A Machine Learning Approach for Specification of Spinal Cord Injuries Using Fractional Anisotropy Values Obtained from Diffusion Tensor Images

Table 4

Results of feature evaluation using the RFS algorithms.

Ranking	Index of feature	Feature	𝑅-value
1	11	(𝐴,𝐿,𝑅)	0.060976
2	8	(𝐴,𝑅)	0.067073
3	9	(𝐿,𝑅)	0.085366
4	14	(𝑃,𝐴,𝐿,𝑅)	0.085366
5	13	(𝑃,𝐿,𝑅)	0.085366
6	12	(𝑃,𝐴,𝑅)	0.085366
7	7	(𝐴,𝐿)	0.091463
8	3	𝑅	0.091463
9	1	𝐴	0.097561
10	2	𝐿	0.103659
11	10	(𝑃,𝐴,𝐿)	0.109756
12	4	(𝑃,𝐴)	0.115854
13	6	(𝑃,𝑅)	0.146341
14	5	(𝑃,𝐿)	0.146341
15	0	𝑃	0.158537