• Users Online: 250
  • Print this page
  • Email this page

Table of Contents
Year : 2020  |  Volume : 69  |  Issue : 4  |  Page : 243-248

In vivo cross-sectional topographic anatomy at sternal angle on magnetic resonance imaging

1 Department of Radiology, 7 Air Force Hospital, Kanpur, Uttar Pradesh, India
2 Department of Commandant, Military Hospital, Kirkee, Maharashtra, India
3 Department of Anatomy, Sri Ramachandra Medical College and Research Institute, Chennai, Tamil Nadu, India
4 Department of Obstetrics and Gynaecology, Armed Forces Medical College, Pune, Maharashtra, India

Date of Submission09-May-2020
Date of Acceptance28-Aug-2020
Date of Web Publication29-Dec-2020

Correspondence Address:
Dr. Rohit Aggarwal
Department of Radiology, 7 Air Force Hospital, Kanpur - 208 004, Uttar Pradesh
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/JASI.JASI_85_20

Rights and Permissions

Introduction: The manubriosternal angle, first described by Louis in 1825, is an important landmark in the anatomy of the thorax and has been conventionally described as corresponding to the T4–5 IV disc level based on cadaveric dissections. The objective of this study was to document the level of the angle of Louis and various anatomic structures that also correspond to the same level in living individuals based on multiplanar magnetic resonance (MR) images. Material and Methods: We reviewed MR scans of the cervicodorsal spine of 262 individuals comprising 174 males and 88 females in the age range 14–76 years. For each individual, the vertebral level of the following structures was noted on T1-weighted (T1W)/T2-weighted (T2W) turbo spin echo (TSE) coronal and sagittal images, namely tracheal bifurcation (TB), aortic arch (AA), and sternal angle (SA). Results: The SA was most commonly seen corresponding to the T5 vertebral body level (45.20%) and at T4–5 IV disc level in only 20.45% of the individuals. The convexity of the arch of the aorta was seen in the majority of the individuals corresponding to the T3 vertebral body level (47.96%). TB was seen at T4 level in 34.35% and only in 22.69% at the T4–5 IV disc level. Discussion and Conclusion: The anatomical level of the SA, AA, and TB in living individuals as assessed on MR images is significantly different from the traditionally held belief based on cadaveric dissections.

Keywords: Aortic arch, magnetic resonance imaging, sternal angle, tracheal bifurcation

How to cite this article:
Aggarwal R, Calicut SM, Sheik RA, Theegala VS, Mukhopadhyay I. In vivo cross-sectional topographic anatomy at sternal angle on magnetic resonance imaging. J Anat Soc India 2020;69:243-8

How to cite this URL:
Aggarwal R, Calicut SM, Sheik RA, Theegala VS, Mukhopadhyay I. In vivo cross-sectional topographic anatomy at sternal angle on magnetic resonance imaging. J Anat Soc India [serial online] 2020 [cited 2021 Jan 24];69:243-8. Available from: https://www.jasi.org.in/text.asp?2020/69/4/243/305380

  Introduction Top

Sternal angle (SA), the forward prominence formed by the manubriosternal joint, is an important landmark in the anatomy of the thorax. Based on cadaveric studies, conventional anatomical textbooks mention the SA plane and surface marking of SA to correspond to many anatomical structures including tracheal bifurcation (TB) and aortic arch (AA).[1],[2],[3] However, this belief has been challenged by many authors using computed tomography (CT)-based in vivo studies. All these studies conducted on different ethnic populations have shown results contrary to this traditional belief, and there has been a very wide variation in the vertebral levels of these structures in living human beings. Furthermore, there is a variable correlation among AA, TB, and SA plane. There is also variable evidence regarding age and gender differences in the levels of these structures.[4],[5],[6],[7],[8] Surface anatomy and markings are essential components of curriculum for medical students, and it is prudent to update this knowledge with availability of modern cross-sectional imaging modalities. This information is vital for planning and execution of many lifesaving interventions. There are few in vivo studies in medical literature that have assessed the anatomy of SA in human beings using magnetic resonance imaging (MRI) as an anatomical tool. MRI by virtue of its direct multiplanar imaging capability and superior contrast resolution is an excellent modality for depicting anatomy. The in vivo anatomical details of various anatomical landmarks that are traditionally believed to lie at the SA plane can be accurately depicted using MRI. In addition, results obtained from free-breathing MRI scans are likely to be closest to the normal physiology in human subjects. This study is an attempt to review the traditional anatomical descriptions of the SA plane and other thoracic structures (AA and TB) with respect to their corresponding vertebral levels and also in relation to each other using free-breathing MRI scans of the upper thorax in living individuals.

  Material and Methods Top

We reviewed 262 MRI scans of the cervicodorsal spine of patients done between September 2018 and December 2019 at our institution using a Philips Achieva 1.5 Tesla scanner (Philips Medical Systems, Veenpluis 4-6, The Netherlands). Institutional ethical committee clearance was obtained. The following categories of patients were excluded from the study:

  • Patients with scoliosis of the cervical and/or dorsal spine
  • Patients with basilar invagination
  • Patients with collapse vertebrae
  • Patients with any congenital vertebral anomaly
  • Patients with evidence of sternotomy
  • Suboptimal visualization of structure under study.

All the MRI scans were performed in supine position with both arms lying parallel to and by the side of the torso. The routine protocol for MRI of the cervicodorsal spine at our institution consisted of T1-weighted (T1W) and T2-weighted (T2W) TSE sagittal and coronal and T2W gradient recalled echo (GRE) axial images, respectively, of which only the T1W and T2W coronal and sagittal images were reviewed for the purpose of this study.

The imaging parameters for the MR scan were as follows:

  • T1W TSE sagittal: Repetition time (TR) – 407 ms and time to echo (TE) – 14 ms
  • T2W TSE sagittal: TR – 4000 ms and TE – 115 ms
  • T1W TSE coronal: TR – 400 ms and TE – 15 ms.

All the scans were done at a field of view of 280 mm, slice thickness 3 mm, and with a matrix size 512 × 512.

For each individual, the vertebral levels corresponding to each of the following structures were noted: TB, AA, and SA. The vertebral levels were counted from above downward with the C2 vertebra as the reference due to its unique appearance on sagittal and coronal images. The sagittal images were used to evaluate the vertebral levels corresponding to the SA and AA, respectively, whereas the coronal images were used to evaluate the vertebral levels of the TB. A reference line was generated across the entire set of images in a series using an inbuilt software tool, which ensured that the line corresponded to the same coordinates on all the images of a series. A horizontal line drawn from the anterior aspect of the manubriosternal joint and another horizontal line drawn along the inferior surface (summit of concavity) of the AA were extended posteriorly across the vertebral column on sagittal images to assess the level of the SA [Figure 1] and the AA [Figure 2], respectively. The carinal angle was identified on coronal images, and a horizontal line corresponding to the inferior surface of the carina was generated across all the coronal images in the series to identify the corresponding vertebral level [Figure 3]. The vertebral body level or the IV disc level at which these lines passed through the vertebral column was noted. For ease of statistical analysis, these vertebral levels were further grouped into three broad groups [Figure 4]:
Figure 1: T2-weighted mid-sagittal image of the cervicodorsal spine with horizontal reference line to evaluate the vertebral level of sternal angle (white arrow)

Click here to view
Figure 2: T2-weighted sagittal images. (a) Reference line across the summit of inferior concavity of aortic arch (star). (b) Reference line extending up to the fourth dorsal vertebra

Click here to view
Figure 3: T1-weighted coronal images. (a) Reference line across carina (star). (b) Reference line extending up to the dorsal vertebra

Click here to view
Figure 4: T2-weighted sagittal image showing grouping of vertebral levels

Click here to view

  • Level 1: Superior to T4 vertebral body level
  • Level 2: At the T4 vertebral body and T4–5 IV disc level
  • Level 3: Inferior to T4–5 IV disc level.

Age groups were classified into three subgroups as <30 years, 30–50 years, and >50 years.

Chi-square test and mean test were used to assess the statistical significance in various data subsets. The relationship between the superior surface of AA and the manubrium sterni was also evaluated by drawing an additional horizontal line posteriorly from the sternal notch toward the vertebral column.

  Results Top

We analyzed the MRI scans of 262 individuals comprising 174 males and 88 females of the age range 14–76 years. Out of the total study group of 262 individuals, only the number of scans, as shown in [Table 1], was included in the data subset for studying the respective structures.
Table 1: Number of cases included in the data subset

Click here to view

The SA corresponded to Level 3 in the majority of the individuals (58.91%) [Table 2]. A statistically significant difference was noted in the levels of the SA between males and females. Among females, the SA corresponded to Level 3 in 74%, whereas in males, it corresponded to Level 2 (46.71%) and 3 (49.63%), respectively (P < 0.001) [Figure 5]. The age distribution was as per [Figure 6].
Figure 5: Clustered stacked chart depicting sex distribution of sternal angle, aortic arch, and tracheal bifurcation

Click here to view
Figure 6: Clustered stacked chart depicting age distribution among various age groups of sternal angle, aortic arch, and tracheal bifurcation

Click here to view
Table 2: Vertebral levels of the sternal angle

Click here to view

The undersurface of the AA corresponded to Level 2 in the majority of the individuals (72.73%) [Table 3]. In the majority of both males (76.11%) and females (67.44%), the concavity of the AA corresponded to Level 2 [Figure 5]. In the majority of the individuals of all the age groups, the undersurface of the AA corresponded to Level 2 [Figure 6].
Table 3: Vertebral levels of the arch of the aorta

Click here to view

The bifurcation of the trachea corresponded to Level 2 in the majority of the individuals (54.07%) [Table 4]. In the majority of both females (55.35%) and males (57.90%), the TB corresponded to Level 2 [Figure 5]. A statistically significant difference was noted in the level of the TB in individuals below 50 years (<30 years – 71.40% and 30–50 years – 57.73% at Level 2) and those above 50 years (57.14% at Level 3) (P < 0.05) [Figure 6].
Table 4: Vertebral levels of tracheal bifurcation

Click here to view

No significant difference was noted in levels of TB and AA in males and females. Furthermore, no significant difference was noted in levels of SA and AA in different age groups (P > 0.05). On comparing levels of various structures in the same individual, AA and SA corresponded to the same level in 36.87% of the individuals (Level 2 in 29.03% and Level 3 in 7.83%). The AA was higher to the SA in 56.68% of the individuals (n = 217).

The TB and SA were at the same level in 41.13% of the individuals (Level 2 in 24.05% and Level 3 in 17.08%). In 42.4% of the individuals, the TB corresponded to a level higher to the SA (n = 158). The AA and TB were at the same level in 57.24% of the individuals (Level 1 in 46.20% and Level 3 in 11.03%). In 35.17%, the TB was below the SA (n = 145). Out of 133 individuals in whom we could evaluate the vertebral level of all the three structures, it was noted that they were at the same level in only 20.30% of the individuals (15.78% at Level 2 and 4.51% at Level 3). In these individuals, the median level of the SA was at Level 3, whereas the median level of the AA and TB, respectively, corresponded to Level 1. This difference was statistically significant using median test (P < 0.001).

The relationship of the upper surface of the AA to the manubrium sterni was separately evaluated in 90 individuals as part of the study. It was noted that in all these individuals, the upper surface of the AA was consistently above the level of the SA. In 61.11% of the individuals, it corresponded to the superior half of the manubrium sterni, whereas in 31.11%, it corresponded to the level of the sternal notch or higher.

  Discussion Top

Pierre Charles Alexander Louis (1787–1872) was an expert French morbid anatomist who has to his credit the description of the manubriosternal joint which is eponymously called the  Angle of Louis More Details. It is a subject of debate among medical historians whether he actually ever described the angle at the junction of the manubrium and the body of the sternum (SA) due to insufficient medical literature on the same. Essom-Sherrier and Neelon[9] state that, in the year 1910, Edward Goodman reviewed Louis's work and found a comprehensive description of the SA. Louis had described the SA as a bulging in the chest of patients with emphysema in the Journal of the Society of Medical Observations in 1837. Morton and Norman state that Louis described the SA in “Recherches antomico – pathologique sur la phthisie” in the year 1825,[10] but in an English translation of the same by Walshe[11] and in another article by Chukwuemeka et al.,[4] no specific description of the SA was found.

The mediastinum is strictly the partition between the lungs and includes the mediastinal pleura, although it is commonly applied to the region between the two pleural sacs. It is bounded anteriorly by the sternum and posteriorly by the thoracic vertebral column extending vertically from the thoracic inlet to the diaphragm. Conventionally, most of the anatomical texts divide the mediastinum into the superior and inferior mediastinum by a plane which passes from the manubriosternal joint to the vertebral column. The vertebral level corresponding to this plane is another debatable issue. Grant et al. mention that “at this level the trachea bifurcates, the arch of the aorta traverses the thorax from right to left in a posterolateral direction.”[1] Last RJ states that this is a horizontal plane passing horizontally through the SA backward to the lower border of the fourth thoracic vertebra.[2] The same author also mentions about the various other structures corresponding to this plane, stating that “the plane passes through the bifurcation of the trachea, the concavity of the arch of the aorta and just above the bifurcation of the pulmonary trunk. On the plane, the azygos vein enters the superior vena cava and the thoracic duct reaches the left side of the esophagus in its passage upward from the abdomen. Also lying in the plane are the ligamentum arteriosum and the superficial and deep parts of the cardiac plexus.” Williams et al. state that “the plane of division into upper and lower mediastinum traverses the manubriosternal joint and the junction of fourth and fifth thoracic vertebra.”[3] In a subsequent edition, Williams et al. state that “the plane of division into upper and lower mediastinum traverses the manubriosternal joint and the lower surface of the fourth thoracic vertebra.”[12]

Various imaging modalities have been used to study the level of SA in relation to the other mediastinal structures and vertebral levels. Arora and Singh in an editorial have comprehensively reviewed the role of imaging in assessing the SA. They suggested that CT and MR are preferred modalities for accurate depiction of mediastinal anatomy. They also noted that there is significant individual variation in the level of mediastinal structures in various studies.[13] Various authors have studied the cross-sectional anatomy of various mediastinal structures on CT.[4],[5],[6],[7],[8] Similarly, Shabshin et al.[14] studied sagittal MRI sections of thorax and localizer images of the cervicothoracic region to assess the reliability of various anatomical landmarks for accurate identification of dorsal vertebrae. In this study, the vertebral levels of the sternal apex, aortic bifurcation, and pulmonary bifurcation were assessed in 67 patients. In another study, Sharan et al.[15] studied the vertebral level of the sternal notch on T1 scout mid-sagittal MRI images in 106 consecutive patients to determine the appropriate surgical approach for thoracic spinal reconstruction without using thoracotomy or sternotomy. Connor et al.[16] have studied the utility of various anatomical landmarks in accurate identification of thoracic vertebrae on thoracic MRI. In this study, a total number of ten bony and soft-tissue anatomical landmarks such as sternal notch, inferior angle of the carina, and superior surface of AA were retrospectively studied in 100 thoracic MRI scans, and the authors found poor interobserver agreement, suggesting that the only reliable way of identifying the thoracic vertebrae is by identifying the C2 vertebra and counting down from thereon.

In the literature review, no other study could be found that has assessed vertebral levels of SA on MRI. However, many authors have assessed the same using CT scan.[4],[5],[6],[7],[8] Chukwuemeka et al. have shown that in the majority of the individuals, SA passed through the upper half of the T5 vertebral body (52.9%) with the level ranging from T4 to T6.[4] Similarly, Garg et al. have reported that the vertebral level of SA varied widely from T3/4 to T6/7, with the most common level being at T4/5 (35%).[5] Badshah et al. report the most common level to be at T5.[6] Uzun et al.[7] and Shen et al.[8] have reported the most common level of SA to lie at or above T4–5 IV disc. In our study, the SA passed through Level 3 in maximum cases (58.9%), followed by Level 2 (38.8%) with a range of T3 to T6/7 [Table 2]. Thus, the findings in our study broadly match the findings of other CT-based studies with respect to wide variation in level. Our findings are also similar to Chukwuemeka et al., Garg et al., and Badshah et al. with respect to the level of SA.[4],[5],[6] However, these findings are in variance with description of the SA in standard anatomy textbooks.[2],[3],[12] One of the possible explanations could be the movement of the thorax during inspiration, while CT images are acquired as opposed to previously described studies in cadavers. We found a statistically significant difference in vertebral levels of SA between males and females, wherein the SA was at a lower level in females. These findings are in variance with the findings of other authors where they found no statistical difference in the level of SA in relation to gender.[4],[5],[7],[8] No plausible explanation could be found or suggested by authors for this difference. With respect to age distribution, our study also does not show any significant difference similar to other studies.[4],[5],[6],[7],[8]

Garg et al.[5] have reported the undersurface of AA to be at or higher than T3–4 IV disc level in 75% of the cases, whereas Uzun et al.[7] reported the same in 62%. Our study revealed that in a majority of individuals (72%), the undersurface of AA corresponded to Level 2 which correlates with the traditionally held belief described in most textbooks of anatomy. Similar results have also been shown by Badshah et al.[6] and Shen et al.[8] The concavity of the arch of the aorta has been variably reported to be corresponding to plane of the SA in 7%–59%.[4],[5],[7],[8] In our study, however, the concavity of the arch of the aorta was at the level of SA in 36.8% and at a higher level in 56.68% of the cases. Interestingly, the level of the arch of the aorta did not show any such sex- or age-related variability in any of the studies. Although wide variation in the vertebral levels of TB has been reported with 63%–91% at or below T5.[5],[6],[7],[8] Our findings are at variance to this as only in 43.03% of the cases TB were at or below T5. With respect to correlation with SA plane, TB and SA are at the same plane in 4%–31% of the individuals and lower in up to 91% of the individuals.[4],[5],[7],[8] Our findings are at variance with these studies as we found TB and SA to be at the same plane in 41.13% of the cases. Furthermore, in another 42.4% of the cases in our study, TB was at a higher level than SA. The TB is seen to lie at a lower level with progression of age in our study. A significant difference was seen in the levels of TB in individuals below 50 years and above 50 years of age. The possible explanation for this difference could be age-related loss of elasticity of tracheobronchial cartilage, however, this needs to be studied further. Uzun et al. have reported a significant difference in levels of TB with respect to gender; however, no such difference was seen in our study.[7]

Shabshin et al.[14] correlated the superior margin of the aorta with the vertebral level and found it to vary from T2–T4 level, with only 8% of the cases above the level of T3 vertebra. Similarly, Connor et al.[16] have found that the mean vertebral level of the superior surface of AA is T3 with a range of T2–T5. In our study, the superior margin of AA was correlated with sternal notch instead of vertebral levels, and we found that in 31% of the cases, the superior margin of AA was above the level of sternal notch. AA was seen to correspond to the upper half of the manubrium or higher in more than 92% of the individuals. From an imaging perspective, the suprasternal approach for ultrasonography of the AA should be effective in the majority of the individuals due to this reason.

  Conclusion Top

The conventional teaching in anatomy textbooks has been that SA, TB, and AA lie at the same cross-sectional plane which corresponds to T4–5 vertebral level. This description is based on historical cadaveric dissections and is probably governed by the dictates of convenience rather than on statistical data in living individuals in contrary to this traditionally held belief. Our study involving a sizeable number of living individuals using MRI as an anatomical tool has shown that these three anatomical landmarks lie at the same plane in only a minority of individuals. Even among these individuals, a further smaller number show that the corresponding cross-sectional plane is at T4–5 level.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Grant JC, John VB, Charles ES. Grant's Method of Anatomy: A Clinical Problem-Solving Approach. 11th ed. Baltimore: Williams & Wilkins; 1989. p. 69.  Back to cited text no. 1
Last RJ, McMinn RM. Last's Anatomy: Regional and Applied. 9th ed. Edinburgh: Churchill Livingstone; 1994. p. 254.  Back to cited text no. 2
Williams PL, Warwick R, Dyson M, Bannister LH. Splanchnology. In: Gray's Anatomy. 37th ed. Edinburgh: Churchill Livingstone; 1989. p. 1245-475.  Back to cited text no. 3
Chukwuemeka A, Currie L, Ellis H. CT anatomy of the mediastinal structures at the level of the manubriosternal angle. Clin Anat 1997;10:405-8.  Back to cited text no. 4
Garg S, Gulati A, Aggarwal A, Gupta T, Mirjalili SA, Sahni D. Retrospective analysis of adult thoracic surface anatomy in Indian population using computed tomography scans. J Anat Soc India 2019;68:39-45.  Back to cited text no. 5
  [Full text]  
Badshah M, Soames R, Khan MJ, Marwat MI, Khan A. Revisiting thoracic surface anatomy in an adult population: A CT evaluation of vertebral level. Clin Anat 2017;30:227-36.  Back to cited text no. 6
Uzun C, Atman ED, Ustuner E, Mirjalili SA, Oztuna D, Esmer TS. Surface anatomy and anatomical planes in the adult Turkish population. Clin Anat. 2016;29:183-190.  Back to cited text no. 7
Shen XH, Su BY, Liu JJ, Zhang GM, Xue HD, Jin ZY, et al. A reappraisal of adult thoracic and abdominal surface anatomy via CT scan in Chinese population. Clin Anat 2016;29:165-1.  Back to cited text no. 8
Essom-Sherrier C, Neelon FA. The names and faces of medicine. Angle of Louis. N C Med J 1989;50:21.  Back to cited text no. 9
Morton LM, Norman JM. Morton's Medical Bibliography. 5th ed. Aldershot England: Scholar Press; 1991. p. 264.  Back to cited text no. 10
Louis PC. Researches on Pthisis Anatomical, Pathological and the Rapeutical. In: Walshe WH, editor. Vol. 8. Translator. London: Sydenham Society; 1844.  Back to cited text no. 11
Williams PL, Bannister LH, Berry MM, Collins P, Dyson M, Dussek JE, et al. Gray's Anatomv: The Anatomical Basis of Medicine and Surgery. 38th ed. Edinburgh: Churchill Livingstone; 1995. p. 1677.  Back to cited text no. 12
Arora VK, Singh V. Sternal angle revisited – From anatomy to radiology. J Anat Soc India 2013;62:95-97.  Back to cited text no. 13
Shabshin N, Schweitzer ME, Carrino NA. Anatomical landmarks and skin markers are not reliable for accurate labelling of thoracic vertebra on MRI. Acta Radiol 2010;51:1038-42.  Back to cited text no. 14
Sharan AD, Przybylski GJ, Tartaglino L. Approaching the upper thoracic vertebrae without sternotomy or thoracotomy. Spine 2000;25:910-6.  Back to cited text no. 15
Connor SE, Shah A, Latifoltojar H, Lung P. MRI-based anatomical landmarks for the identification of thoracic vertebral levels. Clin Radiol 2013;68:1260-7.  Back to cited text no. 16


  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6]

  [Table 1], [Table 2], [Table 3], [Table 4]


    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

  In this article
Material and Methods
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded25    
    Comments [Add]    

Recommend this journal