GÖRSEL ANALOG SKALA ÜZERİNDEKİ OMURGA AĞRISI ŞİDDETİNİN MAKİNE ÖĞRENMESİ TABANLI OBJEKTİF TAHMİNİ

Abstract

Görsel Analog Skala (GAS), klinik uygulamalarda ağrı şiddetinin değerlendirilmesinde yaygın olarak kullanılmaktadır; ancak öznel yapısı ve sınırlı puanlama aralığı hasta kaynaklı hatalara yol açabilmektedir. Bu çalışmanın amacı, hastaların GAS kullanarak ağrı şiddetini değerlendirirken yaptıkları hata payını makine öğrenmesi tabanlı bir model ile tahmin etmektir. Araştırmada kronik omurga ağrısı bulunan 174 hastaya ait klinik veriler kullanılmıştır. Modelin eğitimi için omurga fonksiyon indeksi (OFİ), yaş, cinsiyet ve vücut kitle indeksi (VKİ) gibi objektif parametreler girdi olarak kullanılmıştır. Dinlenme ve hareket durumundaki ağrı şiddetinin tahmini amacıyla beş farklı makine öğrenmesi algoritması değerlendirilmiştir. En yüksek tahmin doğruluğu %83,50 ile Gradient Boosting algoritması ile elde edilirken, en düşük doğruluk %67,90 ile AdaBoost algoritmasıyla sağlanmıştır. Sonuçlar, hastaların GAS kullanarak ağrı şiddetini değerlendirirken %16,5 ile %32,1 arasında hata yapabildiğini göstermektedir. Önerilen yaklaşım, ağrı değerlendirmesinde daha objektif ve güvenilir bir yöntem sunmaktadır.

Keywords

• Makine Öğrenmesi
• Ağrı Şiddeti
• Görsel Analog Skala
• Tahmin Modeli
• Omurga Ağrısı.

MACHINE LEARNING-BASED OBJECTIVE ESTIMATION OF SPINAL PAIN INTENSITY ON THE VISUAL ANALOG SCALE

Abstract

The Visual Analog Scale (VAS) is widely used for assessing pain intensity in clinical practice; however, its subjective nature and limited scoring range may cause patient-related biases. This study aims to predict the error margin in pain severity estimation using the VAS through a machine learning-based model. Clinical data from 174 patients with chronic spinal pain were analyzed. Objective parameters including spinal function index (SFI), age, gender, and body mass index (BMI) were used as inputs for model training. Five machine learning algorithms; Gradient Boosting, XGBoost, AdaBoost, Support Vector Machine, and Random Forest were evaluated to estimate pain intensity at rest and during movement. The Gradient Boosting algorithm achieved the highest prediction accuracy of 83.50%, while AdaBoost showed the lowest accuracy at 67.90%. Comparison between predicted and patient-reported VAS values revealed estimation errors ranging from 16.5% to 32.1%. The findings suggest that machine learning-based approaches can provide a more objective and reliable method for pain intensity assessment, helping to reduce subjective bias in clinical evaluations.

Keywords

• Machine Learning
• Pain Intensity
• Visual Analog Scale
• Prediction Model
• Spinal Pain.

References

1. Azimi P, Yazdanian T, Benzel EC, Aghaei HN, Azhari S, Sadeghi S, et al. A Review on the Use of Artificial Intelligence in Spinal Diseases. Asian Spine J 2020;14(4):543-571.
2. Callan D, Mills L, Nott C, England R, England S. A tool for classifying individuals with chronic back pain: using multivariate pattern analysis with functional magnetic resonance imaging data. PLoS One 2014;9(6):e98007.
3. Chicco D, Warrens MJ, Jurman G. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation. Peerj computer science 2021;7:e623.
4. Coll AM, Ameen JR, Mead D. Postoperative pain assessment tools in day surgery: literature review. J Adv Nurs 2004;46(2):124-133.
5. Finlayson SG, Bowers JD, Ito J, Zittrain JL, Beam AL, Kohane IS. Adversarial attacks on medical machine learning. Science 2019;363(6433):1287-1289.
6. Galbusera F, Casaroli G, Bassani T. Artificial intelligence and machine learning in spine research. 2019;2(1):e1044.
7. Ghaffar Nia N, Kaplanoglu E, Nasab A. Evaluation of artificial intelligence techniques in disease diagnosis and prediction. Discover Artificial Intelligence 2023;3(1):5.
8. González-Fernández M, Ghosh N, Ellison T, McLeod JC, Pelletier CA, Williams K. Moving beyond the limitations of the visual analog scale for measuring pain: novel use of the general labeled magnitude scale in a clinical setting. Am J Phys Med Rehabil 2014;93(1):75-81.

9. Hamet P, Tremblay J. Artificial intelligence in medicine. Metabolism 2017;69s:S36-s40.
10. Hartmannsgruber MW, Swamidoss CP, Budde A, Qadir S, Brull SJ, Silverman DG. A method for overcoming the ceiling effect of bounded pain scales. J Clin Monit Comput 1999;15(7-8):455-459.
11. Holzinger A. Interactive machine learning for health informatics: when do we need the human-in-the-loop? Brain Inform 2016;3(2):119-131.
12. Jakkula V. Tutorial on support vector machine (svm). School of EECS, Washington State University 2006;37(2.5):3.
13. Jiang N, Luk KD, Hu Y. A Machine Learning-based Surface Electromyography Topography Evaluation for Prognostic Prediction of Functional Restoration Rehabilitation in Chronic Low Back Pain. Spine (Phila Pa 1976) 2017;42(21):1635-1642.
14. Korez R, Putzier M, Vrtovec T. A deep learning tool for fully automated measurements of sagittal spinopelvic balance from X-ray images: performance evaluation. Eur Spine J 2020;29(9):2295-2305.
15. Lamichhane B, Jayasekera D, Jakes R, Glasser MF, Zhang J, Yang C, et al. Multi-modal biomarkers of low back pain: A machine learning approach. Neuroimage Clin 2021;29:102530.
16. Lawrence R, Bunn A, Powell S, Zambon M. Classification of remotely sensed imagery using stochastic gradient boosting as a refinement of classification tree analysis. Remote Sens Environ 2004;90(3):331-336.
17. Li W, Yin Y, Quan X, Zhang H. Gene expression value prediction based on XGBoost algorithm. Front Genet. 2019; 10: 1077.
18. Lötsch J, Ultsch A. Machine learning in pain research. Pain 2018;159(4):623-630.
19. Parmar A, Katariya R, Patel V. A review on random forest: An ensemble classifier. International conference on intelligent data communication technologies and internet of things (ICICI) 2018. Springer; 2019. p. 758-763.
20. Pervaiz S, Ul-Qayyum Z, Bangyal WH, Gao L, Ahmad J. A Systematic Literature Review on Particle Swarm Optimization Techniques for Medical Diseases Detection. Computational and Mathematical Methods in Medicine 2021;2021:5990999.
21. Scott J, Huskisson EC. Graphic representation of pain. Pain 1976;2(2):175-184.
22. Sindhu BS, Shechtman O, Tuckey L. Validity, reliability, and responsiveness of a digital version of the visual analog scale. J Hand Ther 2011;24(4):356-363; quiz 364.
23. Solomatine DP, Shrestha DL. AdaBoost. RT: a boosting algorithm for regression problems. 2004 IEEE international joint conference on neural networks (IEEE Cat. No. 04CH37541). IEEE; 2004. p. 1163-1168.
24. Tesler MD, Savedra MC, Holzemer WL, Wilkie DJ, Ward JA, Paul SM. The word-graphic rating scale as a measure of children's and adolescents' pain intensity. Res Nurs Health 1991;14(5):361-371.
25. Tonga E, Gabel CP, Karayazgan S, Cuesta-Vargas AI. Cross-cultural adaptation, reliability and validity of the Turkish version of the spine functional index. Health and quality of life outcomes 2015;13:1-9.
26. Tonga E, Gabel CP, Karayazgan S, Cuesta-Vargas AI. Cross-cultural adaptation, reliability and validity of the Turkish version of the spine functional index. Health Qual Life Outcomes 2015;13:30.
27. Torrance GW, Feeny D, Furlong W. Visual analog scales: do they have a role in the measurement of preferences for health states? Med Decis Making 2001;21(4):329-334.
28. Tuyls K, Weiss G. Multiagent Learning: Basics, Challenges, and Prospects. AI Magazine 2012;33(3):41.
29. Ung H, Brown JE, Johnson KA, Younger J, Hush J, Mackey S. Multivariate classification of structural MRI data detects chronic low back pain. Cereb Cortex 2014;24(4):1037-1044.
30. Williamson A, Hoggart B. Pain: a review of three commonly used pain rating scales. J Clin Nurs 2005;14(7):798-804.