Vibroarthrography for early detection of knee osteoarthritis using normalized frequency features
Vibroarthrography is a radiation-free and inexpensive method of assessing the condition of knee cartilage damage during extension-flexion movements. Acoustic sensors were placed on the patella and medial tibial plateau (two accelerometers) as well as on the lateral tibial plateau (a piezoelectric disk) to measure the structure-borne noise in 59 asymptomatic knees and 40 knees with osteoarthritis. After semi-automatic segmentation of the acoustic signals, frequency features were generated for the extension as well as the flexion phase. We propose simple and robust features based on relative high-frequency components. The normalized nature of these frequency features makes them insusceptible to influences on the signal gain, such as attenuation by fat tissue and variance in acoustic coupling. We analyzed their ability to serve as classification features for detection of knee osteoarthritis, including the effect of normalization and the effect of combining frequency features of all three sensors. The features permitted a distinction between asymptomatic and non-healthy knees. Using machine learning with a linear support vector machine, a classification specificity of approximately 0.8 at a sensitivity of 0.75 could be achieved. This classification performance is comparable to existing diagnostic tests and hence qualifies vibroarthrography as an additional diagnostic tool.
KeywordsVibroarthrography Cartilage degeneration Osteoarthritis Chondromalacia Non-invasive diagnosis
Compliance with Ethical Standards
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent was obtained from all individual participants included in the study.
- 1.Andersen RE, Arendt-Nielsen L, Madeleine P (2016) A review of engineering aspects of vibroarthography of the knee joint. Critical Reviews in Physical and Rehabilitation Medicine 28(1–2):13– 32Google Scholar
- 3.Beverland D, Kernohan G, McCoy G, Mollan R (1985) What is physiological patellofemoral crepitus? Med Biol Eng Comput 23(2):1249–1250Google Scholar
- 4.Bircher E (1913) Zur diagnose der meniscusluxation und des meniscusabrisses. Zentralbl f Chir 40:1852–1857Google Scholar
- 9.Carl H (1885) Grundriss der chirurgie, 3rd edn. FCW Vogel , LeipzigGoogle Scholar
- 12.Cortes C, Vapnik V (1995) Support-vector networks. Mach Learn 20(3):273–297Google Scholar
- 14.Dunbar M, Robertsson O, Ryd L, Lidgren L (2001) Appropriate questionnaires for knee arthroplasty. Bone & Joint Journal 83(3):339–344xGoogle Scholar
- 19.Hsu CW, Chang CC, Lin CJ (2010) A practical guide to support vector classification. Bioinformatics 1 (1):1–16Google Scholar
- 21.Jackson DW, Simon TM, Aberman HM (2001) Symptomatic articular cartilage degeneration: the impact in the new millennium. Clin Orthop Relat Res 391:S14–S25Google Scholar
- 23.Jiang CC, Liu YJ, Yip KM, Wu E (1993) Physiological patellofemoral crepitus in knee joint disorders. Bull Hosp Jt Dis (New York, N.Y.) 53(4):22–6Google Scholar
- 28.Lee TF, Lin WC, Wu LF, Wang HY (2012) Analysis of vibroarthrographic signals for knee osteoarthritis diagnosis. In: Proceedings - 2012 6th international conference on genetic and evolutionary computing, ICGEC 2012, pp 223–228Google Scholar
- 37.Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, Blondel M, Prettenhofer P, Weiss R, Dubourg V, Vanderplas J, Passos A, Cournapeau D, Brucher M, Perrot M, Duchesnay E (2011) Scikit-learn: machine learning in python. J Mach Learn Res 12:2825–2830Google Scholar
- 40.Platt JC (1999) Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods. In: Advances in large margin classifiers. MIT Press, Cambridge, pp 61–74Google Scholar
- 47.Reed ME, Villacis DC, Hatch GFR, Burke WS, Colletti PM, Narvy SJ, Mirzayan R, Vangsness CT (2013) 3.0-tesla MRI and arthroscopy for assessment of knee articular cartilage lesions, vol 36Google Scholar
- 51.Schölkopf B., Burges CJ (1999) Advances in kernel methods: support vector learning. MIT Press, CambridgeGoogle Scholar
- 60.Wise CH (2015) Orthopaedic manual physical therapy from art to evidence. In: Wise CH (ed). F.A. Davis Company, PhiladelphiaGoogle Scholar