Pergamon 0361-9230(95)02015-J BrainResearchBulletin, Vol.38, No. 5, pp. 457-460, 1995 Copyright© 1995Elsevier Science Inc. Printedin the USA. All rightsreserved 0361-9230/95$9.50 + .00 Use of MRI for Measuring Structures in Frozen Postmortem Brain D. LONGSON,* C. E. HUTCHINSON,'~ C. A. DOYLE, .1 M. D. C. SIMPSON,* P. SLATER* AND J. F. W. DEAKINI" *School of Biological Sciences, tDepartment of Diagnostic Radiology, School of Psychiatry and Behavioural Sciences, Universityof Manchester, Manchester M13 9PT, UK [Received 13 March 1995; Revised 15 June 1995; Accepted 20 June 1995] ABSTRACT: A method was developed for magnetic resonance imaging (MRI) of human autopsy brains stored long-term at -70°C. Scanning brains at temperatures between -70 and -8°C gave minimal MRI signals consistent with protons having limited freedom of movement at low temperature. Raising brain tern- perature improved the signal such that scanning at -1°C gen- erated images with good in-plane resolution, grey/white matter conbast, and fine detail of cortical sulcaVgyral patterns. To val- idate the method, volume and area measurements were made using computerized image analysis on stored digital images of 14 brains from adult subjects of both genders and various ages. The data confirmed that brain volume was inversely correlated with age, and female subjects had smaller brains. This is a val- uable new method for acquiring morphometric date from pre- viously unscanned pathologic brains that are to he used for neu- rochemical and molecular investigations. would include MRI examination of the brain for structural ab- normalities followed by histological and neurochemical analyses of any abnormal regions identified. However, antemortem MRI scans are rarely performed on psychiatric patients because of the costs involved and the limited access to the procedure. Further- more, clinical scans would be unlikely to contain the data re- quired for a good morphometric study. Immediate scans at the time of autopsy are usually impossible because of clinical de- mands on MRI scanners. In this article we describe how autopsy brains held in low temperature storage can be imaged with a standard MRI scanner and how morphometric data may be easily obtained without affecting the quality of the tissue for subsequent laboratory investigations. METHODS KEY WORDS: Frozen postmortem brain, Magnetic resonance imaging, Volume and area measurements. INTRODUCTION BRAIN scanning performed in vivo with magnetic resonance imaging (MRI) is an established diagnostic technique. MRI pro- duces images with better contrast and detail than computed to- mography (CT), especially in the temporal lobes and posterior fossa [4,6], and the resolution achieved with MRI has enabled in vivo imaging to examine fine details of brain structure previously only possible with postmortem pathologic techniques [3]. MRI has had great success in identifying some structural abnormalities in the brains of patients suffering from several psychiatric ill- nesses, including schizophrenia, affective disorders, and demen- tias [6], as well as other neurological disorders [1,5]. MRI scans of postmortem brains (fresh or fixed) can be a useful but unusual adjunct to the standard autopsy investigation [2]. Structural abnormalities in the brains of patients with psychi- atric illness sometimes identify the cause of the disease, or may possibly indicate regions in which neurochemical disturbances are likely to occur. Therefore, the ideal investigation of the pa- thology and aetiology of psychiatric and neurological diseases Brains Brains arrived in the laboratory within 2 h after removal at autopsy. The brain surface was cleaned under running water, and the whole brain, held in a polythene bag to restore the normal shape, was quick-frozen by rapid immersion in 10 1of isopentane at -70°C. Once frozen, brains were sealed individually in poly- thene bags, assigned a code number, and stored at -70°C for up to 4 years. Case histories were requested from hospitals and general prac- titioners. All subjects died suddenly or after a short illness, and there were no cases of coma or prolonged agonal hypoxia. His- tories of psychiatric illness, drug or alcohol abuse, psychoactive medication, or degenerative neurological disease were excluded in each case, and there were no family histories of psychiatric disorder. Magnetic Resonance Imaging Optimum defrosting technique and scanning temperature were determined in a short series of pilot experiments. A total of five brains (including cerebellum and brainstem) were used to investigate a range of temperatures (-70 °, -20 °, -8 °, -6 °, -I°C). Previous experience had shown that brains allowed to warm up in room air developed steep core-to-surface temperature gradients. The brains, initially at -70°C, were therefore warmed Requests for reprints should be addressed to C. A. Doyle, School of Biological Sciences,ManchesterUniversity, 1.124, Stopford Building,Oxford Road, Manchester MI3 9PT, UK. E-mail address: CDOYLE@FS2.SCG.MAN.AC.UK 457