A follow-up study of hippocampal subfield atrophy in mild cognitive impairment with automatic segment method
XIANGZHU ZENG1, HUISHU YUAN1, YING LIU1, and ZHENG WANG1

1RADIOLOGY, PEKING UNIVERSITY THIRD HOSPITAL, BEIJING, China, People's Republic of

Synopsis

Mild cognitive impairment (MCI) is often considered as an intermediate stage between normal aging and dementia, as reflected by a higher rate of conversion to clinical Alzheimer’s disease (AD) than in the normal elderly population. The hippocampal formation is a complex brain region with a primary role in memory function and vulnerable to neurodegenerative diseases. The aim of this research is to investigate the atrophy feature of hippocampal subfield and follow up the changes of hippocampal subfield in about two years by using automatic segmentation tool in patients with MCI.Our results suggest that compared to ICV, hippocampal subfield could be a more sensitive to detect cerebral changes of MCI. CA1, CA3 and left entorhinal cortex may be main subfields involved during MCI stage. Volumes of hippocampal subfield decreased in MCI patients and were positive correlation with clinical scores. There were also decreases in volumes of hippocampal subfield in MCI patients with the progress of the disease.

Purpose:

Mild cognitive impairment (MCI) is often considered as an intermediate stage between normal aging and dementia, as reflected by a higher rate of conversion to clinical Alzheimer’s disease (AD) than in the normal elderly population [1]. The hippocampal formation is a complex brain region with a primary role in memory function and vulnerable to neurodegenerative diseases [2]. The aim of this research is to investigate the atrophy feature of hippocampal subfield and follow up the changes of hippocampal subfield in about two years by using automatic segmentation tool in patients with MCI.

Method and materials:

Patients with MCI (n=23, 13 males and 10 females, mean age 71.13±8.25 years) were recruited in this study. Healthy control (HC) subjects (n=12, 7 males and 5 females, mean age 67.10±8.90 years) were also included. All subjects received Mini-Mental State Examination (MMSE) and Cognitive Abilities Screening Instrument (CASI) before MR examination. All MRI examinations were performed using a 3.0T MR scanner. The acquisition parameters: T1-weighted imaging [TR = 2530 ms, TE = 3.44 ms, flip angle = 7°, slice thickness = 1 mm, voxel size = 1.0×1.0×1.0 mm3, slices = 192, scan time = 6 min 3 sec]. T2-weighted imaging [TR = 4140 ms, TE = 97 ms, slice thickness = 2 mm, voxel size = 0.5×0.6×2.0 mm3, scan time = 5 min20 sec]. T2-weighted images were tilted coronal images which were perpendicular to the long axis of the hippocampus. HC subjects were scanned once and 11 of 23 patients were examined twice. The average scanning interval was 23.6±6.28 months in patients with MCI. Software called Automatic Segmentation of Hippocampal Subfields (ASHS, http://www.nitrc.org/projects/ashs/) was used to analyze the T2-images and 3D T1 images and volumes of hippocampal subfields were calculated. Then the volumes of subfields in MCI and HC were compared by independent-samples t-test with SPSS 19.0 and were used further to do correlation analysis [YF1] with MMSE and CASI scores. Volumes of hippocampal subfields of 11 MCI patients in first and second time were compared by paired-samples t-test with SPSS 19.0.

Results:

1. In patients with MCI (first time), mean MMSE scores were 26.26±1.63, mean CASI scores were 90.30±5.06. In Health controls, mean MMSE scores were 29.44±0.53, mean CASI scores were 97.78±1.56. There were significant differences of MMSE scores and CASI scores between patients with MCI and control subjects (pMMSE < 0.0001, pCASI = 0.001), especially in some subitems of CASI, like short-term memory (p = 0.001), focus (p = 0.025) and picture (p = 0.034)(figure.1).

2. Compared to health control, we found significant decreasing volumes of MCI patients (first time) in subsfields of right cornu Ammomis 1 (CA1) (p = 0.04), right CA2 (p = 0.03), right subiculum (SUB) (p = 0.04), left perirhinal cortex (Brodmann35) (p = 0.004). There were no significant differences in intracranial volume (ICV) (p > 0.05).

3. There were positive correlation in bilateral SUB (r = 0.43, p = 0.015 for left; r = 0.40, p = 0.027 for right) and left perirhinal cortex (r = 0.47, p = 0.007) with MMSE scores. Left Perirhinal cortex has positive correlation with focus of CASI score (r = 0.464, p = 0.040).

4. Compared 11 patients with MCI first time to those of second time (paired t-test), we found significant decrease in MCI patients (second) in subsfields of left cornu Ammomis 3 (CA3) (p = 0.025), left entorhinal cortex (p = 0.042). There were no significant differences in intracranial volume (ICV) (p > 0.05) between two times.

Discussion and Conclusion:

Our results suggest that compared to ICV, hippocampal subfield could be a more sensitive to detect cerebral changes of MCI. CA1, CA3 and left entorhinal cortex may be main subfields involved during MCI stage. Volumes of hippocampal subfield decreased in MCI patients and were positive correlation with clinical scores. There were also decreases in volumes of hippocampal subfield in MCI patients with the progress of the disease. Automatic segmentation of hippocampal subfields is efficient and may provide more sensitivity than whole hippocampal volumes.

Acknowledgements

All autors thanks for GE Healthcare, MR Research China, Beijing.

References

1. Petersen RC (2004) Mild cognitive impairment as a diagnostic entity. J Intern Med 256, 183-194.

2. Jack, C.R., Petersen, R.C., Xu, Y., O'Brien, P.C., Smith, G.E., Ivnik, R.J., Boeve, B.F., Tangalos, E.G., Kokmen, E., 2000. Rates of hippocampal atrophy correlate with change in clinical status in aging and AD. Neurology 55, 484–489.

Figures

MMSE and CASI scores of MCI vs HC



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4048