Original Article http://dx.doi.org/10.3349/ymj.2013.54.2.315 Yonsei Med J 54(2):315-320, 2013 pISSN: 0513-5796, eISSN: 1976-2437 Precommissural Fornix in the Human Brain: A Diffusion Tensor Tractography Study Sang Seok Yeo,1 Jeong Pyo Seo,1 Yong Hyun Kwon,2 and Sung Ho Jang1 1Department of Physical Medicine and Rehabilitation, College of Medicine, Yeungnam University, Daegu; 2Department of Physical Therapy, Yeungnam College of Science & Technology, Daegu, Korea. Received: December 21, 2011 Purpose: Other than a single case report, no diffusion tensor tractography (DTT) Revised: April 6, 2012 studies of the precommissural fornix in the human brain have been conducted. In Accepted: April 13, 2012 the current study, we attempted to visualize the precommissural fornix in the human Corresponding author: Dr. Sung Ho Jang, brain using DTT. Materials and Methods: We recruited 36 healthy volunteers for Department of Physical Medicine and this study. Diffusion tensor images were scanned using a 1.5-T scanner, and the pre- Rehabilitation, College of Medicine, commissural fornix was analyzed using Functional Magnetic Resonance Imaging Yeungnam University, 170 Hyeonchung-ro, Nam-gu, Daegu 705-717, Korea. of the Brain (FMRIB) software. Values of fractional anisotropy (FA), mean diffu- Tel: 82-53-620-3269, Fax: 82-53-620-3269 sivity (MD), and tract volume of the precommissural fornix were measured. Re- E-mail: [email protected] sults: The precommissural fornix originated from the hippocampal formation on each hemisphere as a crus; both crura were then joined to the body of the fornix. ∙ The authors have no financial conflicts of The body of the fornix continued anteriorly to the level just superior to the anterior interest. commissure, where it divided into each column of the precommissural fornix. Each column descended anteriorly to the anterior commissure and terminated in the sep- tal nuclei. Values of FA, MD, and tract volumes of the precommissural fornix did not differ between the right and left hemispheres (p>0.05). Conclusion: We believe that the methodology and results of this study would be helpful to future research on the precommissural fornix and in the elucidation of the pathology of diseases in- volving the precommissural fornix. Key Words: D iffusion tensor imaging, precommissural fornix, septal nucleus, hippocampus INTRODUCTION The fornix divides into the precommissural and postcommissural fornix around the anterior commissure. The precommissural fornix is the main neural pathway be- tween hippocampal formation and septal nuclei, and contributes to central regulation © Copyright: of emotional behavior, motivation processes, and memory function.1-5 Many studies Yonsei University College of Medicine 2013 have reported on postcommissural fornix function and injury in the human brain;6-16 This is an Open Access article distributed under the in contrast, little is known about the precommissural fornix, even though it has been terms of the Creative Commons Attribution Non- Commercial License (http://creativecommons.org/ explained in many “neuroanatomy” textbooks and atlases.17-19 This seems to be licenses/by-nc/3.0) which permits unrestricted non- mainly ascribed to the fact that the precommissural fornix is not easily discriminated commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. with adjacent structures on conventional brain MRI. Yonsei Med J http://www.eymj.org Volume 54 Number 2 March 2013 315 Sang Seok Yeo, et al. The recent development of diffusion tensor tractography el, and applied in the present study utilizing tractography rou- (DTT), which is derived from diffusion tensor imaging tines implemented in FMRIB Diffusion (5000 streamline (DTI), allows for visualization and localization of neural samples, 0.5 mm step lengths, curvature thresholds=0.2).24-26 tracts at the subcortical level in three dimensions.20 Accord- The precommissural fornix and the post commissural fornix ingly, DTT has been used for three-dimensional visualiza- were determined by selection of fibers passing through seed tion of the architecture and assessment of the state of the and target regions of interest (ROI). For analysis of the pre- postcommissural fornix, which is associated with cognitive commissural fornix, we placed the seed ROI on the septal nu- function.12-15,21,22 However, except for one case report, no cleus on the coronal image (Fig. 1).18,19 We defined the loca- DTT study on the precommissural fornix in the human tion of the septal nucleus as ventral to the septum pellucidum, brain has been performed.23 superior to the hypothalamus, and below the rostrum of cor- In the current study, using DTT, we attempted to visual- pus callosum.18,19 The target ROI was placed on the crus of the ize the precommissural fornix in the human brain. fornix on the coronal image. For analysis of the postcommis- sural fornix, we placed the seed ROI on the mammillary body on the axial image. We defined the location of the mammilla- MATERIALS AND METHODS ry body as dorsal to the hypothalamus and ventral to the inter- 　　　 peduncular fossa. The target ROI was placed on the crus of Subjects the fornix on the coronal image. Among the 5000 streamline Thirty-six normal healthy subjects (21 males, 15 females; samples generated from each seed voxel, tractography outputs mean age, 30.92±8.27 years; range, 20 to 48) with no histo- from each individual subject were thresholded and binarized ry of neurologic disease were recruited for the study. All to include only those voxels through which at least one participants provided written consent prior to participation streamline sample passed. Values of fractional anisotropy in the study, which was approved by the institutional re- (FA), mean diffusivity (MD), and tract volume of the precom- view board at our hospital. missural and postcommissural fornix were measured. Two authors (Yeo SS and Seo JP) performed random Diffusion tensor image tractography analyses of the data to measure inter-observer and intra-ob- DTI were scanned using a 6-channel head coil on a 1.5 T server variations. The consistency rates of analysis were Philips Gyroscan Intera MRI scanner (Philips, Ltd., Best, moderate for the comparisons made by the two observers the Netherlands) with single-shot echo-planar imaging. For (K=0.654, p=0.000) and the two sets of analyses made by each of the 32 non-collinear diffusion sensitizing gradients, one analyser (Yeo SS) (K=0.654, p=0.000). we acquired 67 contiguous slices parallel to the anterior commissure-posterior commissure line. Imaging parameters Statistical analysis were as follows: acquisition matrix=96×96, reconstructed to Data were analyzed using SPSS software (v.15.0; SPSS matrix=128×128 matrix, field of view=221×221 mm2, repe- Inc., Chicago, IL, USA). Intra-observer and inter-observer tition time=10726 ms, echo time =76 ms, parallel imaging reproducibility of the precommissural fornix was evaluated reduction fac tor (SENSE factor)=2, echo-planar imaging using kappa statistics. Values of FA, MD, and tract volume factor=49 and b=1000 s/mm2, number of excitations=1, and of the precommissural and postcommissural fornix were a slice thickness of 2.3 mm (acquired isotro pic voxel size analyzed with an independent t-test for determination of 2.3×2.3×2.3 mm3). variances between the right and left hemispheres. Statistical differences in DTI parameters between the precommissural Fiber tracking and postcommissural fornix were determined by the inde- Diffusion-weighted imaging data were analyzed using the pendent t-test. Results were considered significant when the Oxford Centre for Functional Magnetic Resonance Imaging p-value was <0.05. of the Brain (FMRIB) Software Library (FSL; www.fmrib. ox.ac.uk/fsl). Head motion effect and image distortion by RESULTS eddy current were corrected using affine multi-scale two-di- mensional registration. Fiber tracking was performed using a probabilistic tractography method based on a multifiber mod- The precommissural fornix originated from the hippocam- 316 Yonsei Med J http://www.eymj.org Volume 54 Number 2 March 2013 Precommissural Fornix A Seed ROI Target ROI Precommissural fornix B Postcommissural fornix C Fig. 1. (A) Seed regions of interest (ROI) were placed in the septal nuclei (yellow rectangle). The target ROI was placed in the crus of the fornix (yellow circle). (B) The pathway of the precommissural fornix is shown at each level of the brain in a normal subject (42-year-old male). (C) The pathway of the postcommissural fornix is shown at the anterior commissure level in a normal subject (42-year-old male). AC, anterior commissure. Yonsei Med J http://www.eymj.org Volume 54 Number 2 March 2013 317 Sang Seok Yeo, et al. pal formation on each hemisphere as a crus; both crura gree of directionality of microstructures, and the MD value were joined to the body of the fornix. The body of the for- indicates the magnitude of water diffusion.29-32 In contrast, nix continued anteriorly to the level just superior to the an- the tract volume was determined by the number of voxels terior commissure, where it divided into each column of the contained within the neural tract. Therefore, a decrement of precommissural fornix. Each column descended anteriorly FA value and increment of MD value without significant to the anterior commissure and terminated in the septal nu- difference of tract volume may indicate less directionality clei. The postcommissural fornix also originated from the of the precommissural fiber compared to the postcommis- hippocampal formation; however, the column passed from sural fiber. the posterior to the anterior commissure and terminated in Several studies have reported on the anatomy of the pre- the mammillary body (Fig. 1). commissural fornix in the animal and human brain.23,27,28,33 Values of FA, MD, and tract volumes of the precommis- In 1977, using an autoradiographic method in the rat brain, sural and postcommissural fornix were not different be- Swanson and Cowan27 found that the precommissural for- tween the right and left hemispheres (p>0.05). FA values of nix originated solely in fields CA1-3 of the hippocampus the postcommissural fornix were significantly higher than proper and the subiculum, while the projection to the mam- those of the precommissural fornix (p<0.05). MD values of millary complex, which comprises a major part of the de- the postcommissural fornix were significantly lower than scending columns of the fornix, has its origin in the dorsal those of the precommissural fornix (p<0.05). Tract volumes subiculum and the pre- and/or parasubiculum. In 2004, us- did not differ between the precommissural and postcom- ing T1-weighted 3D MRI, Watanabe, et al.28 demonstrated missural fornix (p>0.05) (Table 1). that the posterior hippocampal formation was connected to the dorsal part of the lateral septal nucleus via the fimbria and the precommissural fornix in the mouse brain. As for DISCUSSION the human brain, many neuroanatomy textbooks and atlas- es have explained the pathway and function of the precom- In the current study, we used DTT for identification of the missural fornix.17-19,23,33 However, as for regular research precommissural fornix in the human brain. The precom- studies, only two patients who showed injury of the pre- missural fornix, which is divided from the body of the for- commissural fornix have been reported. In 2008, Vann, et nix, descended anteriorly to the anterior commissure and al.33 reported on a patient who had tissue loss in the pre- terminated in the septal nuclei. By contrast, the postcom- commissural and postcommissural fornix following surgi- missural fornix passed posteriorly to the anterior commis- cal removal of a cavernous angioma. In 2005, Rollins23 re- sure and terminated in the mammillary body. These results ported on a neonate with semilobar holoprosencephaly who are compatible with those of previous studies and neuro- showed an abnormal precommissural fornix on DTT. anatomy textbooks.17-19,27,28 As for the DTI parameters, the Therefore, to the best of our knowledge, this is the first FA value for the precommissural fiber was lower than that original DTT study of the precommissural fornix in the hu- of the postcommissural fiber; in contrast, the MD value was man brain following Rollins’s case report.23 for the precommissural fiber was higher than that of the In conclusion, we identified the precommissural fornix in postcommissural fiber, although there was no difference in the human brain using DTT. We believe that the methodol- the value for tract volume. The FA value represents the de- ogy and results of this study would be helpful to future re- Table 1. Diffusion Tensor Parameters of the Precommissural Fornix and Postcommissural Fornix Precommissural fornix Postcommissural fornix t p value Right Left Both Right Left Both FA 0.335±0.039 0.347±0.039 0.341±0.039 0.356±0.045 0.362±0.072 0.359±0.060 -2.209 0.029* MD 1.709±0.366 1.631±0.389 1.670±0.377 1.434±0.235 1.383±0.214 1.407±0.225 5.049 0.000* 356.472 375.722 366.097 349.694 481.361 415.528 Tract volume -0.854 0.395 ±376.610 ±385.679 ±378.601 ±193.423 ±390.869 ±323.291 FA, fractional anisotropy; MD, mean diffusivity. Values represent mean (±standard deviation). MD×10-3 (mm2/s); Independent t-test was used to compare differences between the precommissural and postcommissural fornix. *p<0.05. 318 Yonsei Med J http://www.eymj.org Volume 54 Number 2 March 2013 Precommissural Fornix search on the precommissural fornix and in the elucidation nesia due to fornix infarction. Stroke 2000;31:1418-9. of the pathology of diseases involving the precommissural 12. Hong JH, Jang SH. Degeneration of cingulum and fornix in a pa- tient with traumatic brain injury: diffuse tensor tractography study. fornix. However, several limitations of DTI should be con- J Rehabil Med 2010;42:979-81. sidered in the interpretation of the results.34-37 First, DTI can 13. Chang MC, Kim SH, Kim OL, Bai DS, Jang SH. The relation be- underestimate the fiber tracts. DTI is a powerful anatomic tween fornix injury and memory impairment in patients with dif- fuse axonal injury: a diffusion tensor imaging study. NeuroReha- imaging method that can demonstrate the gross fiber archi- bilitation 2010;26:347-53. tecture; however, it is difficult to reflect all fibers. Second, 14. Jang SH, Kim SH, Kim OL. Fornix injury in a patient with diffuse regions of fiber complexity and crossing can prevent full axonal injury. Arch Neurol 2009;66:1424-5. reflection of the underlying fiber architecture by DTI. We 15. Sugiyama K, Kondo T, Higano S, Endo M, Watanabe H, Shindo K, et al. Diffusion tensor imaging fiber tractography for evaluating suggest that additional studies on the clinical correlation diffuse axonal injury. Brain Inj 2007;21:413-9. and age-related changes of the precommissural fornix be 16. Hong JH, Jang SH. Left fornical crus injury and verbal memory undertaken. impairment in a patient with head trauma. Eur Neurol 2010;63: 252. 17. Malcolm B, Carpenter. Core Text of Neuroanatomy. 4th ed. Balti- more: Williams & Wilkins; 1991. p.373-5. ACKNOWLEDGEMENTS 18. Afifi AK, Bergman RA. Functional Neuroanatomy: Text and At- las. 2nd ed. New York: McGraw-Hill; 2005. p.287-8. 19. Mai JK, Assheuer J, Paxinos G. Atlas of the Human Brain. 3rd ed. This work was supported by the 2011 Yeungnam Universi- San Diego: Elsevier Academic Press; 2008. p.127,129,131,133. ty Research Grant. 20. Mori S, Crain BJ, Chacko VP, van Zijl PC. Three-dimensional tracking of axonal projections in the brain by magnetic resonance imaging. Ann Neurol 1999;45:265-9. REFERENCES 21. Takei K, Yamasue H, Abe O, Yamada H, Inoue H, Suga M, et al. Disrupted integrity of the fornix is associated with impaired mem- ory organization in schizophrenia. Schizophr Res 2008;103:52-61. 1. Sheehan TP, Chambers RA, Russell DS. Regulation of affect by 22. Zhuang L, Wen W, Trollor JN, Kochan NA, Reppermund S, Bro- the lateral septum: implications for neuropsychiatry. Brain Res daty H, et al. Abnormalities of the Fornix in Mild Cognitive Im- Brain Res Rev 2004;46:71-117. pairment are Related to Episodic Memory Loss. J Alzheimers Dis 2. Brisch R, Bernstein HG, Dobrowolny H, Krell D, Stauch R, Trüb- 2012;29:629-39. ner K, et al. A morphometric analysis of the septal nuclei in 23. Rollins N. Semilobar holoprosencephaly seen with diffusion ten- schizophrenia and affective disorders: reduced neuronal density in sor imaging and fiber tracking. AJNR Am J Neuroradiol 2005; the lateral septal nucleus in bipolar disorder. Eur Arch Psychiatry 26:2148-52. Clin Neurosci 2011;261:47-58. 24. Behrens TE, Berg HJ, Jbabdi S, Rushworth MF, Woolrich MW. 3. McNaughton N, Corr PJ. A two-dimensional neuropsychology of Probabilistic diffusion tractography with multiple fibre orienta- defense: fear/anxiety and defensive distance. Neurosci Biobehav tions: what can we gain? Neuroimage 2007;34:144-55. Rev 2004;28:285-305. 25. Behrens TE, Johansen-Berg H, Woolrich MW, Smith SM, Wheel- 4. Henderson J, Greene E. Behavioral effects of lesions of precommis- er-Kingshott CA, Boulby PA, et al. Non-invasive mapping of con- sural and postcommissural fornix. Brain Res Bull 1977;2:123-9. nections between human thalamus and cortex using diffusion im- 5. Thomas GJ. Delayed alternation in rats after pre- or postcommis- aging. Nat Neurosci 2003;6:750-7. sural fornicotomy. J Comp Physiol Psychol 1978;92:1128-36. 26. Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens 6. Park SA, Hahn JH, Kim JI, Na DL, Huh K. Memory deficits after TE, Johansen-Berg H, et al. Advances in functional and structural bilateral anterior fornix infarction. Neurology 2000;54:1379-82. MR image analysis and implementation as FSL. Neuroimage 7. Poreh A, Winocur G, Moscovitch M, Backon M, Goshen E, Ram 2004;23 Suppl 1:S208-19. Z, et al. Anterograde and retrograde amnesia in a person with bi- 27. Swanson LW, Cowan WM. An autoradiographic study of the or- lateral fornix lesions following removal of a colloid cyst. Neuro- ganization of the efferent connections of the hippocampal forma- psychologia 2006;44:2241-8. tion in the rat. J Comp Neurol 1977;172:49-84. 8. Kuroki N, Kubicki M, Nestor PG, Salisbury DF, Park HJ, Levitt 28. Watanabe T, Radulovic J, Spiess J, Natt O, Boretius S, Frahm J, et JJ, et al. Fornix integrity and hippocampal volume in male schizo- al. In vivo 3D MRI staining of the mouse hippocampal system us- phrenic patients. Biol Psychiatry 2006;60:22-31. ing intracerebral injection of MnCl2. Neuroimage 2004;22:860-7. 9. Yasuno F, Hirata M, Takimoto H, Taniguchi M, Nakagawa Y, Ike- 29. Basser PJ, Pierpaoli C. Microstructural and physiological features jiri Y, et al. Retrograde temporal order amnesia resulting from of tissues elucidated by quantitative-diffusion-tensor MRI. J Magn damage to the fornix. J Neurol Neurosurg Psychiatry 1999;67: Reson B 1996;111:209-19. 102-5. 30. Assaf Y, Pasternak O. Diffusion tensor imaging (DTI)-based white 10. Korematsu K, Hori T, Morioka M, Kuratsu J. Memory impair- matter mapping in brain research: a review. J Mol Neurosci ment due to a small unilateral infarction of the fornix. Clin Neurol 2008;34:51-61. Neurosurg 2010;112:164-6. 31. Neil JJ. Diffusion imaging concepts for clinicians. J Magn Reson 11. Moudgil SS, Azzouz M, Al-Azzaz A, Haut M, Gutmann L. Am- Imaging 2008;27:1-7. Yonsei Med J http://www.eymj.org Volume 54 Number 2 March 2013 319 Sang Seok Yeo, et al. 32. Stadlbauer A, Salomonowitz E, Strunk G, Hammen T, Ganslandt anomalies. Radiographics 2005;25:53-65. O. Quantitative diffusion tensor fiber tracking of age-related 35. Parker GJ, Alexander DC. Probabilistic anatomical connectivity changes in the limbic system. Eur Radiol 2008;18:130-7. derived from the microscopic persistent angular structure of cere- 33. Vann SD, Denby C, Love S, Montaldi D, Renowden S, Coakham bral tissue. Philos Trans R Soc Lond B Biol Sci 2005;360:893- HB. Memory loss resulting from fornix and septal damage: im- 902. paired supra-span recall but preserved recognition over a 24-hour 36. Yamada K. Diffusion tensor tractography should be used with delay. Neuropsychology 2008;22:658-68. caution. Proc Natl Acad Sci U S A 2009;106:E14. 34. Lee SK, Kim DI, Kim J, Kim DJ, Kim HD, Kim DS, et al. Diffu- 37. Yamada K, Sakai K, Akazawa K, Yuen S, Nishimura T. MR trac- sion-tensor MR imaging and fiber tractography: a new method of tography: a review of its clinical applications. Magn Reson Med describing aberrant fiber connections in developmental CNS Sci 2009;8:165-74. 320 Yonsei Med J http://www.eymj.org Volume 54 Number 2 March 2013