Towards the Development of a Deep Learning Framework Using Adaptive and Non-Adaptive Time-Frequency Features for EEG-Based Depression Therapy Prediction.

Overview

1. Department of Information Science, University of North Texas, Denton, TX 76205, USA
2. Department of Biomedical Engineering, Science and Research Branch, Islamic Azad University, Tehran 1477893855, Iran; (S.B.); (J.F.S.)
3. School of Computer Science and Electronic Engineering, University of Essex, Colchester CO4 3SQ, UK
4. Department of Psychiatry, University of Tehran, Tehran 141556619, Iran
5. School of Computing, Gachon University, Seongnam-si 13120, Republic of Korea
6. Department of Data Science, University of North Texas, Denton, TX 76205, USA

Abstract

Background/Objectives: Predicting individual response to depression therapy prior to treatment initiation remains a critical clinical challenge, as the response rate to both selective serotonin reuptake inhibitors (SSRIs) and repetitive transcranial magnetic stimulation (rTMS) is approximately 50%, leaving treatment selection largely trial-based. This study presents a computer-aided decision (CAD) framework that predicts depression therapy outcomes from pre-treatment electroencephalogram (EEG) signals using advanced time-frequency representations and pretrained convolutional neural networks (CNNs). Methods: EEG signals from 30 SSRI patients and 46 rTMS patients are transformed into time-frequency images using Continuous Wavelet Transform (CWT), Variational Mode Decomposition (VMD), and their pixel-level fusion. Four pretrained CNN architectures, including ResNet-18, MobileNet-V3, EfficientNet-B0, and TinyViT-Hybrid, are fine-tuned and evaluated under both image-independent and subject-independent 6-fold cross-validation (CV). Results: Results reveal a clear therapy-specific pattern: CWT-based representations yield superior discrimination for SSRI outcome prediction, with ResNet-18 achieving 99.43% image-level accuracy, while VMD-based representations are statistically superior for rTMS outcome prediction, with ResNet-18 reaching 98.77%. Pixel-level fusion of CWT and VMD does not consistently improve performance over the best individual representation in either therapy context. Pairwise Wilcoxon signed-rank tests confirm a two-tier architectural hierarchy in which ResNet-18 and TinyViT-Hybrid significantly outperform MobileNet-V3 and EfficientNet-B0 across all conditions, while remaining statistically indistinguishable from each other. At the subject level, the framework achieves 82.50% and 83.53% accuracy for SSRI and rTMS, respectively, under strict subject-independent evaluation. Per-channel analysis reveals occipital dominance for SSRI under CWT and frontotemporal dominance for rTMS under VMD, consistent with known neurophysiological mechanisms. Conclusions: These findings demonstrate that the choice of time-frequency representation is therapy-specific and at least as important as architectural complexity, and that competitive performance can be achieved without recurrent or attention layers by combining well-designed spectral images with a simple pretrained residual network.

Reproduced under the paper's license (CC BY), from the paper cited above.

Code

The paper links to its data, not to its authors' code: see the Data section.

Tracing map

A tracing map links a paper to the code its authors published: this paper has none, so it has no map.

Data

Datasets cited

• figshare:4244171, at figshare; found in “Data Availability Statement”

Data Availability Statement

The SSRI dataset used in this study are publicly available on Figshare at https://figshare.com/articles/dataset/EEG_Data_New/4244171 (accessed on 1 January 2026). The rTMS dataset was collected at Atieh Hospital, Tehran, Iran, under the ethical approval of Shahid Beheshti University of Medical Sciences, and is available from the corresponding author upon reasonable request.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 30 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 9 authors, 5 keywords, 42 references.

Cite

This paper

Akbari, H., Bagherzadeh, S., Farhadi Sedehi, J., Nawaz, R., Rostami, R., Kazemi, R., Muhammad, S., Chen, H., & Mete, M. (2026). Towards the Development of a Deep Learning Framework Using Adaptive and Non-Adaptive Time-Frequency Features for EEG-Based Depression Therapy Prediction. Brain sciences, 16(3), 301. https://doi.org/10.3390/brainsci16030301

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:

Validation-Aware Retrospective EEG Treatment-Response Modelling Using Chaotic Pattern of Prime Numbers Features: Segment-Level Separability and Subject-Wise Generalisation

Journal: Bioengineering (Basel, Switzerland)

In common: figshare 4244171, depression, EEG, clinical / translational, 14 references

[2] doi:10.3390/s26103065 [code]

Subject-Wise Depression Screening from Eight-Channel Resting-State EEG Using Asymmetry-Aware Spectral Features and Connectivity Ablation.

Journal: Sensors (Basel, Switzerland)

In common: figshare 4244171, depression, EEG

[3] doi:10.3390/brainsci16080873

Evaluating Validation Strategies in Motor Imagery EEG: A Full-Cohort GAF-PLV Analysis and Matched Sensitivity Study.

Journal: Brain sciences

In common: EEG, author Hesam Akbari

[4] doi:10.3389/fnins.2026.1810609

A dual-branch network with brain region-constrained attention for EEG emotion recognition.

Journal: Frontiers in neuroscience

In common: figshare 4244171, EEG

[5] doi:10.3389/fnhum.2026.1660402

EEG-based emotion recognition using phase-space reconstruction with Poincaré sections: a study on the AMIGOS dataset.

Journal: Frontiers in human neuroscience

In common: EEG, 2 references

[6] doi:10.1038/s41398-026-03928-4

Modulatory effects of ketamine on EEG source-based resting state connectivity in treatment resistant depression.

Journal: Translational psychiatry

In common: depression, EEG, clinical / translational, 1 reference

[7] doi:10.1007/s00221-026-07368-w

The role of catastrophizing in anxiety and depression symptom scores among knee osteoarthritis patients: a multimodal assessment with EEG and machine learning.

Journal: Experimental brain research

In common: depression, EEG, clinical / translational, 1 reference

[8] doi:10.1038/s41597-026-07968-9

Atieh Schizophrenia EEG, a novel high-quality dataset designed to advance biomarker and machine learning research.

Journal: Scientific data

In common: EEG, 1 reference

[9] doi:10.1038/s41598-026-53295-9

EEG-based dynamic emotion recognition using multi-scale wavelet transform with a Spatio-Temporal neural network.

Journal: Scientific reports

In common: EEG, 1 reference

[10] doi:10.1038/s41598-026-52750-x

NeuroNetFusion: enhanced EEG abnormality classification via multi-network TF-IDF feature selection.

Journal: Scientific reports

In common: EEG, 1 reference

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.