Skip to main content


Table 1 Characteristics of real datasets

From: A comparison of three clustering methods for finding subgroups in MRI, SMS or clinical data: SPSS TwoStep Cluster analysis, Latent Gold and SNOB

Dataset Data type n Variables
MRI1 dataset Dichotomous, cross-sectional data 2,060 disc levels Disc signal intensity, loss of disc height, disc high intensity zone, location of high intensity zone, type of disc herniation, location of disc herniation, nucleus pulposus shape, annular tear anterior, annular tear posterior, annular tear right, annular tear left, location of nerve root compression, nerve root compression, anterolisthesis, retrolisthesis, top endplate defect, bottom endplate defect, Modic changes top endplate, Modic changes bottom endplate, facet joint degeneration, facet joint asymmetry, central stenosis, foraminal stenosis.
MRI2 dataset Dichotomous, cross- sectional data 3,155 disc levels Disc signal intensity, disc height, disc high intensity zone, disc contour, type of disc herniation, disc herniation signal intensity, anterolisthesis, retrolisthesis, type of endplate changes top, type of endplate changes bottom, size of endplate changes top, size of endplate changes bottom, osteophytes top, osteophytes bottom, endplate defect top, endplate defect bottom, endplate irregularity top, endplate irregularity bottom.
MRI3 dataset Dichotomous, cross-sectional data 20,810 disc levels Disc bulge, disc degeneration, disc herniation, disc high intensity zone, Modic changes Type 1, Modic changes Type 2, nerve root compression, Scheuermann's disease, spondylolisthesis, facet joint degeneration, osteoarthritis, central spinal stenosis, scoliosis, red flag condition (cancer, fracture, infection).
SMS dataset Interval, longitudinal repeated measures data 1,121 people Pain intensity (0 to 10) measured once a week for 52 weeks.
Clinical dataset Mixed (dichotomous, ordinal, interval), cross-sectional data 543 people Dichotomous: gender, living alone, previous episode.
Ordinal: episode duration (3 categories), STarT Back Tool subgroup (3 categories).
Interval: age (years), days of pain in last 2 weeks (0 to 14), Major Depression Inventory sum score (0 to 42), Fear Avoidance Beliefs Questionnaire subscale scores (physical activity 0 to 24, work 0 to 42), Coping Strategies Questionnaire subscale scores (divert attention 0 to 100, ignoring 0 to 100, praying or hoping 0 to 100, catastrophisation 0 to 100, reinterpreting 0 to 100).