RESUMEN
<p><b>OBJECTIVE</b>To investigate the relationship between the mean CT value and bone density value and explore a rapid and accurate image segmentation method for lumbar CT values in different age groups.</p><p><b>METHODS</b>We randomly collected thin lumbar spine CT scan data from 202 patients stratified by gender and age (at the interval of 10 years). The data of each group were imported into Mimics 14.0 software to obtain independent Mask files using the new image segmentation method. The Mask files were used to fill the empty space and calculate mean, maximum, and minimum CT values.</p><p><b>RESULTS</b>In the age groups of 21-30, 31-40, 41-50, 51-60, and 61-70 years, the male subjects had mean CT values of 424.94 ± 52.23 Hu, 405.59 ± 53.60 Hu, 361.76 ± 53.50 Hu, 344.17 ± 47.76 Hu, and 332.88 ± 58.33 Hu, respectively; the values in the female group were 439.89 ± 50.99 Hu, 448.06 ± 65.89 Hu, 421.43 ± 54.74 Hu, 369.07 ± 78.27 Hu, and 304.98 ± 55.05 Hu, respectively. The mean CT value in men aged 61-70 years was significantly lower than the values in men below 50 years (P<0.05), but comparable with that in those aged 51-60 years (P>0.05); The mean CT value was significantly lower in women aged 51-70 years groups than in those aged below 50 years (P<0.05).</p><p><b>CONCLUSIONS</b>The mean CT values are correlated with the bone density values. The new image segmentation method allows rapid and accurate acquisition of the CT values in the regions of interest. The lumbar mean CT values tend to decrease with age.</p>