Derin Öğrenme Algoritmasıyla Obsesif Kompulsif Bozukluk İçin Ayırt Edici Biyobelirteç Tespiti

Yıl 2025, Cilt: 12 Sayı: 3, 75 - 80, 31.12.2025

Güneş Nazik

https://doi.org/10.32739/jnbs.12.3.277

Öz

Amaç: Obsesif Kompulsif Bozukluk (OKB), genellikle ergenlik döneminde başlayan yaygın bir psikiyatrik bozukluktur. Sıklıkla diğer psikiyatrik bozukluklarla birlikte görülmesi, semptomlarının farklı zihinsel hastalıklarla örtüşmesi ve tanının öncelikle klinik görüşmeler ve psikometrik ölçeklere dayanması, obsesif kompulsif bozukluğun tanısını zorlaştırmaktadır. Bu bağlamda, OKB’nin objektif tanı süreçlerine biyobelirteçler ve yapay zekâ destekli yaklaşımlarla katkıda bulunmak amaçlanmaktadır.
Gereç ve Yöntem: Bu çalışmada, OKB tanısı konmuş bireyler, sağlıklı bireylerden iki farklı hibrit derin öğrenme modeli kullanılarak sınıflandırılmıştır: Sırasıyla, bir boyutlu Konvolüsyonel Sinir Ağları (1DCNN) ile Kapılı Tekrarlayan Birim (GRU) ve Transformer Encoder (TE).
Bulgular: 1DCNN-TE modelinde yanlış negatif (11) ve yanlış pozitif (1) sayıları düşük seviyelerde kalırken, 1DCNN-GRU modelinde bu değerler sırasıyla 30 ve 95’tir. 1DCNN-TE modelinin eğitim ve test doğruluğu %95’in üzerinde iken, 1DCNN-GRU modelinin doğruluğu %90’ın üzerindedir. Her iki modelde de eğitim ve test kayıpları azalma eğilimi gösterirken, test kaybındaki dalgalanmalar 1DCNN-TE modelinde daha belirgindir.
Sonuç: Sonuçlar, her iki derin öğrenme modelinin EEG sinyallerine dayanarak OKB’yi yüksek doğrulukla sınıflandırabildiğini ve ayırt edici özellikleri başarılı bir şekilde öğrendiğini göstermektedir. Ancak test verilerinde gözlemlenen dalgalanmalar veya kontrol grubu tespitindeki hatalar, modellerin genellenebilirliği ve yeni veriler üzerindeki güvenilirliği açısından sınırlılıklara işaret etmiştir.

Anahtar Kelimeler

obsesif kompulsif bozukluk , derin öğrenme , elektroensefalografi , biyobelirteç

Etik Beyan

Etik kurul onayına gerek yoktur.

Teşekkür

Yazar, Nörobilim alanındaki Yüksek Lisans tezinin bir parçası olarak yürütülen bu çalışmada kullanılan araştırma altyapısını ve EEG veri setini sağladığı için Üsküdar Üniversitesi'ne teşekkür etmek ister. Yazar ayrıca tez sürecinde akademik rehberlik için Prof. Dr. Türker Tekin ERGÜZEL'e teşekkür etmek ister.

Detecting Discriminative Biomarkers For Obsessive-Compulsive Disorder Using Deep Learning Algorithms

Öz

Objective: Obsessive-Compulsive Disorder (OCD) is a common psychiatric disorder that usually begins in adolescence. The fact that it is frequently seen together with other psychiatric disorders, its symptoms overlap with different mental illnesses, and the diagnosis is primarily based on clinical interviews and psychometric scales makes it difficult to diagnose obsessive-compulsive disorder. In this context, it is aimed to contribute to the objective diagnostic processes of OCD with biomarker and artificial intelligence-supported approaches.
Materials and Methods: In this study, individuals diagnosed with OCD were classified from healthy individuals using two different hybrid deep learning models: Gated Recurrent Unit (GRU) and Transformer Encoder (TE) with one-dimensional convolutional neural networks (1DCNN), respectively.
Results: In the 1DCNN-TE model, false negatives (11) and false positives (1) remain at low levels, while in the 1DCNN-GRU model, these values are 30 and 95, respectively. While the training and test accuracy of the 1DCNN-TE model is over 95%, the accuracy of the 1DCNN-GRU model has reached over 90%. While the training and test losses tend to decrease in both models, the fluctuations in the test loss are more pronounced in the 1DCNN-TE model.
Conclusion: The results indicate that both deep learning models could classify OCD with high accuracy based on EEG signals and successfully learn discriminative features. However, the fluctuations observed in the test data and errors in detecting the control group have indicated limitations regarding the models’ generalizability and reliability on new data.

Anahtar Kelimeler

obsessive-compulsive disorder , deep learning , electroencephalography , biomarker

Etik Beyan

There is no need for ethics committee approval.

Teşekkür

The author would like to thank Üsküdar University for providing the research infrastructure and EEG dataset used in this study, which was conducted as part of the author's MSc thesis in Neuroscience. The author would also like to thank Prof. Dr. Türker Tekin ERGÜZEL for academic guidance during the thesis process.

Kaynakça

• 1. Bruin WB, Taylor L, Thomas RM, Shock JP, Zhutovsky P, Abe Y, et al. Structural neuroimaging biomarkers for obsessive-compulsive disorder in the ENIGMA-OCD consortium: medication matters. Transl Psychiatry. 2020;10(1):342. https://doi.org/10.1038/s41398- 020-01013-y
• 2. Jalal B, Chamberlain SR, Sahakian BJ. Obsessive-compulsive disorder: etiology, neuropathology, and cognitive dysfunction. Brain Behav. 2023;13(6): e3000. https://doi.org/10.1002/brb3.3000
• 3. Yılmaz B. Obsesif Kompulsif Bozukluk Tedavisinde Güncel Yaklaşımlar. Lectio Scientific. 2018;2(1):21-42.
• 4. Gonçalves ÓF, Carvalho S, Leite J, Fernandes-Gonçalves A, Carracedo A, Sampaio A. Cognitive and emotional impairments in obsessive-compulsive disorder: evidence from functional brain alterations. Porto Biomed J. 2016;1(3):92-105. https://doi.org/10.1016/j.pbj.2016.07.005
• 5. Stein DJ, Costa DL, Lochner C, Miguel EC, Reddy YJ, Shavitt RG, et al. Obsessive-compulsive disorder. Nat Rev Dis Primers. 2019;5(1):52. https://doi.org/10.1038/s41572-019-0102-3
• 6. Overduin MK, Furnham A. Assessing obsessive-compulsive disorder (OCD): a review of diagnostic interviews and clinician-rated instruments. Ann Psychiatry Ment Health. 2020;8(3).
• 7. Liu GD, Li YC, Zhang W, Zhang L. A brief review of artificial intelligence applications and algorithms for psychiatric disorders. Engineering. 2020;6(4):462-7. https://doi.org/10.1016/j.eng.2019.06.008
• 8. Farhad S, Metin SZ, Uyulan Ç, Makouei STZ, Metin B, Ergüzel TT, Tarhan N. Application of hybrid deep-learning architectures for identification of individuals with obsessive-compulsive disorder based on EEG data. Clin EEG Neurosci. 2024;55(5):543-52. https:// doi.org/10.1177/15500594231222980
• 9. Rivera MJ, Teruel MA, Mate A, Trujillo J. Diagnosis and prognosis of mental disorders using EEG and deep learning: a systematic mapping study. Artif Intell Rev. 2022;1-43. https://doi.org/10.1007/s10462-021-09986-y
• 10. Zaboski BA, Stern EF, Skosnik PD, Pittenger C.Electroencephalographic correlates and predictors of treatment outcome in OCD: a brief narrative review. Front Psychiatry. 2021;
• 12:703398. https://doi.org/10.3389/fpsyt.2021.703398
• 11. Aydin S, Arica N, Ergul E, Tan O. Classification of obsessive compulsive disorder by EEG complexity and hemispheric dependency measurements. Int J Neural Syst. 2015;25(3):1550010. https://doi.org/10.1142/S0129065715500100
• 12. Hu X, Liu Q, Li B, Tang W, Sun H, Li F, Yang Y, Gong Q, Huang X. Multivariate pattern analysis of obsessive-compulsive disorder using structural neuroanatomy. Eur Neuropsychopharmacol. 2016;26(2):246-254. https://doi:10.1016/j.euroneuro.2015.12.014
• 13. Takagi Y, Sakai Y, Lisi G, Yahata N, Abe Y, Nishida S, Nakamae T, Morimoto J, Kawato M, Narumoto J, Tanaka SC. A Neural Marker of Obsessive-Compulsive Disorder from Whole-Brain Functional Connectivity. Sci Rep. 2017;7(1):7538. https://doi:10.1038/s41598-017-07792-7
• 14. Zhu C, Fu Z, Chen L, Yu F, Zhang J, Zhang Y, Ai H, Chen L, Sui P, Wu Q, Luo Y, Xu P, Wang K. Multi-modality connectome-based predictive modeling of individualized compulsions in obsessivecompulsive disorder. J Affect Disord. 2022;311:595-603. https://doi:10.1016/j.jad.2022.05.120
• 15. Janiesch C, Zschech P, Heinrich K. Machine learning and deep learning. Electron Markets. 2021;31(3):685-95. https://doi.org/10.1007/s12525-021-00475-2
• 16. O’Shea K, Nash R. An introduction to convolutional neural networks. arXiv Preprint. 2015;arXiv:1511.08458. https://doi.org/10.48550/arXiv.1511.08458
• 17. Miotto R, Wang F, Wang S, Jiang X, Dudley JT. Deep learning for healthcare: review, opportunities and challenges. Brief Bioinform. 2018;19(6):1236–46. https://doi.org/10.1093/bib/bbx044
• 18. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015;521(7553):436–44. https://doi.org/10.1038/nature14539
• 19. Sarker IH. Deep learning: a comprehensive overview on techniques, taxonomy, applications and research directions. SN Comput Sci. 2021;2(6):1–20. https://doi.org/10.1007/s42979-021-00815-1
• 20. Zhang L, Wang S, Liu B. Deep learning for sentiment analysis: A survey. Wiley Interdiscip Rev Data Min Knowl Discov. 2018;8(4):e1253. https://doi.org/10.48550/arXiv.1801.07883
• 21. Shiri FM, Perumal T, Mustapha N, Mohamed R. A comprehensive overview and comparative analysis on deep learning models: CNN, RNN, LSTM, GRU. arXiv Preprint. 2023;arXiv:2305.17473.
• 22. https://doi.org/10.48550/arXiv.2305.17473
• 23. Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, et al. Attention is all you need. Adv Neural Inf Process Syst. 2017;30.
• 24. Lu H, Ehwerhemuepha L, Rakovski C. A comparative study on deep learning models for text classification of unstructured medical notes with various levels of class imbalance. BMC Med Res Methodol. 2022;22(1):181. https://doi.org/10.1186/s12874-022-01665-y