Reference resource

Lumbar reference atlas, with the spread

The figures a surgeon consults for vertebral body height, canal dimensions and pedicle width come from cadaveric series of a few dozen specimens, plotted as a single line per measurement with no indication of spread. They are good numbers. What they cannot tell you is whether the patient in front of you is unusual, because they never show what usual looks like.

These are the same measurements from 802 CT records, in the same axes — level on the vertical, measurement on the horizontal — so the two can be laid side by side. Every row carries its median, its interquartile range, its 5th–95th percentile, and its own n.

Six panels of lumbar and lower thoracic morphometry against vertebral level, each row showing a median marker, an interquartile bar and a 5th to 95th percentile line: body height ventral and dorsal, superior endplate width, canal width and depth, transverse pedicle width, disc height by interspace, and trabecular attenuation. — Marker, median; bar, interquartile range; line, 5th–95th percentile; the number at the right of a row is its n. Coverage is not uniform and is drawn rather than assumed: canal depth and pedicle width rest on 330–656 records because they need an axial extent many field-limited series do not carry, while the remaining measures rest on 720–784.

Download the values (CSV) Figure as vector PDF About CTSpinoPelvic1K

What these numbers are, and what they are not. This is a screening cohort aged 50 and over, scanned supine, measured automatically from segmentations. Posture matters for anything postural, age matters for bone, and an automatic measurement is not a radiologist. So this describes that population. It is not a normative atlas of the healthy adult spine, and it should not be used as one. No comparison between levels is tested here and none is implied — the intervals are dispersion, not inference.

The values

Generated from the same file the figure is drawn from, so the table and the plot cannot disagree.

Vertebral body height, ventral (mm)

Level	n	Median	IQR	5th–95th	Mean ± SD
T11	731	25.60	24.00–27.00	21.60–29.00	25.30 ± 2.48
T12	744	27.20	25.60–28.80	24.00–30.40	27.16 ± 2.25
L1	740	28.80	28.00–30.40	25.60–32.80	29.18 ± 2.80
L2	722	30.40	28.80–31.20	27.20–33.60	30.19 ± 2.40
L3	757	30.40	29.60–32.00	28.00–34.40	30.85 ± 2.12
L4	772	31.20	29.60–32.00	28.00–34.40	30.95 ± 2.20
L5	764	32.00	30.40–33.60	28.80–36.00	32.00 ± 2.64

Vertebral body height, dorsal (mm)

Level	n	Median	IQR	5th–95th	Mean ± SD
T11	731	26.40	25.60–28.00	23.20–30.40	26.82 ± 2.41
T12	744	28.80	27.20–30.40	24.80–32.80	28.65 ± 2.35
L1	739	30.40	28.80–31.20	26.40–33.60	30.14 ± 2.52
L2	721	30.40	28.80–32.00	27.20–33.60	30.31 ± 2.07
L3	757	30.40	28.80–31.20	27.20–33.60	30.09 ± 2.04
L4	771	29.60	28.00–30.40	26.40–33.60	29.46 ± 2.14
L5	763	30.40	28.00–32.00	25.64–36.00	30.38 ± 3.29

Superior endplate width (mm)

Level	n	Median	IQR	5th–95th	Mean ± SD
T11	730	37.60	34.50–40.90	30.55–45.31	37.75 ± 4.67
T12	743	39.90	36.80–43.70	33.00–48.70	40.27 ± 4.97
L1	740	40.70	37.60–44.30	34.50–49.10	40.96 ± 4.72
L2	723	42.20	38.90–45.60	36.10–50.30	42.39 ± 4.58
L3	757	44.10	40.80–48.30	37.20–53.10	44.61 ± 5.19
L4	768	47.00	43.70–50.60	39.10–56.00	47.05 ± 5.61
L5	753	51.00	46.50–56.00	40.36–64.80	51.42 ± 8.13

Spinal canal width (mm)

Level	n	Median	IQR	5th–95th	Mean ± SD
T11	737	20.30	18.80–21.90	16.40–24.42	20.36 ± 2.46
T12	742	23.15	21.80–24.98	19.40–27.40	23.28 ± 2.47
L1	741	23.90	22.50–25.60	20.30–28.10	24.11 ± 2.46
L2	730	24.20	22.70–25.88	20.70–28.35	24.36 ± 2.40
L3	771	25.00	23.60–26.70	21.30–29.30	25.19 ± 2.61
L4	784	26.20	24.60–28.10	22.10–31.58	26.44 ± 3.02
L5	761	30.00	27.90–32.40	24.60–36.50	30.27 ± 3.75

Spinal canal depth (AP) (mm)

Level	n	Median	IQR	5th–95th	Mean ± SD
L1	324	18.60	17.38–19.70	15.70–21.48	18.54 ± 1.75
L2	421	17.80	16.50–18.80	15.00–21.10	17.85 ± 1.79
L3	612	17.10	15.80–18.40	14.55–20.30	17.13 ± 1.79
L4	651	17.50	16.20–18.80	14.65–21.50	17.60 ± 2.20
L5	517	19.90	18.00–21.80	15.98–24.62	20.03 ± 2.78

Transverse pedicle width (mm)

Level	n	Median	IQR	5th–95th	Mean ± SD
L1	330	7.00	5.70–8.57	4.25–10.50	7.18 ± 1.97
L2	421	6.60	5.50–7.90	4.10–9.80	6.81 ± 1.88
L3	615	8.60	7.00–10.35	5.30–12.93	8.76 ± 2.39
L4	656	11.30	9.57–13.20	7.35–15.60	11.44 ± 2.63
L5	530	18.60	15.60–21.90	11.25–26.60	18.82 ± 4.68

Disc height (mm)

Level	n	Median	IQR	5th–95th	Mean ± SD
L1–L2	798	8.80	8.00–9.60	6.40–11.20	8.88 ± 1.62
L2–L3	793	9.60	8.80–11.20	5.60–12.80	9.74 ± 2.14
L3–L4	797	10.40	8.80–11.20	6.40–13.60	10.19 ± 2.11
L4–L5	786	10.40	8.80–12.00	4.80–13.95	10.20 ± 2.57
L5–S1	747	9.60	7.20–12.00	4.00–14.40	9.53 ± 3.33

Trabecular attenuation (HU)

Level	n	Median	IQR	5th–95th	Mean ± SD
L1	800	157.20	130.47–186.10	98.52–235.11	160.47 ± 42.47
L2	802	157.30	129.50–189.10	95.91–237.57	161.00 ± 43.33
L3	801	154.40	125.20–185.70	92.20–236.90	157.92 ± 44.11
L4	800	158.40	132.20–189.80	97.18–241.05	163.50 ± 44.98

Two checks worth noting. Body height crosses over — dorsal exceeds ventral at T11–T12 and the order reverses by L4–L5 — and pedicle width widens steeply into the lower lumbar spine. Both are properties the classical figures report, recovered here independently from CT. Panel (f) is a substitution rather than a reproduction: the textbook plots vertebral compression strength measured by loading cadaveric specimens to failure, and trabecular attenuation is a different quantity that stands in for it in living patients.