Cobb Angle Calculator (Scoliosis)
Cobb Angle Calculator (Scoliosis)
The Cobb angle from the two endplate inclinations, and the change from a previous film read against the published measurement error — because a difference under 10 degrees is not progression.
Cobb angle, and the change since the last film
Difference of two endplate inclinationsA standing PA spine film: the superior endplate of the upper end vertebra tilts 14 degrees one way and the inferior endplate of the lower end vertebra 13 degrees the other, and the previous film measured 22 degrees
A difference, not an arctangent
change since the previous film = this angle − the previous angle
- why this page does not use an arctangent
- because the Cobb angle is natively an angle. It is the angle between two vertebral endplates, and as Radiopaedia notes, “Most PACS will have a dedicated angle tool to measure this without needing the lines to intersect”. Forcing a rise and a run on a reader who has two angles would invent a measurement nobody took. The angle between two lines is the difference of their inclinations, so the arithmetic is a subtraction
- the sign convention, and why it only has to be consistent
- both endplates are measured from the same film horizontal, with one sign for each direction of tilt. In a curve they tilt opposite ways, so one is positive and one negative and the difference is the sum of the two magnitudes. Which direction you call positive does not change the answer; using two different conventions for the two endplates does, and that is the error worth guarding against
- this formula IS symmetric, unlike every other on this site
- the absolute value means swapping the two endplate readings gives exactly the same answer, which is correct — the angle between two lines does not depend on which you name first. So the transposition test that catches a swapped pair everywhere else is empty here, and pretending otherwise would be a test that proves nothing. What the proof asserts instead is that the signed difference underneath rises with the upper endplate and falls with the lower one, separately, and that the absolute value is symmetric on purpose
- which vertebrae and which endplates
- the end vertebrae are “the vertebra whose endplates are most tilted towards each other”, and the convention Wikipedia records from Caffey is the SUPERIOR endplate of the superior vertebra and the INFERIOR endplate of the inferior one — with the caution that “not all sources include usage of a superior versus inferior endplate in the definition”. Carman and colleagues found that pre-selecting the end vertebrae did NOT significantly improve reliability, so the choice of vertebrae is not where the error comes from
- derivation
- Cobb JR, 1948. The method as described here is Radiopaedia’s and Wikipedia’s rendering of it; Cobb’s own chapter was not reachable and nothing is attributed to it directly
Worked example
A standing PA spine film: the superior endplate of the upper end vertebra tilts 14 degrees one way and the inferior endplate of the lower end vertebra 13 degrees the other, and the previous film measured 22 degrees
Cobb angle = | 14 − (−13) | = | 27 | = 27.0 degrees, above the Scoliosis Research Society's threshold of greater than 10 degrees for scoliosis
Change since the previous film = 27.0 − 22 = 5.0 degrees. That is not progression on the published figures. Carman, Browne and Birch computed that "the difference would have to be 10 degrees for scoliosis radiographs" to exceed measurement variation, and Radiopaedia states that "if a greater than 10 degree change in Cobb angle is measured, it is 95% likely to represent a true difference"
Swap the two endplate readings: | −13 − 14 | = 27.0 degrees, exactly the same. This formula is symmetric on purpose, because the angle between two lines does not depend on which line you name first — and that makes it the one record in this batch where a transposed pair is harmless rather than invisible
Use two different sign conventions by mistake, entering both endplates as positive: | 14 − 13 | = 1.0 degree, which would read as a straight spine. That, and not a transposition, is the error this page's inputs are built to make visible
A curve at the diagnostic threshold: endplates at 5 and −5 degrees give exactly 10.0 degrees, which the Scoliosis Research Society calls postural asymmetry rather than scoliosis, since its definition is "any curve greater than 10 degrees"; at 5.5 and −5 it is 10.5 and the same definition calls it scoliosis. At 28 and −24 degrees it is 52.0, where a 10-degree measurement variation is still 10 degrees
The measurement error, which leads this page
| Finding | Figure | Cohort |
|---|---|---|
| Average difference between readings | “For scoliosis, the average difference between readings was 3.8 degrees” | Carman, Browne and Birch, J Bone Joint Surg Am 1990. Four staff orthopaedists and one physical therapist measured eight scoliosis radiographs and twenty kyphosis radiographs, on two occasions in random order |
| Spread of those differences | “95 per cent of the differences were 8 degrees or less (range, 0 to 10 degrees)” | The same study |
| How big a change has to be | “the difference would have to be 10 degrees for scoliosis radiographs” (and 11 degrees for kyphosis) | The same study. Radiopaedia states the same figure independently: “if a greater than 10 degree change in Cobb angle is measured, it is 95% likely to represent a true difference” |
| Does choosing the end vertebrae cause it? | “Reliability was not significantly improved when the end-vertebrae of the curve had been pre-selected” | The same study, which pre-selected them for some radiographs and not others |
| Patient rotation between films | “minor rotation of patients between examinations can significantly change measurements (may be as high as 20 degree variation)” | Radiopaedia |
| Posture and time of day | Lying down “decreases Cobb angles by around 7-10 degrees”; within one day “curvature increases during the day (~5 degree variation)” | Wikipedia, citing Keenan et al 2014, and Radiopaedia |
Published thresholds, with the body that publishes each
| Threshold | What the source says | Source |
|---|---|---|
| Greater than 10 degrees | “any curve greater than 10 degrees is considered scoliosis”; 0 to 10 degrees is “‘postural asymmetry’ which is not true scoliosis” | Scoliosis Research Society patient information. Radiopaedia states the same: “Scoliosis is defined as a lateral spinal curvature with a Cobb angle of >10 degrees” |
| 10 degrees or more | “AIS is defined as a lateral spine curvature of 10 degrees or more in the coronal plane” without congenital or neuromuscular comorbidities | AAFP, FP Essentials |
| Less than 20, then 20 to 26, then 26 to 45 degrees | “watchful waiting and/or referral for physical therapy are indicated”; then “Referral to a spine specialist for bracing is reasonable”; then bracing “is recommended” | AAFP, FP Essentials — that body’s statements, not this page’s |
| Greater than 40, then greater than 50 degrees | “Surgical intervention is considered for initial Cobb angles greater than 40 degrees and recommended for Cobb angles greater than 50 degrees” | AAFP, FP Essentials |
| Greater than 60 degrees | “some, especially curves greater than 60 degrees, continue to progress during adulthood”; respiratory complications reported above this figure | Scoliosis Research Society; Wikipedia citing the ERS Handbook 2015 |
A subtraction, and why five degrees is not a change
The Cobb angle is the angle between the endplates of the two end vertebrae of a spinal curve — the vertebrae whose endplates are most tilted towards each other. The convention Wikipedia records from Caffey is the superior endplate of the superior vertebra and the inferior endplate of the inferior one, with the honest caveat that not all sources include that distinction. Because the measured quantity is an angle and not a distance, this page takes two inclinations and subtracts them: that is what a PACS angle tool gives a reader, and the angle between two lines is the difference of their inclinations. The sign convention has to be the same for both endplates, which is the one input error that changes the answer.
The measurement error leads this page because it is bigger than readers expect and because the thresholds are packed more tightly than it is. Carman, Browne and Birch gave eight scoliosis radiographs to four staff orthopaedists and a physical therapist, twice each in random order: the average difference between readings was 3.8 degrees, 95% of differences were 8 degrees or less with a range to 10, and they computed that a difference would have to reach 10 degrees before it could be distinguished from measurement variation. Radiopaedia states the same figure independently, that a change greater than 10 degrees is 95% likely to be real, and records the widely quoted 5-degree measurement error. Pre-selecting the end vertebrae did not improve reliability significantly, so the choice of vertebrae is not the problem.
Put that beside the thresholds. The Scoliosis Research Society defines scoliosis as a curve greater than 10 degrees and calls 0 to 10 degrees postural asymmetry. The AAFP’s FP Essentials review describes watchful waiting below 20 degrees, bracing referral as reasonable from 20 to 26 degrees and recommended from 26 to 45, and surgery considered above 40 and recommended above 50. Those are that body’s statements, quoted here with attribution and not advice from this page. The gap between two of those bands is 6 degrees, and 95% of differences between two readings of one film are 8 degrees or less.
An angle is only as good as the two lines it is drawn between. Every threshold on this page assumes the measurement convention stated above, and a different landmark gives a different number from the same hip, heel, knee or spine — not a more or less accurate one. Agreement between observers is part of what a measured angle means, and for several of these angles the published spread between experienced readers is wider than the interval between the thresholds they are read against. Each page here gives the reliability figures with the number of observers and the number of radiographs they came from. This page renders no clinical or surgical decision; it computes the published quantity and states what the literature reports for it, with every threshold and performance figure given with the cohort it came from. A measurement is not a diagnosis, and a figure from a published series is a property of that series rather than a fact about one patient.
Frequently asked questions
What Cobb angle counts as scoliosis?
The Scoliosis Research Society’s own patient material states that any curve greater than 10 degrees is considered scoliosis and that 0 to 10 degrees is postural asymmetry rather than true scoliosis. Radiopaedia gives the same threshold, and the AAFP’s FP Essentials review defines adolescent idiopathic scoliosis as a curvature of 10 degrees or more in the coronal plane.
How much change between films is real?
On the published figures, 10 degrees. Carman, Browne and Birch computed that a difference would have to be 10 degrees on scoliosis radiographs to exceed measurement variation, and Radiopaedia states that a change greater than 10 degrees is 95% likely to represent a true difference. A 5-degree difference between two films is within the published variation of the measurement.
Why does this calculator ask for two angles rather than distances?
Because the Cobb angle is natively an angle. It is measured between two endplates, and as Radiopaedia notes, most PACS have a dedicated angle tool that reports it without the lines needing to intersect. Asking for a rise and a run would invent a measurement the reader never took. The angle between two lines is the difference of their inclinations.
Does it matter which way round I enter the two endplates?
No, and that is a real property of the measurement rather than a convenience: the angle between two lines does not depend on which you name first, so this calculation is symmetric in its two inputs. What does matter is using the SAME sign convention for both. Entering both as positive when they tilt in opposite directions gives a small number for a large curve.
Why does standing or lying down change the answer?
Because the curve itself changes. Lying down decreases Cobb angles by around 7 to 10 degrees, which is why measurements are preferably made standing; the curvature also increases through the day by about 5 degrees in one patient, and minor rotation between examinations has been reported to change the measurement by as much as 20 degrees.
Related calculators
References
- Radiopaedia. Cobb angle. The method: “one must first decide which vertebrae are the end vertebrae of the curve deformity”, “the vertebra whose endplates are most tilted towards each other”; “Lines are then drawn along the endplates (or the pedicles if the endplates are not properly visualized)” “and the angle between the two lines, where they intersect, measured.” Where they do not meet on the film, “a further two lines can be plotted, each at right angles to the previous lines”, and “Most PACS will have a dedicated angle tool to measure this without needing the lines to intersect”. “Scoliosis is defined as a lateral spinal curvature with a Cobb angle of >10°.” On measurement error: “a 5° measurement error is well established”; “if a greater than 10° change in Cobb angle is measured, it is 95% likely to represent a true difference”; “caution should be used in assuming that sequential measurements are correct when little change is evident”; “minor rotation of patients between examinations can significantly change measurements (may be as high as 20° variation)”; and in one patient on one day “curvature increases during the day (~5° variation)”.
- Carman DL, Browne RH, Birch JG. Measurement of scoliosis and kyphosis radiographs. Intraobserver and interobserver variation. J Bone Joint Surg Am. 1990;72(3):328–333. “Four staff orthopaedists and one physical therapist measured eight radiographs that showed scoliosis and twenty that showed kyphosis”, on two occasions and in random order. “For scoliosis, the average difference between readings was 3.8 degrees” and “95 per cent of the differences were 8 degrees or less (range, 0 to 10 degrees)”. For a change to be distinguishable from measurement error, “the difference would have to be 10 degrees for scoliosis radiographs and 11 degrees for kyphosis radiographs”. “The end-vertebrae were pre-selected for some radiographs and were freely chosen by the interpreter for others”, and “Reliability was not significantly improved when the end-vertebrae of the curve had been pre-selected” — so choosing the end vertebrae is not where the error comes from. The abstract reports differences between readings without separating intra- from inter-observer values, so neither is attributed separately here.
- Scoliosis Research Society. Adolescent idiopathic scoliosis (patient and family information). “A straight spine has a curve of 0º; any curve greater than 10º is considered scoliosis.” “Between 0º and 10º is considered ‘postural asymmetry’ which is not true scoliosis.” On adult progression: “some, especially curves greater than 60°, continue to progress during adulthood”. The page gives no numeric bracing or surgical threshold and does not discuss measurement variability. Its footer carries “© 2025 Scoliosis Research Society. All Rights Reserved”, so it is cited here for these short attributed factual statements and nothing structural is reproduced from it.
- American Academy of Family Physicians, FP Essentials: Adolescent idiopathic scoliosis. “AIS is defined as a lateral spine curvature of 10° or more in the coronal plane” without congenital or neuromuscular comorbidities. Its management thresholds, quoted as that body’s own statements and not as advice from this page: “For Cobb angles less than 20°, watchful waiting and/or referral for physical therapy are indicated”; “Referral to a spine specialist for bracing is reasonable for curves between 20° and 26°” and “is recommended for curves between 26° and 45°”; “Surgical intervention is considered for initial Cobb angles greater than 40° and recommended for Cobb angles greater than 50°”. The fetched page is short — about 1,800 characters in total, of which roughly 750 are clinical text — and names no author, so its thresholds are printed as that source’s position alongside the Scoliosis Research Society’s diagnostic definition rather than as a consensus.
- Wikipedia. Cobb angle. Useful for one thing the other sources leave out: which endplates. “measured from the superior endplate of a superior vertebra to the inferior endplate of an inferior vertebra” (cited there to Coley, Caffey’s Pediatric Diagnostic Imaging, p. 1429), with the caution that “not all sources include usage of a superior versus inferior endplate in the definition”. Also “Cobb angles are preferably measured while standing, since lying down decreases Cobb angles by around 7–10°” (cited to Keenan et al, 2014). Its severity table (<10, 10–30, 30–45, >45 degrees) is unattributed on the page and is not used for this record’s bands.
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/