AÇIKLANABİLİR YAPAY ZEKA İLE EĞRİ BOYUN HASTALIĞI’NIN HASTALIK DERECELERİ VE TEDAVİLERİNİ BELİRLEME

Abstract

Cervical spine diseases, particularly neck flatness, pose significant diagnostic and treatment challenges due to the complexity of spinal structures. This study explores the application of Explainable Artificial Intelligence (XAI) techniques, specifically Random Forest and Decision Tree algorithms, to classify and assess the severity of cervical spine diseases. The dataset consists of cervical spine curvature measurements, demographic information, and clinical features. To enhance model interpretability, SHAP (Shapley Additive Explanations) and LIME (Local Interpretable Model-Agnostic Explanations) methods were integrated. These techniques provide a transparent framework for decision-making, allowing medical professionals to understand the reasoning behind AI-driven predictions. The study highlights the impact of feature selection and hyperparameter tuning on model performance, optimizing the classification process. experimental results indicate that the Random Forest algorithm achieved the highest classification accuracy at 88%, demonstrating robust predictive capabilities. The Decision Tree algorithm provided an interpretable alternative with an accuracy of 83%, enabling clear visualization of feature importance. A comparative analysis was conducted with existing literature, and findings suggest that XAI-powered models significantly improve diagnostic reliability. Additionally, application images from the dataset were incorporated into the findings section to provide a more comprehensive representation of the study. The results obtained by testing the models with independent data were also included. This research underscores the importance of integrating explainable AI into medical diagnosis, offering trustworthy, transparent, and clinically relevant insights for cervical spine disease assessment.

Keywords

• Açıklanabilir Yapay Zeka
• Random Forest
• Decision Tree
• Hiperparametre optimizasyonu
• Tıbbi görüntü analizi
• SHAP
• LIME

DETERMINING THE STAGES AND TREATMENTS OF CERVICAL SPINE DISEASES WITH EXPLAINABLE ARTIFICIAL INTELLIGENCE

Abstract

Cervical spine diseases, particularly neck flatness, pose significant diagnostic and treatment challenges due to the complexity of spinal structures. This study explores the application of Explainable Artificial Intelligence (XAI) techniques, specifically Random Forest and Decision Tree algorithms, to classify and assess the severity of cervical spine diseases. The dataset consists of cervical spine curvature measurements, demographic information, and clinical features. To enhance model interpretability, SHAP (Shapley Additive Explanations) and LIME (Local Interpretable Model-Agnostic Explanations) methods were integrated. These techniques provide a transparent framework for decision-making, allowing medical professionals to understand the reasoning behind AI-driven predictions. The study highlights the impact of feature selection and hyperparameter tuning on model performance, optimizing the classification process. experimental results indicate that the Random Forest algorithm achieved the highest classification accuracy at 88%, demonstrating robust predictive capabilities. The Decision Tree algorithm provided an interpretable alternative with an accuracy of 83%, enabling clear visualization of feature importance. A comparative analysis was conducted with existing literature, and findings suggest that XAI-powered models significantly improve diagnostic reliability. Additionally, application images from the dataset were incorporated into the findings section to provide a more comprehensive representation of the study. The results obtained by testing the models with independent data were also included. This research underscores the importance of integrating explainable AI into medical diagnosis, offering trustworthy, transparent, and clinically relevant insights for cervical spine disease assessment.

Keywords

• Explainable Artificial Intelligence
• Cervical Spine Diseases
• Random Forest
• Decision Tree
• LIME
• Medical Diagnosis
• Hyperparameter Optimization
• Neck Disorders
• Feature Importance
• SHAP

References

1. [1] Ahamed, Z. (2023). Comparative analysis of chatgpt and human decision-making in thyroid and neck swellings: a case-based study. Barw Medical Journal. https://doi.org/10.58742/bmj.v1i2.43
2. [2] Corp, N., Mansell, G., Stynes, S., Wynne‐Jones, G., Morsø, L., Hill, J., … & Windt, D. (2020). Evidence‐based treatment recommendations for neck and low back pain across europe: a systematic review of guidelines. European Journal of Pain, 25(2), 275-295. https://doi.org/10.1002/ejp.1679
3. [3] Doya, L., Doya, L., & Ghanem, A. (2022). salmonella typhi: a rare cause of neck abscess. Oxford Medical Case Reports, 2022(11). https://doi.org/10.1093/omcr/omac120
4. [4] Chinnery, T., Arifin, A., Tay, K., Leung, A., Nichols, A., Palma, D., … & Lang, P. (2020). Utilizing artificial intelligence for head and neck cancer outcomes prediction from imaging. Canadian Association of Radiologists Journal, 72(1), 73-85. https://doi.org/10.1177/0846537120942134
5. [5] Falla, D., Lindstrøm, R., Rechter, L., Boudreau, S., & Petzke, F. (2013). Effectiveness of an 8‐week exercise programme on pain and specificity of neck muscle activity in patients with chronic neck pain: a randomized controlled study. European Journal of Pain, 17(10), 1517-1528. https://doi.org/10.1002/j.1532-2149.2013.00321.x
6. [6] Fujima, N. (2023). Current state of artificial intelligence in clinical applications for head and neck mr imaging. Magnetic Resonance in Medical Sciences, 22(4), 401-414. https://doi.org/10.2463/mrms.rev.2023-0047
7. [7] Giorgini, F. (2023). Artificial intelligence in endocrinology: a comprehensive review. Journal of Endocrinological Investigation, 47(5), 1067-1082. https://doi.org/10.1007/s40618-023-02235-9
8. [8] Al-Shoteri, A. (2022). The role of methods and applications of artificial intelligence tools in the field of medicine to diagnose and discover various diseases. Journal of Applied Data Sciences, 3(1), 01-14. https://doi.org/10.47738/jads.v3i1.48

9. [9] Kim, Y., Park, J., Choi, K., Moon, B., & Lee, J. (2017). Case reports about an overlooked cause of neck pain. Medicine, 96(46), e8343. https://doi.org/10.1097/md.0000000000008343
10. [10] Barbosa, J. (2023). Effect of a telerehabilitation exercise program versus a digital booklet with self-care for patients with chronic non-specific neck pain: a protocol of a randomized controlled trial assessor-blinded, 3 months follow-up. *Trials, 24(1).* https://doi.org/10.1186/s13063-023-07651-z
11. [11] Katz, R., Leavitt, F., Cherny, K., Small, A., & Small, B. (2022). The vast majority of patients with fibromyalgia have a straight neck observed on a lateral view radiograph of the cervical spine. JCR Journal of Clinical Rheumatology. https://doi.org/10.1097/rhu.0000000000001912
12. [12] Ran, Y., Qin, W., Qin, C., Li, X., Liu, Y., Xu, L., Mu, X., Yan, L., Wang, B., Dai, Y., Chen, J., & Han, D. (2024). A high-quality dataset featuring classified and annotated cervical spine X-ray atlas. Scientific Data, 11, 625. https://doi.org/10.1038/s41597-024-03383-0
13. [13] Soellner, M., & Koenigstorfer, J. (2021). Compliance with medical recommendations depending on the use of artificial intelligence as a diagnostic method. BMC Medical Informatics and Decision Making, 21(1). https://doi.org/10.1186/s12911-021-01596-6