1998 Volume 41 Issue 2 Pages 97-105
Foot morphology was analyzed using methods of classification. Data concerning 18 parameters of bare feet was collected from 362 feet of subjects aged in their 20s, 50s or between 60 and79. The data were subjected to a cluster analysis by Kmeans method, using 3, 4, 5 and 6 as numbers of partitions. The results of this analysis were subjected to Kruskal's similarity assessment. This assessment, which was carried out while the number of partitions was increased, allowed us to extract morphological features of the feet, as shown below.
(1) The clusters for each age group at partition numbers of 3, 4 and 6 could be two-dimensionally represented, at a stress below 0.20. The characteristics of clusters at each of the three different partition numbers (3, 4 and 6) resembled each other. Thus, re-grouping of clusters with different partition numbers was possible.
(2) When changes in clusters of 362 feet following an increase of the partition number from 3 to 4 and 6 were analyzed, six types of change (cluster-combination types) were identified and the frequencies of these types differed among different age groups. These 6 types were seen in 70.7% of the 362 feet examined.
(3) When foot characteristics were extracted from the six change patterns mentioned above, feet for the young group were characterized by the lack of specific features in the anterior portion and the presence of specific features in the 5th toe side, and the feet for the middle and aged groups were characterized by specific shapes of the 1st toe side.
