Consistent Metric Learning for Outcomes of Different Measurement Tools of Cervical Spondylosis: Towards Better Therapeutic Effectiveness Evaluation
Various measuring tools have been developed to evaluate the therapeutic effectiveness of neck pain caused by cervical spondylosis (CS). However, due to different evaluation bias of these tools, it has been observed empirically that there may be some inconsistency between the outcomes of different measuring tools, leading to great challenges to precisely evaluate therapeutic effectiveness. We propose to apply a supervised metric learning algorithm to learn a metric integrating the concerned outcomes of measuring tools with least inconsistency according to the training set. Through a pair-wise constraints metric learning algorithm, the metric is expressed as a parameterized transformation. We evaluate the learnt metric and the original outcomes with three well-known learning models on the clinical data from a multi-center clinical study on acupuncture for neck pain caused by CS. The result shows that the learned integrated metric has better performance than the original outcomes.
KeywordsCervical spondylosis Metric learning Pair-wise constraints Therapeutic effectiveness evaluation
This work is supported by the 2012 College Student Career and Innovation Training Plan Project (1184512043), the 2011 Higher Education Research Fund of GDUT (2011YZ09) and Science and Technology Planning Project of Haizhu District, Guangzhou (2011-YL-05).
- 1.Hogg-Johnson S, van der Velde G, Carroll LJ, Holm LW, Cassidy JD, Guzman J et al (2008) The burden and determinants of neck pain in the general population: results of the bone and joint decade 2000–2010 task force on neck pain and its associated disorders. Spine 33(4 Suppl):S39–S51CrossRefGoogle Scholar
- 7.Zhang G, Fang J, Liang Z, Fu W, Liu J, Xu N (2011) Therapeutic effectiveness evaluation algorithm based on KCCA for neck pain caused by different diagnostic sub-types of cervical spondylosis. In: 2011 4th International conference on biomedical engineering and informatics, BMEI 2011, October 15, 2011–October 17, 2011. vol. 4 of Proceedings - 2011 4th international conference on biomedical engineering and informatics, BMEI 2011. IEEE Computer Society. p. 1794–1798Google Scholar
- 8.Di Z, Zhang HL, Zhang G, Liang ZH, Jiang L, Liu JH, et al. A clinical outcome evaluation model with local sample selection: A study on efficacy of acupuncture for cervical spondylosis. In: 2011 IEEE international conference onbioinformatics and biomedicine workshops, BIBMW 2011, November 12, 2011–November 15, 2011. 2011 IEEE international conference on bioinformatics and biomedicine workshops, BIBMW 2011. IEEE Computer Society. p. 829–833Google Scholar
- 9.Yang L, Jin R (2006) Distance metric learning: a comprehensive survey. Department of computer science and engineering, Michigan State UniversityGoogle Scholar
- 11.Lu Z, Jain P, Dhillon IS (2009) Geometry-aware metric learning. In: Proceedings of the 26th annual international conference on machine learning. ICML’09. New York, NY, USA. ACM, pp 673–680Google Scholar
- 12.Zhang Y, Zhou ZH (2009) Non-metric label propagation. In: Boutilier C, (ed). IJCAI pp 1357–1362Google Scholar
- 13.Liang Z, Zhang G, Xu S, Ou A, Fang J, Xu N, et al (2011) A kernel-decision tree based algorithm for outcome prediction on acupuncture for neck pain: A new method for interim analysis. In: BIBM Workshops. IEEE pp 760–764Google Scholar
- 14.Davis JV, Kulis B, Jain P, Sra S, Dhillon IS (2007) Information-theoretic metric learning. In: Proceedings of the 24th international conference on machine learning. ICML’07. New York, NY, USA: ACM pp 209–216Google Scholar