Description of the sagittal alignment of the degenerative human spine

  • Amer Sebaaly
  • Pierre Grobost
  • Lisa Mallam
  • Pierre Roussouly
Original Article



To present the description of sagittal alignment of the degenerative human spine and its possible evolution.

Materials and methods

This is a retrospective observational study of degenerative evolution in spinal alignment in low back pain patients. Full spine EOS® sagittal X-rays were analyzed, and pelvic and spinal parameters were measured. Spinal shapes were classified on the hypothesis that the possible sagittal shapes of degenerative spine would be divided into four categories: “classical” Roussouly types 1–4, anteverted types (PT ≤ 5), retroverted types (PT ≥ 25) and kyphotic types.


A total of 331 patients (280 women and 51 men) were included. “Classic” types 1–4 represented the majority in this cohort (71.9%). Retroverted types made the second most common category with 20.8% of the cohort. Kyphosis group (lumbar and global) make only 5.8% of this cohort, while anteverted group make the lowest incidence (1.5%). Retroverted type 2 with thoracic kyphosis should be considered a separate type and made 1.5% of this cohort. Two theoretical subtypes, retroverted type 1 and type 4 were not found.


This is the first description of degenerative spine disease based on its shape and based on the classification of the normal variation in the sagittal alignment of the human lumbar spine described by Roussouly. Eleven types, divided into classical types, anteverted types, false shapes (retroverted) and kyphotic shapes, are described and an evolution pathway is proposed. An evaluation of surgical results in order to propose a treatment algorithm based on this classification should follow.

Level of evidence

Level IV cross sectional observational study.


Degenerative spine Adult spinal deformity Scoliosis Pelvic parameters Roussouly classification Sagittal balance 


Sagittal balance of the spine is a recent and booming concept for understanding and treating spinal pathologies [1]. It first started when Duval-Beaupère et al. described pelvic parameters [2, 3] that are the pelvic incidence (PI), the sacral slope (SS) and the pelvic tilt (PT) where PI is the sum of SS and PT (Fig. 1). PI tends to increase linearly with adolescent growing age and becomes a constant anatomical parameter in adulthood [1, 4, 5, 6]. Even more, global and local spinal parameters [lumbar lordosis (LL) and thoracic kyphosis (TK)] depend greatly on the PI [7, 8]. Based on that, Roussouly et al. proposed a classification for the normal spine balance depending on the SS and spinal shape with four types of normal spine [9, 10] (Fig. 2). This classification helps in determining the high local stress zones in the spine: The more the lumbar spine is curved, the more is the contact force on the posterior elements (facets in particular); the lower the lumbar curvature or flat back, the higher the impact is on the disks [11]. This classification was recently updated adding anteverted types 3 and 4 for the normal population (high SS and very low PI) [12]. Nonetheless, this description was established on asymptomatic population and was criticized for being unusable in pathological conditions. To add to that, spine degeneration modifies the organization of the curvatures and is responsible for compensation mechanisms above (at the spine level) or below the deformity (in the pelvis, hips and knees), making it difficult to use this classification in degenerative conditions [13].
Fig. 1

Drawing of the pelvis showing the three pelvic parameters: pelvic incidence (PI), sacral slope (SS) and pelvic tilt (PT) with PI = PT + SS

Fig. 2

Drawing of the Roussouly classification of the asymptomatic spine in four types based on sacral slope: type 1 with low SS, a long thoracolumbar kyphosis and a short lumbar lordotic curve; type 2 with low SS and the lumbar spine having a flat back appearance; type 3 with higher SS and an almost equal length of the kyphotic and lumbar lordosis curves; and finally type 4 with a very high SS, along lumbar lordosis and a shorter kyphosis

Many attempts have been made to classify degenerative spine diseases, such as the classification of ankylosing spondylitis [14], the degenerative spondylolisthesis [15, 16] or adult deformity (SRS-Schwab classification) [17]. Nonetheless, no attempt was made to classify sagittal imbalance beyond the primitive reasoning of balanced, unbalanced and compensated unbalanced spine [13]. This classification depends on the position of C7 with numerous described parameters to evaluate it (SSA, SVA [13], global tilt [18], TPA [19], Barrey ratio (BR) [13]). Again, all these parameters are positional parameters without any attempt to analyze the shape of the spine and its pathological evolution.

The aim of this study is to present the possible evolution of sagittal alignment of the degenerative human spine based on the original Roussouly classification of normal spine. It also describes the possible evolution in degenerative cases in order to determine the initial type of the patient’s spine helping the surgeon to plan the shape of the lumbar curvature in order to restore the optimal profile for each patient.

Materials and methods

This retrospective study was conducted after institutional review board clearance and evaluated 800 adults who consulted at our department between 2010 and 2016 for symptomatic back pain. Inclusion criteria included: (1) age > 35 years, (2) patients with mechanical low back pain and degenerative changes on imaging (degenerative disk disease, facet hypertrophy, degenerative spondylolisthesis) and (3) indication for lumbar fusion (failed non-surgical treatment). Exclusion criteria were: (1) patients who already had a surgery on the spine, excepted microdiscectomy, (2) patients with incomplete radiological records and (3) history of trauma, tumor or infection to the spine.

Each patient had a simultaneous standing anteroposterior and left lateral full spine radiographs including the pelvis. These were obtained using the EOS™ low-dose imaging system [20]. The following spinal and pelvic radiographic parameters were then measured: Pelvic parameters consisted of PI to assess sacropelvic morphology, PT and SS (Fig. 1), spinal parameters included LL, TK and number of vertebra in the lordosis (NVL), global spinal parameters included the BR which is the ratio between the horizontal distances from C7 plumb line to hip axis (HA) and between HA and posterior endplate of S1 to evaluate global balance.

All parameters mentioned above were measured by an independent operator using the KEOPS software (SMAIO, Lyon, France) as it was found to have a better reliability than manual measuring for sagittal pelvic parameter as well as for coronal Cobb angles [21].

Degenerative disease of the spine induces local or global kyphosing events, which may be compensated by pelvic retroversion in order to keep the plumb line above the femoral heads. We hypothesized that the possible sagittal shapes of degenerative spine would be divided into four categories according to the initial classification of the normal spinal shapes by Roussouly et al. [10] and to compensation mechanism involving the spine. These are the following:

Classical types As described by the initial report of Roussouly et al. [10] and is based on the value of SS and the form of the lordosis.
  • Type 1 Characterized by low SS (< 35°), where the LL is short and the kyphosis is long extending to the thoracolumbar area. It is a non-harmonious back with thoracolumbar kyphosis and short hyperlordosis.

  • Type 2 Characterized by low SS (< 35°) and by a longer but flat LL, close to a straight line. It is a harmonious flat back.

  • Type 3 Characterized by medium-range SS (35° ≤ SS ≤ 45°) with a well-balanced LL between its two arches. It is a harmonious regular back.

  • Type 4 Characterized by high SS (> 45°) with increased LL length and curvature. It is a harmonious hypercurved back.

It was noted that types 1 and 2 had low PI (< 50°), whereas types 3 and 4 had high PI values (> 50°).

Anteverted types As described by the revision of the Roussouly classification [12]. Anteverted types 3 and 4 have, respectively, the same characteristics of the classical types 3 and 4 (medium and high SS) but with a low PT (PT ≤ 5°). It was noted that these anteverted types had low PI [12].

Retroverted types These types arise from the four classical types with a pelvic retroversion. Retroverted types 1, 2, 3 and 4 have, respectively, the same characteristics of the classical types according to SS and LL shape, but they have a high PT (PT ≥ 25).

Kyphotic types When all compensatory mechanisms are consumed, kyphosis occurs. It could be divided into two subtypes according to the possibility of compensation in the thoracic spine.
  • Lumbar kyphosis shape Characterized by lumbar kyphosis with a hypokyphotic thoracic compensation curve.

  • Global kyphosis shape Characterized by lumbar kyphosis without a hypokyphotic thoracic compensation curve (and thus an unbalanced spine).

Statistical analysis

18.0 (International Business Machines Corporation, Armonk NY) was used for statistical analysis. Shapiro normality test was used to test the normal distribution of our population. Pearson correlation test (R) was used to evaluate different correlations. In accordance with Cohen, statically significant correlation were considered large clinically if R > 0.5, moderate if 0.3 ≤ R ≤ 0.5 and small if R < 0.3 [22]. p = 0.05 was chosen as significance level.


A total of 331 patients (280 women and 51 men), with mean age of 56.4 ranging from 35 to 82, met the inclusion and exclusion criteria and were included for the final analysis. Table 1 shows the incidence of each of the described types in this population. Subtype analysis showed that “classic” types 1–4 constituted the majority of the types in this population (71.9%). Retroverted types made the second most common category with 20.8% of the population. Kyphosis group made only 5.8% of this population, while anteverted group made the lowest incidence (1.5%). Two theoretical subtypes, retroverted type 1 and type 4 were not found in this population. However, a subtype of type 2 shape (classic and retroverted types) should be considered as a separate type: retroverted type 2 with thoracic kyphosis that made 1.5% of this population (SS < 35°, low LL, high TK).
Table 1

Incidence of each type of spinal shape of the proposed classification

Spinal shape type


Incidence (%)

Classical subtypes


 Type 1



 Type 2



 Type 3



 Type 4





 Type 3 Ant



 Type 4 Ant





 Retrov T1



 Retrov T2



 Retrov T2 + TK



 Retrov T3



 Retrov T4





 Global kyphosis



 Lumbar kyphosis






In this population of patients with degenerative spinal disease, the sagittal parameters of the spine varied significantly (Table 2). PI averaged 56.97° with a range of 26°–97°. The average value for LL was 49.4° with a range from − 13° to 94°. The average value for the TK was 32.3° with a range of 3°–77°. The average NVL was 4.6 with a range from 1 to 11. The average SS was 35°, with a range from 2° to 69°. PT averaged 21.4°, with a range from − 11° to 49°. The BR averaged 102% ranging between − 160 and 630%. Using the Shapiro normality test, all above variables except the TK had a normal distribution (Fig. 3). There was no difference in pelvic parameters between the two sexes.
Table 2

Studied variables with mean, standard deviation and range



















− 11







Number of vertebra in lordosis








− 13





− 12


Fig. 3

Histograms showing the normal distribution of the measured variables according to the Shapiro–Wilk test

The statistical correlations between the geometric parameters of lumbar and pelvic alignment are listed in Table 3. Some important correlations should be pointed out. Correlation between the SS and LL (R = 0.69; p < 0.001) indicates that the total amount of lordosis is still determined by the position of the superior endplate of S1 with respect to the horizontal axis. There is also a high correlation between SS and PI (R = 0.72; p < 0.001). Statistically significant weak correlation is present between PI and LL (R = 0.35; p < 0.01) as well as between SS and PT (R = − 0.26; p < 0.01).
Table 3

Correlation between the different pelvic and spinal parameters


















− 0.26

− 0.40







− 0.03





− 0.06

− 0.21








All correlations are significant at the 0.01 level (two-tailed)

Strong correlations are showed in bold

PI was analyzed according to the different described shapes (Fig. 4). Types 1, 2, anteverted types 3 and 4 and lumbar kyphosis were characterized by low PI. In contrast, type 3 and retroverted type 2 (and its subgroup with TK) had medium-range PI. Type 4 and retroverted type 3 share the same characteristics of very high PI. Finally, global kyphosis type could have low to high PI values.
Fig. 4

Histogram showing the mean pelvic incidence in each described subtype. Note that types 1, 2, anteverted 3 and lumbar kyphosis has low-range PI, whereas types 4 and retroverted 3 has high PI. The others types have medium-range PI. Ant anteverted, Retrov retroverted, TK thoracic kyphosis


Sagittal balance understanding is a primordial factor in implementing an accurate surgical strategy in degenerative spine. There is a paucity of degenerative spine classifications: The most recognized classification, the SRS-Schwab classification, relies on coronal and the sagittal imbalance [17]. Nonetheless, this classification doesn’t help neither in planning the surgical strategy nor in predicting the possible surgical outcome, especially when there is no sagittal imbalance and spinal shape is not taken into account. Surgical strategy could not be generated and rod bending could not be predicted by angles. When Roussouly et al. first presented their original classification based on 160 healthy subjects [10], it was received with mixed critics. Many criticized it for being described in normal population and unusable in degenerative disease, while others stated that it is not widely applied and poorly understood in the physician population. Other authors correlated normal spinal shape with pathological spine conditions. In fact, many of them found increased degenerative disk diseases in types 1 and 2 as proposed by Roussouly [23, 24], while others applied this classification to a population of mature idiopathic scoliosis [25]. Yet, this initial classification was made in an asymptomatic population and extrapolation to adult degenerative deformity was not made.

This paper presents the first description of the sagittal alignment of the degenerative human spine based on Roussouly classification. In this classification, 11 types of spinal degenerative shapes are described: The “classical” types 1–4, “anteverted” types 3 and 4, false type 2, false type 2 with TK, false type 3 and finally the “kyphotic” group (lumbar and global). Some interesting results concerning this classification are worth to be highlighted. First of all, the most common spinal types are the classical types (1–4). This gives reliability and clinical value to the initial description of Roussouly et al. However, compared to normal population, there is increasing incidence of type 1 and 2 shapes, possibly because type 1 could be a degenerative evolution of low PI types, whereas type 2 could be a degenerative evolution of higher PI types (see below). Secondly, one should note that there were no cases of retroverted type 1 and retroverted type 4. This could be explained by the fact that retroverted type 1 and global kyphosis (when associated with small PI) share the same characteristics and that type 4 loses lordosis to give raise to false type 3. Finally, same correlations exist between pelvic and spinal parameters in degenerative spine compared to the asymptomatic population. As a matter of fact, large correlation exists between SS and LL as well as between SS and PI. On the other hand, there is a moderate correlation between PI and PT (compared to its large correlation in asymptomatic population) [10]. This fact could be explained by the position of the pelvis in a spine affected by the degenerative disease.

D. Polly stated that “Life is a kyphosing event.” Compensation potential depends greatly on the PI; hence, low PI types have little compensation potential, whereas high PI types have bigger potential, with type 4 having the greatest potential of compensation. On the other hand, if the spine is flexible around this kyphosing event, there is an increased extension of the flexible spine above and/or below this local kyphosis. If the spine is rigid, with progressive kyphosis, the gravity line moves forward and the pelvis rotates backward (retroversion) inducing a decrease in the SS [13]. Considering these facts, this description of the sagittal alignment of the degenerative human spine is created by adding kyphosis to the four classical types. Hence, when the kyphosing event occurs in a type 1, compensation is done below in the lumbar spine by increasing more the lordosis on a short segment (accentuating the type 1). On the other hand, when compensation mechanisms are consumed, LL disappears and extension in the normally kyphotic thoracolumbar area is insufficient, giving raise to “global kyphosis” type with a very small PI (Fig. 5). Likewise, type 2 spines have a little range of compensation. When the kyphosing event affects the thoracolumbar area or lumbar area, either LL increases on a small arch generating a type 1 spine (thoracolumbar kyphosis), or lumbar spine lordosis disappears generating “lumbar kyphosis” type (if the thoracic spine could compensate with a hypokyphosis) or “global kyphosis type” (with a low PI) (Fig. 6). When the loss of lordosis in lumbar spine affects type 3 spines, compensation begins in the mobile spine above (loss of TK) and evolves enough to create retroversion with a straight spine with decrease in SS giving raise to false type 2. When thoracic spine is not mobile enough, thoracic hypokyphosis does not occur giving raise to false type 2 + TK with the only compensation being the increase in PT (retroversion of the pelvis). The ultimate evolution of type 3 is global kyphosis (with high PI) maximally retroverted pelvis (Fig. 7). Finally, decrease in lordosis in type 4 spines induces a type 3 shape with retroverted pelvis, the false type 3 spine. Subsequent evolution is identical to type 3 spines (Fig. 8).
Fig. 5

Drawing showing the possible evolution of type 1 shape. PI pelvic incidence, LL lumbar lordosis, SS sacral slope

Fig. 6

Drawing showing the possible evolution of type 2 shape. PI pelvic incidence, LL lumbar lordosis, SS sacral slope, TLK thoracolumbar kyphosis

Fig. 7

Drawing showing the possible evolution of type 3 shape. PI pelvic incidence, LL lumbar lordosis, SS sacral slope

Fig. 8

Drawing showing the possible evolution of type 4 shape. PI pelvic incidence, LL lumbar lordosis, SS sacral slope

This is a proposed explanation of evolution of normal type shapes into their pathological shapes. The comparison of PI between different types (Fig. 4) consolidates this evolution hypothesis. In fact, type 3 and false type 2 (with or without TK) have the same PI, whereas type 4 and false type 3 are characterized by very high PIs. Furthermore, it could help the surgeon predict the best correction magnitude and the best rods bending to ensure optimal results. For example, a global kyphosis shape would rather be restored to type 1 with short lordosis if PI is low or otherwise to type 3 or 4 with a longer lordosis if PI is high.

This study has several limitations. First of all, no explanation could be generated concerning the evolution of anteverted type 3 and type 4. These types are encountered in the asymptomatic population and in the degenerative population (even with a decrease in its incidence: 16% in the asymptomatic population to 1.8% in this population), and the evolution profile of these types is still unknown. It may be hypothesized that they may evolve to classical type 2 since they share the same PI range. Second of all, there were no correlation of classification types to functional scores (Oswestry, SRS-30, SF12, etc.). Moreover, this is a cross-sectional study with data collection unable to identify trends of degenerative evolution of the spines. Even more, the sample of this population could be considered small and further multicenter study could give more validity to this classification. Finally, even though this radiological population is a presurgical population, there was no correlation of described types to surgical results. Though, there is an ongoing study to evaluate the surgical results of degenerative spine fusion, performed according to this description, as well as to evaluate mechanical complications (pseudarthrosis, proximal junctional kyphosis, etc.) according to different types of this classification. This study should generate a treatment algorithm of degenerative spinal disease based on this classification.


This is the first description of degenerative spine disease based on its shape and based on the classification of the normal variation in the sagittal alignment of the human lumbar spine described by Roussouly. Eleven types, divided into classical types, anteverted types, false shapes (retroverted) and kyphotic shapes, are described and an evolution pathway is proposed. An evaluation of surgical results in order to propose a treatment algorithm based on this classification should follow.



The authors thank M. Rizkallah, MD for his valuable help in proofreading this manuscript.

Compliance with ethical standards

Conflict of interest

No conflict of interest for all authors regarding this paper.


  1. 1.
    Roussouly P, Pinheiro-Franco JL (2011) Sagittal parameters of the spine: biomechanical approach. Eur Spine J 20(Suppl 5):578–585. CrossRefPubMedPubMedCentralGoogle Scholar
  2. 2.
    Duval-Beaupère G, Schmidt C, Cosson P (1992) A barycentremetric study of the sagittal shape of spine and pelvis: the conditions required for an economic standing position. Ann Biomed Eng 20:451–462CrossRefPubMedGoogle Scholar
  3. 3.
    Legaye J, Duval-Beaupère G, Hecquet J, Marty C (1998) Pelvic incidence: a fundamental pelvic parameter for three-dimensional regulation of spinal sagittal curves. Eur Spine J 7:99–103CrossRefPubMedPubMedCentralGoogle Scholar
  4. 4.
    Mac-Thiong J-M, Roussouly P, Berthonnaud E, Guigui P (2011) Age- and sex-related variations in sagittal sacropelvic morphology and balance in asymptomatic adults. Eur Spine J 20(Suppl 5):572–577. CrossRefPubMedPubMedCentralGoogle Scholar
  5. 5.
    Mac-Thiong J-M, Berthonnaud E, Dimar JR et al (2004) Sagittal alignment of the spine and pelvis during growth. Spine (Phila Pa 1976)9:1642–1647CrossRefGoogle Scholar
  6. 6.
    Mac-Thiong J-M, Labelle H, Roussouly P (2011) Pediatric sagittal alignment. Eur Spine J 20(Suppl 5):586–590. CrossRefPubMedPubMedCentralGoogle Scholar
  7. 7.
    Vialle R, Levassor N, Rillardon L et al (2005) Radiographic analysis of the sagittal alignment and balance of the spine in asymptomatic subjects. J Bone Jt Surg Am 87:260–267. CrossRefGoogle Scholar
  8. 8.
    Boulay C, Tardieu C, Hecquet J et al (2006) Sagittal alignment of spine and pelvis regulated by pelvic incidence: standard values and prediction of lordosis. Eur Spine J 15:415–422. CrossRefPubMedGoogle Scholar
  9. 9.
    Berthonnaud E, Dimnet J, Roussouly P, Labelle H (2005) Analysis of the sagittal balance of the spine and pelvis using shape and orientation parameters. J Spinal Disord Tech 18:40–47CrossRefPubMedGoogle Scholar
  10. 10.
    Roussouly P, Gollogly S, Berthonnaud E, Dimnet J (2005) Classification of the normal variation in the sagittal alignment of the human lumbar spine and pelvis in the standing position. Spine (Phila Pa 1976) 30:346–353. CrossRefGoogle Scholar
  11. 11.
    Roussouly P, Pinheiro-Franco JL (2011) Biomechanical analysis of the spino-pelvic organization and adaptation in pathology. Eur Spine J 20(Suppl 5):609–618. CrossRefPubMedPubMedCentralGoogle Scholar
  12. 12.
    Laouissat F, Sebaaly A, Gehrchen M, Roussouly P (2017) Classification of normal sagittal spine alignment: refounding the Roussouly classification. Eur Spine J. Google Scholar
  13. 13.
    Barrey C, Roussouly P, Perrin G, Le Huec J-C (2011) Sagittal balance disorders in severe degenerative spine. Can we identify the compensatory mechanisms? Eur Spine J 20:626–633. CrossRefPubMedPubMedCentralGoogle Scholar
  14. 14.
    Debarge R, Demey G, Roussouly P (2011) Sagittal balance analysis after pedicle subtraction osteotomy in ankylosing spondylitis. Eur Spine J 20(Suppl 5):619–625. CrossRefPubMedPubMedCentralGoogle Scholar
  15. 15.
    Kepler CK, Hilibrand AS, Sayadipour A et al (2015) Clinical and radiographic degenerative spondylolisthesis (CARDS) classification. Spine J 15:1804–1811. CrossRefPubMedGoogle Scholar
  16. 16.
    Gille O, Challier V, Parent H et al (2014) Degenerative lumbar spondylolisthesis: cohort of 670 patients, and proposal of a new classification. Orthop Traumatol Surg Res 100:S311–S315. CrossRefPubMedGoogle Scholar
  17. 17.
    Schwab F, Ungar B, Blondel B et al (2012) Scoliosis Research Society-Schwab adult spinal deformity classification: a validation study. Spine (Phila Pa 1976) 37:1077–1082. CrossRefGoogle Scholar
  18. 18.
    Obeid I, Boissière L, Yilgor C et al (2016) Global tilt: a single parameter incorporating spinal and pelvic sagittal parameters and least affected by patient positioning. Eur Spine J. Google Scholar
  19. 19.
    Ryan DJ, Protopsaltis TS, Ames CP et al (2014) T1 pelvic angle (TPA) effectively evaluates sagittal deformity and assesses radiographical surgical outcomes longitudinally. Spine (Phila Pa 1976) 39:1203–1210. CrossRefGoogle Scholar
  20. 20.
    Dubousset J, Charpak G, Dorion I et al (2005) A new 2D and 3D imaging approach to musculoskeletal physiology and pathology with low-dose radiation and the standing position: the EOS system. Bull Acad Natl Med 189:287–297PubMedGoogle Scholar
  21. 21.
    Maillot C, Ferrero E, Fort D et al (2015) Reproducibility and repeatability of a new computerized software for sagittal spinopelvic and scoliosis curvature radiologic measurements: KeopsÒ. Eur Spine J. PubMedGoogle Scholar
  22. 22.
    Cohen J (1988) Statistical power analysis for the behavioural sciences, 2nd edn. Lawrence Erlbaum Associates, HillsdaleGoogle Scholar
  23. 23.
    Mardare M, Oprea M, Popa I et al (2016) Sagittal balance parameters correlate with spinal conformational type and MRI changes in lumbar degenerative disc disease: results of a retrospective study. Eur J Orthop Surg Traumatol 26:735–743. CrossRefPubMedGoogle Scholar
  24. 24.
    Bae J, Lee S-H, Shin S-H et al (2016) Radiological analysis of upper lumbar disc herniation and spinopelvic sagittal alignment. Eur Spine J 25:1382–1388. CrossRefPubMedGoogle Scholar
  25. 25.
    Hong J-Y, Kim K-W, Suh S-W et al (2016) Effect of coronal scoliotic curvature on sagittal spinal shape—analysis of parameters in mature adolescent scoliosis patients. Clin spine Surg. Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2017

Authors and Affiliations

  • Amer Sebaaly
    • 1
    • 2
  • Pierre Grobost
    • 1
  • Lisa Mallam
    • 3
  • Pierre Roussouly
    • 1
  1. 1.Department of Orthopedic SurgeryClinique médico-chirurgicale des MassuesLyonFrance
  2. 2.Saint Joseph University, Faculty of MedicineBeirutLebanon
  3. 3.National Institute of Applied Sciences of Lyon (INSA)LyonFrance

Personalised recommendations