are latent structure models with discrete multivariate latent variables. Recently, extensions of DCMs to mixed membership have been proposed. In this article, ordinary DCMs, mixed and partial membership models, and multidimensional item response theory (IRT) models are compared through analytical derivations, three example datasets, and a simulation study. It is concluded that partial membership DCMs are similar, if not structurally equivalent, to sufficiently complex multidimensional IRT models.

Detecting Alzheimer's Disease Stages and Frontotemporal Dementia in Time Courses of Resting-State fMRI Data Using a Machine Learning Approach.

Early, accurate diagnosis of neurodegenerative dementia subtypes such as Alzheimer's disease (AD) and frontotemporal dementia (FTD) is crucial for the effectiveness of their treatments. However, distinguishing these conditions becomes challenging when symptoms overlap or the conditions present atypically. Resting-state fMRI (rs-fMRI) studies have demonstrated condition-specific alterations in AD, FTD, and mild cognitive impairment (MCI) compared to healthy controls (HC). Here, we used machine learning to build a diagnostic classification model based on these alterations. We curated all rs-fMRIs and their corresponding clinical information from the ADNI and FTLDNI databases. Imaging data underwent preprocessing, time course extraction, and feature extraction in preparation for the analyses. The imaging features data and clinical variables were fed into gradient-boosted decision trees with fivefold nested cross-validation to build models that classified four groups: AD, FTD, HC, and MCI. The mean and 95% confidence intervals for model performance metrics were calculated using the unseen test sets in the cross-validation rounds. The model built using only imaging features achieved 74.4% mean balanced accuracy, 0.94 mean macro-averaged AUC, and 0.73 mean macro-averaged F1 score. It accurately classified FTD (F1 = 0.99), HC (F1 = 0.99), and MCI (F1 = 0.86) fMRIs but mostly misclassified AD scans as MCI (F1 = 0.08). Adding clinical variables to model inputs raised balanced accuracy to 91.1%, macro-averaged AUC to 0.99, macro-averaged F1 score to 0.92, and improved AD classification accuracy (F1 = 0.74). In conclusion, a multimodal model based on rs-fMRI and clinical data accurately differentiates AD-MCI vs. FTD vs. HC.

A Method of Empirical Q-Matrix Validation for Multidimensional Item Response Theory

ABSTRACT A Q-matrix is a binary matrix that defines the relationship between items and latent variables and is widely used in diagnostic classification models (DCMs), and can also be adopted in multidimensional item response theory (MIRT) models. The construction process of the Q-matrix is typically carried out by experts in the subject area of the items and statistical procedures can be used to verify its suitability. In DCMs, different approaches have been proposed for validating the Q-matrix through iterative algorithms. This article presents a method of empirical Q-matrix validation for MIRT models. A simulation study and an application to real data of morphology skills of elementary students are conducted to examine the viability of the method. Relevant issues regarding the implementation of the method and the results obtained are discussed.

Modeling Hierarchical Attribute Structures in Diagnostic Classification Models with Multiple Attempts

AbstractIn classroom assessments, examinees can often answer test items multiple times, resulting in sequential multiple‐attempt data. Sequential diagnostic classification models (DCMs) have been developed for such data. As student learning processes may be aligned with a hierarchy of measured traits, this study aimed to develop a sequential hierarchical DCM (sequential HDCM), which combines a sequential DCM with the HDCM, and investigate classification accuracy of the model in the presence of hierarchies when multiple attempts are allowed in dynamic assessment. We investigated the model's impact on classification accuracy when hierarchical structures are correctly specified, misspecified, or overspecified. The results indicate that (1) a sequential HDCM accurately classified students as masters and nonmasters when the data had a hierarchical structure; (2) a sequential HDCM produced similar or slightly higher classification accuracy than nonhierarchical sequential LCDM when the data had hierarchical structures; and (3) the misspecification of the hierarchical structure of the data resulted in lower classification accuracy when the misspecified model had fewer attribute profiles than the true model. We discuss limitations and make recommendations on using the proposed model in practice. This study provides practitioners with information about the possibilities for psychometric modeling of dynamic classroom assessment data.

Diagnostic Classification Models for Testlets: Methods and Theory.

Diagnostic classification models (DCMs) have seen wide applications in educational and psychological measurement, especially in formative assessment. DCMs in the presence of testlets have been studied in recent literature. A key ingredient in the statistical modeling and analysis of testlet-based DCMs is the superposition of two latent structures, the attribute profile and the testlet effect. This paper extends the standard testlet DINA (T-DINA) model to accommodate the potential correlation between the two latent structures. Model identifiability is studied and a set of sufficient conditions are proposed. As a byproduct, the identifiability of the standard T-DINA is also established. The proposed model is applied to a dataset from the 2015 Programme for International Student Assessment. Comparisons are made with DINA and T-DINA, showing that there is substantial improvement in terms of the goodness of fit. Simulations are conducted to assess the performance of the new method under various settings.

Ultrasound radiomics-based artificial intelligence model to assist in the differential diagnosis of ovarian endometrioma and ovarian dermoid cyst.

Accurately differentiating between ovarian endometrioma and ovarian dermoid cyst is of clinical significance. However, the ultrasound appearance of these two diseases is variable, occasionally causing confusion and overlap with each other. This study aimed to develop a diagnostic classification model based on ultrasound radiomics to intelligently distinguish and diagnose the two diseases. We collected ovarian ultrasound images from participants diagnosed as patients with ovarian endometrioma or ovarian dermoid cyst. Feature extraction and selection were performed using the Mann-Whitney U-test, Spearman correlation analysis, and the least absolute shrinkage and selection operator (LASSO) regression. We then input the final features into the machine learning classifiers for model construction. A nomogram was established by combining the radiomic signature and clinical signature. A total of 407 participants with 407 lesions were included and categorized into the ovarian endometriomas group (n = 200) and the dermoid cyst group (n = 207). In the test cohort, Logistic Regression (LR) achieved the highest area under curve (AUC) value (0.981, 95% CI: 0.963-1.000), the highest accuracy (94.8%), and the highest sensitivity (95.5%), while LightGBM achieved the highest specificity (97.1%). A nomogram incorporating both clinical features and radiomic features achieved the highest level of performance (AUC: 0.987, 95% CI: 0.967-1.000, accuracy: 95.1%, sensitivity: 88.0%, specificity: 100.0%, PPV: 100.0%, NPV: 88.0%, precision: 93.6%). No statistical difference in diagnostic performance was observed between the radiomic model and the nomogram (P > 0.05). The diagnostic indexes of radiomic model were comparable to that of senior radiologists and superior to that of junior radiologist. The diagnostic performance of junior radiologists significantly improved with the assistance of the model. This ultrasound radiomics-based model demonstrated superior diagnostic performance compared to those of junior radiologists and comparable diagnostic performance to those of senior radiologists, and it has the potential to enhance the diagnostic performance of junior radiologists.

What affects the continued learning about energy? Evidence from a 4‐year longitudinal study

AbstractEnergy is a central concept across the sciences and an important goal of science education is to support all students so that they develop a full understanding of the energy concept. However, given the abstract and complex nature of the energy concept, only a few students develop an understanding so that they can use energy ideas to make sense of phenomena. Research into energy learning progressions aims at developing models of learning about energy to guide instruction so that students can be best supported in developing competence and has provided a rich model of how students' understanding of energy develops over time. Being largely based on cross‐section data, however, the extent to which this model can guide instruction is limited, especially concerning the continued learning of students about energy. To address this gap—the limited evidence regarding what supports students' continued learning about energy—it was investigated how holding non‐normative ideas and the integratedness of students' energy knowledge affect students' continued learning about energy. Drawing on data from a 4‐year longitudinal study covering Grades 6–9 on students' learning about energy, diagnostic classification models were used to characterize students' non‐normative idea profiles and the integratedness of their knowledge and then related both to their continued learning. The results suggest no detrimental effects of holding non‐normative ideas and strong positive effects of holding integrated knowledge for students' continued learning about energy. Implications for teaching and future research are discussed.

Machine learning classification of autism spectrum disorder based on reciprocity in naturalistic social interactions

Autism spectrum disorder is characterized by impaired social communication and interaction. As a neurodevelopmental disorder typically diagnosed during childhood, diagnosis in adulthood is preceded by a resource-heavy clinical assessment period. The ongoing developments in digital phenotyping give rise to novel opportunities within the screening and diagnostic process. Our aim was to quantify multiple non-verbal social interaction characteristics in autism and build diagnostic classification models independent of clinical ratings. We analyzed videos of naturalistic social interactions in a sample including 28 autistic and 60 non-autistic adults paired in dyads and engaging in two conversational tasks. We used existing open-source computer vision algorithms for objective annotation to extract information based on the synchrony of movement and facial expression. These were subsequently used as features in a support vector machine learning model to predict whether an individual was part of an autistic or non-autistic interaction dyad. The two prediction models based on reciprocal adaptation in facial movements, as well as individual amounts of head and body motion and facial expressiveness showed the highest precision (balanced accuracies: 79.5% and 68.8%, respectively), followed by models based on reciprocal coordination of head (balanced accuracy: 62.1%) and body (balanced accuracy: 56.7%) motion, as well as intrapersonal coordination processes (balanced accuracy: 44.2%). Combinations of these models did not increase overall predictive performance. Our work highlights the distinctive nature of non-verbal behavior in autism and its utility for digital phenotyping-based classification. Future research needs to both explore the performance of different prediction algorithms to reveal underlying mechanisms and interactions, as well as investigate the prospective generalizability and robustness of these algorithms in routine clinical care.

Properties and performance of the one-parameter log-linear cognitive diagnosis model

Diagnostic classification models (DCMs) are psychometric models that yield probabilistic classifications of respondents according to a set of discrete latent variables. The current study examines the recently introduced one-parameter log-linear cognitive diagnosis model (1-PLCDM), which has increased interpretability compared with general DCMs due to useful measurement properties like sum score sufficiency and invariance properties. We demonstrate its equivalence with the Latent Class/Rasch Model and discuss interpretational consequences. The model is further examined in a DCM framework. We demonstrate the sum score sufficiency property and we derive an expression for the cut score for mastery classification. It is shown by means of a simulation study that the 1-PLCDM is fairly robust to model constraint violations in terms of classification accuracy and reliability. This robustness in combination with useful measurement properties and ease of interpretation can make the model attractive for stakeholders to apply in various assessment settings.

Knowledge Graph Enhanced Transformers for Diagnosis Generation of Chinese Medicine.

Chinese medicine (CM) diagnosis intellectualization is one of the hotspots in the research of CM modernization. The traditional CM intelligent diagnosis models transform the CM diagnosis issues into classification issues, however, it is difficult to solve the problems such as excessive or similar categories. With the development of natural language processing techniques, text generation technique has become increasingly mature. In this study, we aimed to establish the CM diagnosis generation model by transforming the CM diagnosis issues into text generation issues. The semantic context characteristic learning capacity was enhanced referring to Bidirectional Long Short-Term Memory (BILSTM) with Transformer as the backbone network. Meanwhile, the CM diagnosis generation model Knowledge Graph Enhanced Transformer (KGET) was established by introducing the knowledge in medical field to enhance the inferential capability. The KGET model was established based on 566 CM case texts, and was compared with the classic text generation models including Long Short-Term Memory sequence-to-sequence (LSTM-seq2seq), Bidirectional and Auto-Regression Transformer (BART), and Chinese Pre-trained Unbalanced Transformer (CPT), so as to analyze the model manifestations. Finally, the ablation experiments were performed to explore the influence of the optimized part on the KGET model. The results of Bilingual Evaluation Understudy (BLEU), Recall-Oriented Understudy for Gisting Evaluation 1 (ROUGE1), ROUGE2 and Edit distance of KGET model were 45.85, 73.93, 54.59 and 7.12, respectively in this study. Compared with LSTM-seq2seq, BART and CPT models, the KGET model was higher in BLEU, ROUGE1 and ROUGE2 by 6.00-17.09, 1.65-9.39 and 0.51-17.62, respectively, and lower in Edit distance by 0.47-3.21. The ablation experiment results revealed that introduction of BILSTM model and prior knowledge could significantly increase the model performance. Additionally, the manual assessment indicated that the CM diagnosis results of the KGET model used in this study were highly consistent with the practical diagnosis results. In conclusion, text generation technology can be effectively applied to CM diagnostic modeling. It can effectively avoid the problem of poor diagnostic performance caused by excessive and similar categories in traditional CM diagnostic classification models. CM diagnostic text generation technology has broad application prospects in the future.

Supervised diagnostic classification of cognitive attributes using data augmentation.

Over recent decades, machine learning, an integral subfield of artificial intelligence, has revolutionized diverse sectors, enabling data-driven decisions with minimal human intervention. In particular, the field of educational assessment emerges as a promising area for machine learning applications, where students can be classified and diagnosed using their performance data. The objectives of Diagnostic Classification Models (DCMs), which provide a suite of methods for diagnosing students' cognitive states in relation to the mastery of necessary cognitive attributes for solving problems in a test, can be effectively addressed through machine learning techniques. However, the challenge lies in the latent nature of cognitive status, which makes it difficult to obtain labels for the training dataset. Consequently, the application of machine learning methods to DCMs often assumes smaller training sets with labels derived either from theoretical considerations or human experts. In this study, the authors propose a supervised diagnostic classification model with data augmentation (SDCM-DA). This method is designed to utilize the augmented data using a data generation model constructed by leveraging the probability of correct responses for each attribute mastery pattern derived from the expert-labeled dataset. To explore the benefits of data augmentation, a simulation study is carried out, contrasting it with classification methods that rely solely on the expert-labeled dataset for training. The findings reveal that utilizing data augmentation with the estimated probabilities of correct responses substantially enhances classification accuracy. This holds true even when the augmentation originates from a small labeled sample with occasional labeling errors, and when the tests contain lower-quality items that may inaccurately measure students' true cognitive status. Moreover, the study demonstrates that leveraging augmented data for learning can enable the successful classification of students, thereby eliminating the necessity for specifying an underlying response model.

DBPF-net: dual-branch structural feature extraction reinforcement network for ocular surface disease image classification.

Pterygium and subconjunctival hemorrhage are two common types of ocular surface diseases that can cause distress and anxiety in patients. In this study, 2855 ocular surface images were collected in four categories: normal ocular surface, subconjunctival hemorrhage, pterygium to be observed, and pterygium requiring surgery. We propose a diagnostic classification model for ocular surface diseases, dual-branch network reinforced by PFM block (DBPF-Net), which adopts the conformer model with two-branch architectural properties as the backbone of a four-way classification model for ocular surface diseases. In addition, we propose a block composed of a patch merging layer and a FReLU layer (PFM block) for extracting spatial structure features to further strengthen the feature extraction capability of the model. In practice, only the ocular surface images need to be input into the model to discriminate automatically between the disease categories. We also trained the VGG16, ResNet50, EfficientNetB7, and Conformer models, and evaluated and analyzed the results of all models on the test set. The main evaluation indicators were sensitivity, specificity, F1-score, area under the receiver operating characteristics curve (AUC), kappa coefficient, and accuracy. The accuracy and kappa coefficient of the proposed diagnostic model in several experiments were averaged at 0.9789 and 0.9681, respectively. The sensitivity, specificity, F1-score, and AUC were, respectively, 0.9723, 0.9836, 0.9688, and 0.9869 for diagnosing pterygium to be observed, and, respectively, 0.9210, 0.9905, 0.9292, and 0.9776 for diagnosing pterygium requiring surgery. The proposed method has high clinical reference value for recognizing these four types of ocular surface images.

M2DC: A Meta-Learning Framework for Generalizable Diagnostic Classification of Major Depressive Disorder.

Psychiatric diseases are bringing heavy burdens for both individual health and social stability. The accurate and timely diagnosis of the diseases is essential for effective treatment and intervention. Thanks to the rapid development of brain imaging technology and machine learning algorithms, diagnostic classification of psychiatric diseases can be achieved based on brain images. However, due to divergences in scanning machines or parameters, the generalization capability of diagnostic classification models has always been an issue. We propose Meta-learning with Meta batch normalization and Distance Constraint (M2DC) for training diagnostic classification models. The framework can simulate the train-test domain shift situation and promote intra-class cohesion, as well as inter-class separation, which can lead to clearer classification margins and more generalizable models. To better encode dynamic brain graphs, we propose a concatenated spatiotemporal attention graph isomorphism network (CSTAGIN) as the backbone. The network is trained for the diagnostic classification of major depressive disorder (MDD) based on multi-site brain graphs. Extensive experiments on brain images from over 3261 subjects show that models trained by M2DC achieve the best performance on cross-site diagnostic classification tasks compared to various contemporary domain generalization methods and SOTA studies. The proposed M2DC is by far the first framework for multi-source closed-set domain generalizable training of diagnostic classification models for MDD and the trained models can be applied to reliable auxiliary diagnosis on novel data.

Identification of potential immune-related hub genes in Parkinson's disease based on machine learning and development and validation of a diagnostic classification model.

Parkinson's disease is the second most common neurodegenerative disease in the world. However, current diagnostic methods are still limited, and available treatments can only mitigate the symptoms of the disease, not reverse it at the root. The immune function has been identified as playing a role in PD, but the exact mechanism is unknown. This study aimed to search for potential immune-related hub genes in Parkinson's disease, find relevant immune infiltration patterns, and develop a categorical diagnostic model. We downloaded the GSE8397 dataset from the GEO database, which contains gene expression microarray data for 15 healthy human SN samples and 24 PD patient SN samples. Screening for PD-related DEGs using WGCNA and differential expression analysis. These PD-related DEGs were analyzed for GO and KEGG enrichment. Subsequently, hub genes (dld, dlk1, iars and ttd19) were screened by LASSO and mSVM-RFE machine learning algorithms. We used the ssGSEA algorithm to calculate and evaluate the differences in nigrostriatal immune cell types in the GSE8397 dataset. The association between dld, dlk1, iars and ttc19 and 28 immune cells was investigated. Using the GSEA and GSVA algorithms, we analyzed the biological functions associated with immune-related hub genes. Establishment of a ceRNA regulatory network for immune-related hub genes. Finally, a logistic regression model was used to develop a PD classification diagnostic model, and the accuracy of the model was verified in three independent data sets. The three independent datasets are GES49036 (containing 8 healthy human nigrostriatal tissue samples and 15 PD patient nigrostriatal tissue samples), GSE20292 (containing 18 healthy human nigrostriatal tissue samples and 11 PD patient nigrostriatal tissue samples) and GSE7621 (containing 9 healthy human nigrostriatal tissue samples and 16 PD patient nigrostriatal tissue samples). Ultimately, we screened for four immune-related Parkinson's disease hub genes. Among them, the AUC values of dlk1, dld and ttc19 in GSE8397 and three other independent external datasets were all greater than 0.7, indicating that these three genes have a certain level of accuracy. The iars gene had an AUC value greater than 0.7 in GES8397 and one independent external data while the AUC values in the other two independent external data sets ranged between 0.5 and 0.7. These results suggest that iars also has some research value. We successfully constructed a categorical diagnostic model based on these four immune-related Parkinson's disease hub genes, and the AUC values of the joint diagnostic model were greater than 0.9 in both GSE8397 and three independent external datasets. These results indicate that the categorical diagnostic model has a good ability to distinguish between healthy individuals and Parkinson's disease patients. In addition, ceRNA networks reveal complex regulatory relationships based on immune-related hub genes. In this study, four immune-related PD hub genes (dld, dlk1, iars and ttd19) were obtained. A reliable diagnostic model for PD classification was developed. This study provides algorithmic-level support to explore the immune-related mechanisms of PD and the prediction of immune-related drug targets.

Mixed Methods Investigation into Test Score Users’ Perspectives about IELTS Reading Skill Profiles

Research on the use of standardized test scores in higher education reveals significant variations in attitudes and perceptions of language proficiency tests among test score users. Most test score users have limited knowledge about test score interpretations in terms of what English as additional language (EAL) students typically know and can do at the language proficiency levels associated with admission cut scores. To address this critical gap, the language testing field has actively investigated the potential of Diagnostic Classification Models (DCMs) to offer useful information, facilitating test score users in their decision-making processes. The present two-phase mixed methods study examined the characteristics of reading skill profiles across various IELTS band scores, specifically focusing on the most frequently used admission cut scores: 6.0, 6.5, and 7.0. The study further explored test score users’ perspectives about these admission test scores, challenges encountered by EAL students, and the usefulness of reading skill profiles derived from DCMs. Findings from the application of DCMs to IELTS reading test responses (N = 5,222) showed a lack of advanced skills, such as inferential reasoning, at these commonly employed cut scores. Test score users perceived the skill profiles as instrumental in distinguishing reading abilities across various band scores and discussed the EAL students’ lack of critical reasoning, especially in inferential and synthesis tasks in coursework. The results underscore the potential of DCM-based skill profiles in providing test score users with detailed information about test score interpretations.

Neuroimaging-based classification of PTSD using data-driven computational approaches: A multisite big data study from the ENIGMA-PGC PTSD consortium

BackgroundRecent advances in data-driven computational approaches have been helpful in devising tools to objectively diagnose psychiatric disorders. However, current machine learning studies limited to small homogeneous samples, different methodologies, and different imaging collection protocols, limit the ability to directly compare and generalize their results. Here we aimed to classify individuals with PTSD versus controls and assess the generalizability using a large heterogeneous brain datasets from the ENIGMA-PGC PTSD Working group. MethodsWe analyzed brain MRI data from 3,477 structural-MRI; 2,495 resting state-fMRI; and 1,952 diffusion-MRI. First, we identified the brain features that best distinguish individuals with PTSD from controls using traditional machine learning methods. Second, we assessed the utility of the denoising variational autoencoder (DVAE) and evaluated its classification performance. Third, we assessed the generalizability and reproducibility of both models using leave-one-site-out cross-validation procedure for each modality. ResultsWe found lower performance in classifying PTSD vs. controls with data from over 20 sites (60 % test AUC for s-MRI, 59 % for rs-fMRI and 56 % for d-MRI), as compared to other studies run on single-site data. The performance increased when classifying PTSD from HC without trauma history in each modality (75 % AUC). The classification performance remained intact when applying the DVAE framework, which reduced the number of features. Finally, we found that the DVAE framework achieved better generalization to unseen datasets compared with the traditional machine learning frameworks, albeit performance was slightly above chance. ConclusionThese results have the potential to provide a baseline classification performance for PTSD when using large scale neuroimaging datasets. Our findings show that the control group used can heavily affect classification performance. The DVAE framework provided better generalizability for the multi-site data. This may be more significant in clinical practice since the neuroimaging-based diagnostic DVAE classification models are much less site-specific, rendering them more generalizable.