A 3D-CNN and ICA-Guided Model for Bipolar Disorder Diagnosis via Resting-State fMRI

Noor Ayesha; Roaa Khalil Mohamed Ali Abed; Saeed Ali Bahaj

doi:10.48084/etasr.18633

Authors

Noor Ayesha CYBEX, Prince Sultan University, Riyadh, Saudi Arabia
Roaa Khalil Mohamed Ali Abed College of Sciences and Humanities (CSH), Prince Sultan University, Riyadh, Saudi Arabia
Saeed Ali Bahaj MIS Department College of Business Administration, Prince Sattam Bin Abdulaziz University, Al‑Kharj, Saudi Arabia

Volume: 16 | Issue: 3 | Pages: 35981-35987 | June 2026 | https://doi.org/10.48084/etasr.18633

Received: 10 March 2026 | Revised: 11 April 2026 | Accepted: 17 April 2026 | Online: 6 June 2026

Corresponding author: Saeed Ali Bahaj

Abstract

Bipolar Disorder (BD) is a persistent mental health condition marked by intense mood fluctuations. These mood episodes range from emotional highs (mania or hypomania) to lows (depression), significantly affecting an individual's energy, activity levels, sleep patterns, behavior, and cognitive clarity. This study utilizes resting-state functional MRIs (rs-fMRI) to capture detailed structural and functional insights into brain tissue samples. Independent Component Analysis (ICA) was applied to rs-fMRI scans of 45 Healthy Controls (HCs) and 45 subjects with BD (BDs) from the OpenNeuro database to identify significant features. The top five features (IC 12, IC 15, IC 16, IC 18, and IC 22) were selected based on kurtosis, skewness, and variability, and then used as input to a 3D Convolutional Neural Network (3D-CNN) model for BD diagnosis. Of 90×5=450 independent components, the model was trained on 70% (315) and tested on the remaining 30% (135). Evaluation metrics confirmed high performance in distinguishing BD cases from HCs, achieving 94% accuracy. This approach demonstrates a reliable, high-accuracy method for diagnosing BD using rs-fMRIs, offering insights into associated functional activation patterns and neurobiological mechanisms. This multimodal approach could improve clinicians' diagnostic confidence and advance understanding of the underlying pathology of BD.

Keywords:

bipolar disorder, rsfMRI, Independent Component Analysis (ICA), deep learning, health risks

References

S. Z. Kurdi, "Machine Learning–Based Classification Framework for Human Health Care Monitoring," International Journal of Theoretical & Applied Computational Intelligence, vol. 2026, pp. 1–15, Jan. 2026.

M. S. Shobini and M. Sudha, "BiPoP: Bipolar Disorder Optimized Preprocessing Framework for Stress Disorder Identification through Gene Expression Data using Deep Learning," Engineering, Technology & Applied Science Research, vol. 15, no. 2, pp. 22126–22130, Apr. 2025.

T. Saba, A. Rehman, M. N. Shahzad, R. Latif, S. A. Bahaj, and J. Alyami, "Machine learning for post-traumatic stress disorder identification utilizing resting-state functional magnetic resonance imaging," Microscopy Research and Technique, vol. 85, no. 6, pp. 2083–2094, 2022.

H. Ali, "A Meta-Review of Computational Intelligence Techniques for Early Autism Disorder Diagnosis," International Journal of Theoritical & Applied Computational Intelligence, pp. 01–21, 2025.

H. K. Aljobouri, "Independent Component Analysis with Functional Neuroscience Data Analysis," Journal of Biomedical Physics and Engineering, vol. 13, no. 2, Apr. 2023.

A. Y. Abdzaid, D. N. Khadum, S. S. Al-Barrak, and A. Saad, "White Blood Cells Classification Based on a New Strategy to Cluster Channels RGB of Blood Smear Image Using Resnet-18 Model," International Journal of Theoretical & Applied Computational Intelligence, vol. 2026, pp. 120–131, Apr. 2026.

B. Shehu Aliyu, J. Isuwa, A. Abdulrahim, M. Abdullahi, I. Hayatu Hassan, and T. Sikder Momi, "Enhanced Feature Selection for Imbalanced Microarray Cancer Gene Classification Using Chaotic Salp Swarm Algorithm," International Journal of Theoritical & Applied Computational Intelligence, pp. 259–283, 2025.

J. Sawalha et al., "Individualized identification of first-episode bipolar disorder using machine learning and cognitive tests," Journal of Affective Disorders, vol. 282, pp. 662–668, Mar. 2021.

P. Jakobsen et al., "Applying machine learning in motor activity time series of depressed bipolar and unipolar patients compared to healthy controls," PLOS ONE, vol. 15, no. 8, Aug. 2020, Art. no. e0231995.

C. Saylan and K. Yilancioglu, "Classification of schizophrenia and bipolar disorder by using machine learning algorithms," The Journal of Neurobehavioral Sciences, vol. 3, no. 3, pp. 92–95, 2016.

M. Chen et al., "Distinguishing schizophrenia and bipolar disorder through a Multiclass Classification model based on multimodal neuroimaging data," Journal of Psychiatric Research, vol. 172, pp. 119–128, Apr. 2024.

S. Poletti et al., "A peripheral inflammatory signature discriminates bipolar from unipolar depression: A machine learning approach," Progress in Neuro-Psychopharmacology and Biological Psychiatry, vol. 105, Mar. 2021, Art. no. 110136.

Turker Tekin Erguzel, Gokben Hizli Sayar, and N. Tarhan, "Artificial Intelligence Approach to Classify Unipolar and Bipolar Depressive Disorders." ResearchGate, 2015.

R. Dou et al., "Machine learning algorithm performance evaluation in structural magnetic resonance imaging-based classification of pediatric bipolar disorders type I patients," Frontiers in Computational Neuroscience, vol. 16, Aug. 2022, Art. no. 915477.

X. Jiang et al., "Identifying misdiagnosed bipolar disorder using support vector machine: feature selection based on fMRI of follow-up confirmed affective disorders," Translational Psychiatry, vol. 14, no. 1, Jan. 2024, Art. no. 9.

N. F. Jie et al., "Discriminating Bipolar Disorder From Major Depression Based on SVM-FoBa: Efficient Feature Selection With Multimodal Brain Imaging Data," IEEE Transactions on Autonomous Mental Development, vol. 7, no. 4, pp. 320–331, Sept. 2015.

M. Li et al., "Clinical utility of a short resting‐state MRI scan in differentiating bipolar from unipolar depression," Acta Psychiatrica Scandinavica, vol. 136, no. 3, pp. 288–299, Sept. 2017.

N. F. Jie et al., "Discriminating Bipolar Disorder from Major Depression using Whole-Brain Functional Connectivity: a Feature Selection Analysis with SVM-FoBa Algorithm," Journal of Signal Processing Systems, vol. 90, no. 2, pp. 259–271, Feb. 2018.

R. Bilder et al., "UCLA Consortium for Neuropsychiatric Phenomics LA5c Study." Openneuro, 2020.

"CONN toolbox." https://web.conn-toolbox.org/.

H. A. Jaber, H. K. Aljobouri, İ. Çankaya, O. M. Koçak, and O. Algin, "Preparing fMRI Data for Postprocessing: Conversion Modalities, Preprocessing Pipeline, and Parametric and Nonparametric Approaches," IEEE Access, vol. 7, pp. 122864–122877, 2019.

Y. Behzadi, K. Restom, J. Liau, and T. T. Liu, "A component based noise correction method (CompCor) for BOLD and perfusion based fMRI," NeuroImage, vol. 37, no. 1, pp. 90–101, Aug. 2007.

M. N. Shahzad and H. Ali, "Deep learning based diagnosis of PTSD using 3D-CNN and resting-state fMRI data," Psychiatry Research: Neuroimaging, vol. 343, Sept. 2024, Art. no. 111845.

S. Nasir, M. Bilal, and H. Khalidi, "Detection and Classification of Skin Cancer by using CNN-Enabled Cloud Storage Data Access Control Algorithm based on Blockchain Technology," International Journal of Theoritical & Applied Computational Intelligence, pp. 146–159, 2025.

M. N. Shahzad, H. Ali, T. Saba, A. Rehman, H. Kolivand, and S. A. Bahaj, "Identifying Patients With PTSD Utilizing Resting-State fMRI Data and Neural Network Approach," IEEE Access, vol. 9, pp. 107941–107954, 2021.

M. H. Saghayan, M. H. Zolfagharnasab, A. Khadem, F. Matinfar, and H. Rashidi, "Diagnosing Bipolar Disorder from 3-D Structural Magnetic Resonance Images Using a Hybrid GAN-CNN Method." arXiv, 2023.