Plantar Basınç Dağılımı Sinyalleri Kullanılarak Erken MSlilerde Ataksinin Hybrt CNN Modelleri ile Belirlenmesi

Öz

Multipl Skleroz (MS), ataksi ve denge bozukluklarına neden olan bir merkezi sinir sistemi hastalığıdır. Atakside genellikle ilk semptom yürüyüş bozukluğu olarak görülmektedir. Yürüyüş ataksisi klinik olarak artmış çift destek süresi, kısalmış adım uzunluğu ve düzensiz adımlar ile tanımlanabilir. Bu yüzden ataksi tespitinde yürüme bozukluğunun değerlendirilmesi doğru bir yol olacaktır. Derin öğrenme çok sayıda girdi verisinden özellik çıkararak çıktı verisini tahmin eden bir makine öğrenmesi yöntemidir. Derin öğrenme nesne tanıma, sınıflandırma ve sinyal işleme gibi alanlarda sıklıkla kullanılmaktadır. Bu çalışmada plantar basınç dağılım sinyalleri içeren görüntüler kullanılarak MS’li bireyler (PwMS) için ataksi tespiti yapılması amaçlanmıştır. Bu amaçla PwMSi olan ve sağlıklı olan bireylerin plantar basınç dağılım sinyallerini içeren toplam 418 görüntü önceden eğitilmiş Hybrit CNN ağlar yardımıyla sınıflandırılmıştır. Veri setinden özellik çıkarılırken VGG16, VGG19, ResNet, MobilNet ve NasNEt derin öğrenme mimarileri kullanıldı. Daha sonra elde edilen özellik vektörleri SVM, KNN ve ANN sınıflandırıcıları kullanılarak sınıflandırıldı. Bu çalışma sonucunda en iyi sınıflandırma performansı,SVM sınıflandırıcısı ile VGG19 %85.71 Acc %81.81 Sen, %88.23 Spe derin öğrenme mimarisi kullanılarak elde edilmiştir. Yapılan bu çalışmanın yapay zeka yardımı ile PwMS’de ataksi tespitinde hekime yardımcı olacağı kanaatine varılmıştır.

Anahtar Kelimeler

• Multiple Skleroz
• Ataksi
• Derin Öğrenme
• Sınıflandırma

Determination of Ataxia With Hybrt CNN Models in Early MS Using Plantar Pressure Distribution Signals

Abstract

Multiple sclerosis (MS) is a disease of the central nervous system that causes ataxia and deficits in balance.In ataxia, the first symptom is usually seen as gait disturbance. Gait ataxia can be clinically defined by increased double support time, shortened stride length, and irregular strides. In this direction, the evaluation of deterioration in the detection of ataxia would be the right way. Deep learning is a machine learning method that predicts output data by extracting features from a large number of input data. Deep learning is frequently used in areas such as object recognition, classification and signal processing. In this study, it was aimed to detect ataxia for individuals with MS (PwMS) using images containing plantar pressure distribution signals. For this purpose, a total of 418 images containing the plantar pressure distribution signals of healthy individuals with PwMSi were classified with the help of pre-trained Hybrid CNN networks. VGG16, VGG19, ResNet, MobilNet and NasNEt deep learning architectures were used to extract features from the dataset. Then the obtained feature vectors were classified using SVM, KNN and ANN classifiers. As a result of this study, the best classification performance was obtained by using the SVM classifier and VGG19 85.71% Acc 81.81% Sen, 88.23% Spe deep learning architecture. It was concluded that this study will help the physician in the detection of ataxia in PwMS with the help of artificial intelligence.

Keywords

• Multiple Sclerosis
• Ataxia
• Deep Learning
• Classification

Kaynakça

1. McDonald, W. I., Compston, A., Edan, G., Goodkin, D., Hartung, H. P., Lublin, F. D., McFarland, H. F., Paty, D. W., Polman, C. H., Reingold, S. C., Sandberg-Wollheim, M., Sibley, W., Thompson, A., van den Noort, S., Weinshenker, B. Y., & Wolinsky, J. S. (2001). Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the diagnosis of multiple sclerosis. Annals of neurology, 50(1), 121–127. https://doi.org/10.1002/ana.1032
2. Finlayson, M., Multiple sclerosis rehabilitation: from impairment to participation. (2012): Crc Press, https://doi.org/10.1201/b12666.
3. Bethoux, F., & Bennett, S. (2011). Evaluating walking in patients with multiple sclerosis: which assessment tools are useful in clinical practice?. International journal of MS care, 13(1), 4–14. https://doi.org/10.7224/1537-2073-13.1.4
4. Givon, U., Zeilig, G., & Achiron, A. (2009). Gait analysis in multiple sclerosis: characterization of temporal-spatial parameters using GAITRite functional ambulation system. Gait & posture, 29(1), 138–142. https://doi.org/10.1016/j.gaitpost.2008.07.011
5. Nutt, J. G., Horak, F. B., & Bloem, B. R. (2011). Milestones in gait, balance, and falling. Movement disorders : official journal of the Movement Disorder Society, 26(6), 1166–1174. https://doi.org/10.1002/mds.23588
6. Heesen, C., Böhm, J., Reich, C., Kasper, J., Goebel, M., & Gold, S. M. (2008). Patient perception of bodily functions in multiple sclerosis: gait and visual function are the most valuable. Multiple sclerosis (Houndmills, Basingstoke, England), 14(7), 988–991. https://doi.org/10.1177/1352458508088916
7. Benedetti, M. G., Piperno, R., Simoncini, L., Bonato, P., Tonini, A., & Giannini, S. (1999). Gait abnormalities in minimally impaired multiple sclerosis patients. Multiple sclerosis (Houndmills, Basingstoke, England), 5(5), 363–368. https://doi.org/10.1177/135245859900500510
8. Martin, C. L., Phillips, B. A., Kilpatrick, T. J., Butzkueven, H., Tubridy, N., McDonald, E., & Galea, M. P. (2006). Gait and balance impairment in early multiple sclerosis in the absence of clinical disability. Multiple sclerosis (Houndmills, Basingstoke, England), 12(5), 620–628. https://doi.org/10.1177/1352458506070658

9. Morel, E., Allali, G., Laidet, M., Assal, F., Lalive, P. H., & Armand, S. (2017). Gait Profile Score in multiple sclerosis patients with low disability. Gait & posture, 51, 169-173.
10. Morel, E., Allali, G., Laidet, M., Assal, F., Lalive, P. H., & Armand, S. (2017). Gait Profile Score in multiple sclerosis patients with low disability. Gait & posture, 51, 169–173. https://doi.org/10.1016/j.gaitpost.2016.10.013
11. DeLisa, J. A. (Ed.). (1998). Gait analysis in the science of rehabilitation (Vol. 2). Diane Publishing.
12. LeMoyne, R., Heerinckx, F., Aranca, T., De Jager, R., Zesiewicz, T., & Saal, H. J. (2016, June). Wearable body and wireless inertial sensors for machine learning classification of gait for people with Friedreich's ataxia. In 2016 IEEE 13th International Conference on Wearable and Implantable Body Sensor Networks (BSN) (pp. 147-151). IEEE.
13. Phan, D., Nguyen, N., Pathirana, P. N., Horne, M., Power, L., & Szmulewicz, D. (2019). A random forest approach for quantifying gait ataxia with truncal and peripheral measurements using multiple wearable sensors. IEEE Sensors Journal, 20(2), 723-734.
14. Prochazka, A., Dostal, O., Cejnar, P., Mohamed, H. I., Pavelek, Z., Valis, M., & Vysata, O. (2021). Deep Learning for Accelerometric Data Assessment and Ataxic Gait Monitoring. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society, 29, 360–367. https://doi.org/10.1109/TNSRE.2021.3051093
15. Ilg, W., Seemann, J., Giese, M., Traschütz, A., Schöls, L., Timmann, D., & Synofzik, M. (2020). Real-life gait assessment in degenerative cerebellar ataxia: Toward ecologically valid biomarkers. Neurology, 95(9), e1199–e1210. https://doi.org/10.1212/WNL.0000000000010176
16. Marquer, A., Barbieri, G., & Pérennou, D. (2014). The assessment and treatment of postural disorders in cerebellar ataxia: a systematic review. Annals of physical and rehabilitation medicine, 57(2), 67–78. https://doi.org/10.1016/j.rehab.2014.01.002
17. Pan, S. J., & Yang, Q. (2009). A survey on transfer learning. IEEE Transactions on knowledge and data engineering, 22(10), 1345-1359, doi: 10.1109/TKDE.2009.191.
18. Mateen, M., Wen, J., Song, S., & Huang, Z. (2019). Fundus image classification using VGG-19 architecture with PCA and SVD. Symmetry, 11(1), 1.
19. Zagoruyko, S., & Komodakis, N. (2016). Wide residual networks. arXiv preprint arXiv:1605.07146.
20. Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., ... & Adam, H. (2017). Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861.
21. Zoph, B., Vasudevan, V., Shlens, J., & Le, Q. V. (2018). Learning transferable architectures for scalable image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 8697-8710).
22. Vapnik, V. (1998). The support vector method of function estimation. In Nonlinear modeling (pp. 55-85). Springer, Boston, MA.
23. He, C., Ma, M., & Wang, P. (2020). Extract interpretability-accuracy balanced rules from artificial neural networks: a review. Neurocomputing, 387, 346-358.