Nuclear M edicine and Ultrasound in the Evaluation of Neurologic Diseases Charles A. Whelan, Anthony M. Passalaqua, and Philip Braunstein Echoencephalography and radionuclide brain imaging are used to aid in the diagnosis of a variety of intracranial abnormalities. Because of technical considerations, A - m o d e rather than B-mode echography must, under most circumstances, be employed for examination of the head. There are technical limitations to echoencephalography, and it tends to be relatively subjective and dependent on the skill of the sonographer. These factors have, to some extent, been responsible for the more limited application of ultrasound to neurologic diagnosis, as compared to the application of radionuclides. The most c o m m o n use of echoencephalography is in the detection of midline shifts associated w i t h various sequelae of head trauma. Ultrasound has also been used to detect space-occupying lesions, either directly or more usually by deter-
mining associated midline displacements. The characterization and follow-up of k n o w n lesions with ultrasound has also been described. In the above instances its usefulness is usually as a noninvasive technique c o m p l e m e n t a r y to nuclear medicine studies, which under most circumstances are more effective than ultrasound. The evaluation of ventricular enlargement, which is not ususally possible per se with radionuclides, is possible with ultrasound. Computerized transverse tomography (CTT) can be applied more efficiently in a routine w a y in many of the above circumstances, including evaluation of ventricular size. The combination of CTT and nuclear medicine procedures in the brain area is so effective and comprehensive that where these are both available the demand for ultrasound will probably decrease but will not be eliminated.
U R I N G the last quarter century remarkable advances have been made in the diagnosis of neurologic disorders. Plain film x-ray examinations, electroencephalography, pneumoencephalography, cerebral angiography, radionuclide cerebral imaging, and the recently introduced technique of computerized transmission tomography (CTT) have all contributed to improved diagnostic accuracy. Both nuclear medicine and ultrasound techniques are noninvasive, safe, relatively simple, and reasonably comfortable. Each method is capable of aiding in the detection and localization of neuropathologic disorders.
D
ROUTINE ECHOENCEPHALOGRAPHIC TECHNIQUES
A-mode echoencephalography initially was introduced as a method of midline determination by Leskell' in 1956. Unlike the case for other regions of the body, where the B scan has assumed greater significance, the unidimensional A-mode display utilizing a stationary transducer has remained the more popular type of echoencephalogram in studies involving the head, because of several factors that severely hamper successful B scanning. The irregular shape of the skull inhibits continuous contact between the transducer and the scalp as the two move relative to one another. The need for continuous contact can be solved with immersion Front the Division of Nuclear Medicine, Department of Radiology, New York UniversiO' Medical Center, New York, N. Y. Charles A. Whelan, M.D.; Anthony M. Passalaqua, M.D.; Philip Braunstein, M.D.; Division of Nuclear Medicine, Department of Radiolog)', New York University Medical Center. © 1975 by Grune & Stratton, Inc. Seminars in Nuclear Medicine, Vol. 5, No. 4 (October), 1975
419
420
WHELAN, PASSALAQUA, AND BRAUNSTEIN
scanning,~ but other considerations have limited this technique. Probably the most serious limitation in all types of B scanning is the topographic distortion produced by the skull itself.3The bones of the skull are often asymmetrical and of variable thickness. The velocity of sound traveling through the skull is approximately twice as fast as in the brain, so a variable skull thickness distorts the echo pattern. The skull attenuates the sound wave irregularly, and this creates a need for greater intensity in thicker areas, which in turn further limits the resolution of two-dimensional echoencephalography.4 The skull therefore poses formidable problems for B scanning, and although certain workers 5.8 feel that twodimensional scanning is feasible and rewarding, the majority of work done in echoencephalography is with A scans. Theoretically any abnormality that could be seen on a B scan should also be recognizable on an A scan, since the former is basically a summation of a large number of unidimensional (A-scan) determinations. The routine technique for A-mode echoencephalography involves positioning the probe against the fiat thin portions of the squamous temporal and parietal bones above and slightly anterior to the pinna of the ear. By applying a small amount of acoustical gel to the scalp an efficient acoustic interface can be obtained. A theoretical midline may be obtained by through-transmission scanning (two probes) where a single echo will appear to occur halfway between the probes, since the sound wave travels the interposed distance only once. An alternative method for determining the theoretical midline is to bisect the distance between the inner-table echoes if the equipment is capable of determining near-side echoes. The theoretical midline echo (T echo) serves as a reference standard on the triple-exposed photographic documentation of the oscilloscope scans. Several structures in the median plane of the brain have the potential of reflecting the ultrasound wave. 7 The pineal gland, particularly one containing calcium, causes sharp echoes. 8 However, midline echoes can be obtained even when the pineal is absent, or in children when there is no calcification present within the pineal body. The posterior part of the third ventricle is also midline, and its walls provide a sharp acoustic interface? Depending on the actual placement of the transducer and direction of sounding, the interhemispheric fissure, septum pelucidum, and the cerebral falx may also be echo sources in bitemporai scanning. The midline echo (M echo) is.recognized as the most stable echo complex and the one that has the greatest amplitude. Because nonmidline structures, such as lateral ventricle walls, may also give strong echoes, it is best to obtain several tracings of the M echo, rotating the probe as necessary to make certain that the echo representing the presumed midline is indeed the most stable and persistent echo visualized. The midline cannot be determined accurately from unilateral sounding, since the distance from the near-side echo to the midline echo includes the thickness of scalp and bone on the side of sounding and hence is greater than the midline-far-side echo distance. Therefore the M echo is analyzed from each side of the head. A photographic record (Fig. 1) is made from which the midlinefar-side echo distances can be determined. With the transducer on the right side of the head the M echo-far-side-echo measurement represents the distance between the midline structures and the inner table of the left side of the skull. With
NEUROLOGIC DISEASES
421
Fig. 1. Normal echoencephalogram. This figure is composed of near-side, midline, and far-side echo complexes. By convention the upper trace is obtained with the transducer on the right side of the skull, and the lower inverted trace is obtained with the transducer on the left side of the skull. The farside echoes for both traces are aligned and located on the reader's right side. The midline echoes from both sides of the head coincide, indicating no midline shift.
the transducer on the left side the distance between the midline and the inner table of the right side is obtained. The photographic record of the echoencephalogram must meet certain criteria to be considered diagnostic.~° The far-side echoes, representing the inner tables on the distal side of the skull, must be aligned within 2 mm of each other on the two tracings. If the midline echo has a double-spike configuration, then measurement of the width of the Complex should not vary more than 2 mm when comparing the tracings from the two sides. Any shift of the midline complex from the theoretical midline must show identical deviation on both the right-to-left and left-to-right scans if it is to be considered verified. There is some controversy as to what constitutes an abnormal M echo displacement. Several large series 9.'.~ suggest that when the center of the midline echo complex is displaced from the theoretical midline by more than 2 mm there is pathologic displacement. A 3-mm shift is an even more definite indication of diencephalic displacement, and many authorities ~3-~5 consider a shift of less than 3 mm insignificant. With the above somewhat simplified summary of A-mode encephalography in mind, we will consider where the technique can be utilized alone or in conjunction with radionuclide techniques to deduce neurologic diagnoses. There are four major areas of interest where correlation is appropriate: (I) in the evaluation of
422
WHELAN, PASSALAQUA, AND BRAUNSTEIN
trauma and post-traumatic disorders, (2) in the detection of tumors and in the evaluation of tumor growth, (3) in the characterization of intracranial lesions, and (4) in the evaluation of hydrocephalus and the analysis of ventricular shunting procedures. HEAD INJURIES
Echoencephalography has achieved its widest application and acceptance in the clinical setting of head trauma. Subdural hematomas, epidural hematomas, intracerebral hematomas, and the edema associated with cerebral trauma have all been shown to produce sufficient mass effect to create a recognizable shift of the midline echo complex. The degree of shift is largely dependent on the location of the lesion, with collections in the same plane as the pineal gland and posterior recesses of the third ventricle producing the greatest shift. Anterior and posterior masses, even when large, often do not produce shifts of the diencephalic structures, x8 The potential of echoencephalography in the detection of cerebral midline displacement was recognized early. Both European ~.~5,1r and American 9.~8 investigators reported diagnostic accuracy rates of over 90%. Yet others xg.~° were unable to achieve such high accuracy. White ~4recognized the primary flaw in the procedure. He showed that the examination was inherently subjective and that the validity of the results was dependent on the experience of the sonographer3 x Clinical prejudice.may influence the selection of the M echo. The lack of anatomic landmarks in the A scans prevents definitive recognition of improper echo recordings on the video display. A valid screening test should be objective, and an examination that is vulnerable to recorder and interpreter bias does not meet this fundamental criterion. Brylski22 in 1965 cited the reluctance of the medical community to accept sonography as a reliable procedure. A similar statement could be made today, and the extent of utilization of A-mode echoencephalography varies considerably from one institution to another. Two ways to eliminate subjectiviiy recently have been investigated. An automatic midline computer apparatus that renders gated histograms rather than simple echo spikes can reduce operator bias. White has recorded his experience with over 3300 automatic midline examinations.23 The histograms produced are quite reliable in excluding a midline shift at the level of the third ventricle. The results are adequately objective provided that the transducer is properly positioned. False positive results (15%-20%) are a bothersome problem with the automatic midline device. Since false negative examinations are quite rare, this device may well make a contribution as a screening examination in cases of head trauma: To the extent that this technique permits an objective evaluation of one part of the cerebral midline, the automatic midline device may provide the clinician with a certain degree of confidence when there is no shift demonstrated. It is possible with more sensitive equipment to increase the amount of information available on the A scan. Recently Rothman and Gershowitz24 reported their experience with a Hoffrel 21A ultrasonoscope that possesses improved near-side resolution and obtains actual echoes from structures at or adjacent to the lateral surface of the brain. By measuring the distance from the inner dural echo to the gyral surface echo they were able to localize 13of 17 subdural hema-
NEUROLOGIC DISEASES
423
.,,
4
.
f
B Fig. 2. Bilateral subdural hematomas. (A) Echoencephalogram showing a 5-mm shift of the midline echo toward the left side. (B} Anger-camera brain scintiphoto revealing a bilateral increased peripheral activity compatible with bilateral subdural hematomas. (C) Cerebral angiogram confirming bilateral subdural hematomas with a midline shift. Note that subdural on the right side is larger than subdural on the left side.
tomas. An additional benefit to be derived from equipment rendering greater anatomic detail is recognition of bilateral subdural'hematomas. Routine midline A scans may be normal in the presence of bilateral subdurals, since there may be no significant diencephalic shift. Figui'e 2A is an example of a relatively small displacement of the midline in a patient with a large right subdural hematoma because of another smaller subdural hematoma on the left side.
,
424
WHELAN, PASSALAQUA, AND BRAUNSTEIN
Specific techniques and procedures have been well established for the evaluation of cerebral trauma in nuclear medicine. Dynamic arterial perfusion studies permit reliable detection of displaced peripheral flow in the presence of a subdural hematoma. 25-~r Although earlier series suggested that brain scans were about 80% accurate ~9in the detection of subdural collections, the routine use of delayed scans and a complementary flow study may well result in greater sensitivity,s° Two recent investigations have compared the accuracy of A-mode echoencel~halography and brain scanning in the detection of chronic subdural hematomas. Raskind et al. 3t found the echoencephalogram positive in 20 of 32 patients (63%) as compared to 35 positive brain scans in 41 patients (85%) studied. Hurwitz et al. x6showed an even greater disparity in favor of the nuclear medicine procedure, with 13 of 14 patients studied having positive scans, whereas the echo was only 44% accurate in the same gi'oup. The particularly unsatisfactory result with echoencephalography in this particular study is partially attributable to the fact that 5 of the hematomas were eccentrically located and therefore did not in fact produce any diencephalic shift. Under certain circumstances echoencephalography may make a positive contribution in evaluating patients with cerebral trauma. The information gained may complement the results of the brain scan. Brain scanning is most appropriate and reliable with chronic subdural hematomas, whereas the echoencephalogram is most often helpful in acute situations and may be a valuable tool in the emergency.ward. Echoencephalography is a more appropriate study than the radionuclide scan in a patient suspected of having an-epidural or acute subdural hematoma, since these are not likely to be detected by radionuclide scanning. When the brain scan shows a "crescent sign ''3s and the analysis of its significance is complicated by ipsilateral scalp laceration or other superficial injury, then the demonstration of coexisting diencephalie displacement will help clarify the situation. When the .clinical situation involves an adult in a nonemergent situation and an obscure diagnosis, then the greater sensitivity and objectivity of the radionuclide angiogram with delayed static scans merit utilization as the screening examination of choice. 33 But here, too, there may be occasional value in supplementary echoencephalography if the brain scan appears normal. Activity in the tempo.ral muscles hinders radionuclide recognition of a subdural collection near the temporal lobe. This type ofhematoma is most likely to produce a sizable diencephalic displacement and may result in a dramatically abnormal A scan or histogram. The echoencephalogram also may substitute for serial brain scanning in certain situations. If surgical intervention is not undertaken for a subdural hematoma, then serial echography may be utilized to evaluate increasing or decreasing diencephalic displacement. Resolution of a cerebral contusion also can be followed by means of serial echography.
TUMORS In our experience, recourse to the sonogram for the detection of tumor is virtually unknown. By collating the information obtained from brain scanning with computerized transmission tomography we have found that it is possible to detect accurately over 95% of brain tumors, 34 and given such sensitivity echoen-
NEUROLOGIC DISEASES
425
i
•
•
i
Fig. 3A. Right-sided cerebral lesion. Echoencephalogramshowing a 5 - m m shift of the midline toward the left side. Fig. 3B. Anterior (left) and lateral (right) Anger-camera brain scintlphotos with99mTc-pertech netate showing a localized area of increased activity in the right cerebral hemisphere.
cephalography has little to add. Other investigators, however, have employed echoencephalography profitably, and we shall refer briefly to their endeavors. Space-occupying lesions in certain areas of the cerebral hemispheres may be expected to cause a shift of the midline echo complex to the contralateral side (Fig. 3). However, if only midline displacement is analyzed the echo will be positive in only about 50% of brain tumors. To be effective as a screening device
426
WHELAN, PASSALAQUA, AND BRAUNSTEIN
Fig. 3C. Cerebral angioconfirming the gram presence of a right cerebral neoplasm and a midline shift toward the left side.
for brain tumors, the echoencephalogram must display additional signs. This demands an experienced sonographer and clinical competence. Tumor impingement on the normal pathway of cerebrospinal flow may result in obstructive hYdrocephalus. Evaluating ventricular size by means of echoencephalography will be discussed in greater detail below in the section dealing with hydrocephalus. The mass effect of tumors affecting the brain stem has also been described. Tenner et al. 36 outlined a technique for localizing the aqueduct of Sylvius and quadrigeminal plate when the ultrasound beam is directed from a point midway between the occipital protuberance and the vertex toward the level of the lower orbital rim. They have been able to identify anterior and posterior displacement of the aqueduct in this fashion. The tumor itself may have a sufficiently distinct composition to provide abnormal echoes as a result of the particular acoustic impedance of the tumor contents. The velocity of sound in calcified tumors is particularly rapid, and such masses may produce high-amplitude echoes. It is difficult for most investigators tO recognize a particular echo as one emanating from a tumor. Often the echoes from the tumor appear as just one part of a dazzling array of confusing reflections. This is particularly true with A scanning. When immersion techniques ~ or other sophisticated B scans are obtained the tumor echoes may be more reproducible and more easily identified. Lombroso 37 reported visualizing tumor echoes in 34% (18/44)of supratentorial tumors evaluated with intensity-modulated twodimensional B scans.
NEUROLOGIC DISEASES
427
Fig. 3D. CTT scan revealing right cerebral mass with midline shift and ventricular dilatation.
CHARACTERIZATION OF BRAIN MASSES
The potential for distinguishing solid from cystic masses by sonographic techniques has been evaluated in several other articles in this Seminar. The reliability of such ultrasonic differentiation has been established in renal masses, and the rewards in pancreatic, hepatic, and thyroid sonography may be equally great. Unfortunately no similar success has been obtained with echoencephalography. It is, however, occasionally possible to differentiate solid from cystic lesions. The mass must first be localized with other neurologic diagnostic procedures. Transonic or echo-free zones may then be detected by sounding from the opposite side of the head. The reliability of the A scan in this regard is questionable. The difficulties of obtaining transcalvarial B scans have been pointed out above. When the skull is thin or has a defect, as was the case with the patient whose scan is shown in Fig. 4A, it is possible to obtain reliable B scans. Direct sounding of the brain through areas of limited acoustic impedance has been reported by Heimburger. 38 In this situation the resolution and attenuation problems inherent in transcalvarial scanning are absent, and it is feasible to determine cystic components and tumor growth rates, as well as ventricular outlines. Similarly, resolution is considerably better when scans are obtained through the very thin skulls of newborns. Brain scans are positive in approximately 85% of primary and metastatic brain tumors. 34.39.4° Dynamic flow studies and static radionuclide scanning permit
428
WHELAN, PASSALAQUA, AND BRAUNSTEIN
L
Fig. 4A. Hydrocephalus. B scan (gray tone) taken through a right temporal cranlotomy defect shows the contour of the enlarged lateral ventricles with thinning of the cortex and no shift of the midline.
!A
limited characterization of abnormal masses in the brain. Similarly, some information about the neoplastic nature of brain abnormalities can be obtained by studying masse.~ with different radiopharmaceuticals. Except in rare instances the echoencephalogram is not reliably helpful in characterizing lesions seen by radionuclide imaging. HYDROCEPHALUS
Prior to the introduction of computerized axial tomography, the echoencephalogram was virtually the only effective noninvasive technique for evaluating ventricular size. Radionuclide brain imaging is neither informative nor appropriate in the evaluation of ventricular size. Radionuclide cisternography does provide useful information about ventricular patterns and is particularly informative in the presence of communicating hydrocephalus. Cisternography, however, does involve a lumbar puncture and is not a noninvasive screening procedure. The detection of ventricular size by A-mode echoencephalography has been
Fig. 4B. Hydrocephalus. A-mode study (reversed to correlate with B scan} shows echoes from the midline and the lateral wall of the left lateral ventricle illustrating marked ventricular enlargement.
NEUROLOGIC DISEASES
429
Fig. 4C. H y d r o c e p h a lus, C1-1"scan taken through s a m e level as ultrasound studies. Note the craniotomy defect in the right temporal area.
proposed by many investigators.9.3s.a6,"-4a Careful examination of midline echo complexes usually shows that the relatively high M-echo peak is surrounded by two smaller echo patterns. These represent the posterior part of the third ventricle and may be used to evaluate size. When the midline tracings are being sought, lateral echoes are often detected approximately halfway between the Mecho complex and the echoes emanating from the inner table. The source of these echoes may be the walls of the lateral ventricles, although the Sylvian fissure echoes occasionally cause confusion. The atrial region and the temporal horns of the lateral ventricles both produce echoes, but the two usually can be distinguished by the experienced sonographer~ The atrial echo will be detected when the beam is directed horizontally, whereas the temporal-horn reflection is noted when the beam is directed downward and somewhat posteriorly from a position 2 cm above the root of the ear. A-mode scanning can be used to measure ventricular size (Fig. 4B), but an experienced examiner must perform such an examination. The ventricles may actually be measured from the A scan or an index relating the position of the temporal horn to the midline. When the ventricles are dilated the echos are often more easily recognized, since their echo amplitudes are increased. It may be difficult to obtain good reflections from the ventricular interfaces where a tumor itself displaces the wall. This may distort the normal perpendicular arrangement necessary for sharp echo display patterns. Therefore A-mode sc~/nning has been utilized as a reliable means of detecting and following children with hydrocephalus because of the thinness of the skull mentioned previously. In the adult, hydrocephalus is probably present when the third ventricle diameter exceeds 10 •
430
WHELAN, PASSALAQUA, AND BRAUNSTEIN
m m and the lateral ventricles m e a s u r e greater than 2.8 cm. S t a n d a r d reference planes have been established for obtaining echoes from the ventricular walls. 3s Patients with ventricular shunts m a y be followed with either serial echoenc e p h a l o g r a m s or nuclear medicine studies. Serial echoes obtained f r o m identical external l a n d m a r k s will indicate progressive ventricular dilatation in situations o f shunt malfunction. Intraventricular clearance studies offer an objective radionuclide test of C S F shunt patency, R u d d et al. 4s have found that in 25 patients with n o r m a l l y functioning shunts the Tt/2 o f 99mTc-pertechnetate injected into the central reservoir (flushing device) averaged 3 min. CONCLUSION
The role o f echoencephalography in neurologic diagnosis depends to some extent upon the availability o f computerized transmission t o m o g r a p h y . W h e r e C T T is available, its superior ability to objectively evaluate the midline and ventricular size has largely eliminated the use o f A - m o d e echoencephalography. C T T is capable o f detecting acute subdural h e m a t o m a s , and the relative absorption coefficients provid e reliable characterization of the nature o f a mass lesion. 4~ The combination o f radionuclide brain scanning and C T T provides c o m plementary information with r e m a r k a b l e sensitivity, to which A - m o d e echoencephalography has little to add. When C T T is not available and in situations where the appropriate information can be obtained utilizing ultrasound, A - m o d e echoencephalography employed by an experienced u l t r a s o n o g r a p h e r may be of clinical value in detecting midline shifts, evaluating hydrocephalus in children, and following patients with C S F shunts. Finally, echoeneephalography m a y provide reassuring information to the clinician dealing with patients with acute head t r a u m a . REFERENCES
h Leskell L: Echoencephalography. I. Detection of intracranial complications following head injury. Acta Chit Seand 110:301, 1956 " 2. Makow DM, McRae DL: Two-dimensional echoencephalography using immersion scanning: Recent results, in Kazner E, Schiefer W, Zulch KJ (eds): Proceedings in Echoencephalography. New York, Springer-Verlag, 1968 3. White DN, Clark JM, Chesebrough JN, et al: The effect of the skull in degrading resolution in echo-encephalographie B- and C-scans, in Kazner E, Schiefer W, Zulch KJ (eds): Proceedings in Echoencephalography. New York, Springer-Verlag, 1968 4. Fry WJ: lntracranial anatomy visualized in-vivo by ultrasound. Invest Radiol 3:243, 1968 5. Erba G, Lombroso CT: Detection of ventricular landmarks by two dimensional ultrasonography. J Neurol Neurosurg Psychiatry 31:232, 1968 6. Lombroso CT, Erba G, YogoT, et al: Two dimensional sonar scanning for detection of intracranial lesions. Arch Neurol 23:518, 1970 7. Schiefer W, Kazner E, Kunze S: Clinical
Echoencephalography. New York, SpringerVerlag, 1968, p 38 8 . Leskell L: Echoencephalography. 11. Midline echo from the pineal body as an index of pineal displacement. Acta Chir Scand 115:255, 1958 9. Brinker RD, King DL, Taveras JM: Echoeneephalography. Am J Roentgenol 93:781, 1965 10. Sandok BA, Henson TE, Skaggs H: Analysis of echoencephalograms. Neurology 20:933, 1970 II. Schiefer W, Kazner E, Kunze S: Clinical Echoencephalography. New York, SpringerVerlag, 1968, p.49 12. Jeppson S: Echoencephalography. IV. The midline echo; an evaluation of its usefulness for .diagnosing intracranial expansivities and investigation into its sources. Acta Chir Scand [Suppl] 1961, p 272 13. Elizondo-Martel E, Gersham-Cohen J: Medical ultrasonics. Am J Roentgenol 93:79J, 1965 14. White DN, Kraus AS, Clark JM, Camp-
NEUROLOGIC DISEASES
bell JK: Interpreter error in echoencephalography. Neurology 19:775, 1969 15. Ford R, Ambrose J: Echoencephalography. Brain 86:189, 1963 16. Hurwitz SR, Halpern SE, Leopold G: Brain scans and echoencephalography in the diagnosis of chronic subdural hematoma. J Neurosurg 40:347, 1974 17. Kazner E: Recognition and differential diagnosis of intracranial complications following head injuries by means of echoencephalography, in Kazner E, Schiefer W, Zulch KJ (eds): Proceedings in Echoencephalography. New York, Springer-Verlag, 1968 18. Rothman J, Shatsky S, Kricheff I, Chase N: Ultrasonic diagnosis of subdural hematomas. Am J Roentgenol 105:413, 1969 19. Drydale LN: Echoencephalography. The use of pulsed ultra-sound in neurologic diagnosis. Audt Radiol 10:207, 1969 20. Hanser LH, tlolmes HL: EchoencephaIography in a large city hospital. Rocky Mt Med J 63:34, 1966 21. White DN: The limitations of echoencephalography. Ultrasonics 5:88, 1967 22. Brylski MD, Ibenstark JL: Anatomic localization of midline echo in s°n°grams of the brain. Am J Roentgeno193:81 I, 1965 23. White DN, Hanna LF: Automatic midline echoencephalography. Neurology 24:80, 1974 24. Rothman J, Gershowitz PS: Ultrasonic diagnosis of subdural hematomas, epidural hematomas, brain edema and brain atrophy. Am J Roentgenol 122:531, 1974 25. Fish NB, Koch RL, Pollycove M, et al: Basis for accurate scintophotographic detection of subdural hematoma. J Nucl Med 9:316, 1968 26. Rosenthal L: Intravenous and intracarotid radionuclide cerebral angiography. Semin Nucl Med 1:70, 1971 27. Beiker A: Displacement of anterior cerebral vessels in cerebral dynamic studies in cases of chronic subdural hematomas. J Nucl Med 16;86, 1975 28. Apfelbaum RI, Newman SA, Fingesser LH: Dynamics of technetium scanning of subdural hematomas. Radiology 107:571, 1973 29. Cowan R J, Maynard CD, Lassiter KR: Technetium 99m pertechnetate brain scan in the detection of subdural hematomas. A study of the lesion as related to the development of a positive scan. J Neurosurg 32:30, 1970 30. Cowan R J, Maynard CD: Trauma to the brain .and extracranial structures. Semin Nucl Med 4:319, 1974 31. Raskind R, Glover MB, Weiss SR: Chronic subdural hematoma in the elderly: A
431
challenge in diagnosis and treatment. J Am Geriatr Soc 20:330, 1972 32. Hersen WJ, Quinn JL III, Mellihan WV: The crescent pattern of increased radioactivity in brain scanning. Radiology 87:483, 1966 33. Gilday DL, Coates G, Goldenbar D: Subdural bematoma--What is the role of brain scanning in its diagnosis. J Nucl Med 14:283, 1973 34. Passalaqua AM, Braunstein P, Kricheff I, et al: Clinical comparison of radionuclide brain imaging and computerized transmission tomography. Symposium on Past, Present and Future of Non-invasive Brain Imaging. (in press) 35. Ford RM, McRae DL; Echoencephalography--A standardized technique for the measurement of the width of the third and lateral ventricles, in Grossman CC (ed): Diagnostic Ultrasound, Proceedings of First International Conference. New York, Plenum, 1966, p 117 36. Tenner MS, Wodrasha G, Adapon BD: Newer ultrasound techniques in the evaluation of neurologic disorders. Radiol. Clin North Am 12:283, 1974 37. Lombroso CT, Erba G, Yogo T, et al: Two dimensional sonar scanning for detection of intracranial lesions. Arch Neurol 23:518, 1970 38. Heimburger RF, Eggleton RC, Fry FJ: Ultrasonic visualization in detection of tumor growth rate. JAMA 224:497, 1973 39. O'Mara RE, Mozley JN: Current status of brain scans. Semin Nucl Meal 1:7, 1971 40. Gottschalk A: Brain scanning--Is it becoming unnecessarily complicated? Am J Roentgenol I 11:864, 1971 41. Schiefer W, Kazner E, Kunze S: Clinical Echoencephalography. New York, SpringerVerlag, 1968, p 174 42. Kirkpatrick J: Echoencephalography in the evaluation of hydrocephalics. Radiology 86:1052, 1966 43. Brahme F, Traghard B: Echoencephalographic estimation of size of the lateral ventricles in normal children. Radiology 92:60, 1969 44. Erba G, Lombroso CT: Detection of ventricular landmarks by two dimensional ultrasonography. J Neurol Neurosurg Psychiatry 31:232, 1968 45. Rudd TG, Shurtleff DB, Loeser JD, Neph WB: Radionuclide assessment of cerebrospinal fluid shunt function in children. J Nuci Med 14:683; 1973 . 46. Ambrose J: Computerized transverse axial scanning (tomography). Part 2: Clinical application. Br J Radio146:1023, 1973 47. New PF, Scott WR, Schnur JA, et al: Computerized axial tomography with the EMI scanner. Radiology 110:109, 1974
mining associated midline displacements. The characterization and follow-up of k n o w n lesions with ultrasound has also been described. In the above instances its usefulness is usually as a noninvasive technique c o m p l e m e n t a r y to nuclear medicine studies, which under most circumstances are more effective than ultrasound. The evaluation of ventricular enlargement, which is not ususally possible per se with radionuclides, is possible with ultrasound. Computerized transverse tomography (CTT) can be applied more efficiently in a routine w a y in many of the above circumstances, including evaluation of ventricular size. The combination of CTT and nuclear medicine procedures in the brain area is so effective and comprehensive that where these are both available the demand for ultrasound will probably decrease but will not be eliminated.
U R I N G the last quarter century remarkable advances have been made in the diagnosis of neurologic disorders. Plain film x-ray examinations, electroencephalography, pneumoencephalography, cerebral angiography, radionuclide cerebral imaging, and the recently introduced technique of computerized transmission tomography (CTT) have all contributed to improved diagnostic accuracy. Both nuclear medicine and ultrasound techniques are noninvasive, safe, relatively simple, and reasonably comfortable. Each method is capable of aiding in the detection and localization of neuropathologic disorders.
D
ROUTINE ECHOENCEPHALOGRAPHIC TECHNIQUES
A-mode echoencephalography initially was introduced as a method of midline determination by Leskell' in 1956. Unlike the case for other regions of the body, where the B scan has assumed greater significance, the unidimensional A-mode display utilizing a stationary transducer has remained the more popular type of echoencephalogram in studies involving the head, because of several factors that severely hamper successful B scanning. The irregular shape of the skull inhibits continuous contact between the transducer and the scalp as the two move relative to one another. The need for continuous contact can be solved with immersion Front the Division of Nuclear Medicine, Department of Radiology, New York UniversiO' Medical Center, New York, N. Y. Charles A. Whelan, M.D.; Anthony M. Passalaqua, M.D.; Philip Braunstein, M.D.; Division of Nuclear Medicine, Department of Radiolog)', New York University Medical Center. © 1975 by Grune & Stratton, Inc. Seminars in Nuclear Medicine, Vol. 5, No. 4 (October), 1975
419
420
WHELAN, PASSALAQUA, AND BRAUNSTEIN
scanning,~ but other considerations have limited this technique. Probably the most serious limitation in all types of B scanning is the topographic distortion produced by the skull itself.3The bones of the skull are often asymmetrical and of variable thickness. The velocity of sound traveling through the skull is approximately twice as fast as in the brain, so a variable skull thickness distorts the echo pattern. The skull attenuates the sound wave irregularly, and this creates a need for greater intensity in thicker areas, which in turn further limits the resolution of two-dimensional echoencephalography.4 The skull therefore poses formidable problems for B scanning, and although certain workers 5.8 feel that twodimensional scanning is feasible and rewarding, the majority of work done in echoencephalography is with A scans. Theoretically any abnormality that could be seen on a B scan should also be recognizable on an A scan, since the former is basically a summation of a large number of unidimensional (A-scan) determinations. The routine technique for A-mode echoencephalography involves positioning the probe against the fiat thin portions of the squamous temporal and parietal bones above and slightly anterior to the pinna of the ear. By applying a small amount of acoustical gel to the scalp an efficient acoustic interface can be obtained. A theoretical midline may be obtained by through-transmission scanning (two probes) where a single echo will appear to occur halfway between the probes, since the sound wave travels the interposed distance only once. An alternative method for determining the theoretical midline is to bisect the distance between the inner-table echoes if the equipment is capable of determining near-side echoes. The theoretical midline echo (T echo) serves as a reference standard on the triple-exposed photographic documentation of the oscilloscope scans. Several structures in the median plane of the brain have the potential of reflecting the ultrasound wave. 7 The pineal gland, particularly one containing calcium, causes sharp echoes. 8 However, midline echoes can be obtained even when the pineal is absent, or in children when there is no calcification present within the pineal body. The posterior part of the third ventricle is also midline, and its walls provide a sharp acoustic interface? Depending on the actual placement of the transducer and direction of sounding, the interhemispheric fissure, septum pelucidum, and the cerebral falx may also be echo sources in bitemporai scanning. The midline echo (M echo) is.recognized as the most stable echo complex and the one that has the greatest amplitude. Because nonmidline structures, such as lateral ventricle walls, may also give strong echoes, it is best to obtain several tracings of the M echo, rotating the probe as necessary to make certain that the echo representing the presumed midline is indeed the most stable and persistent echo visualized. The midline cannot be determined accurately from unilateral sounding, since the distance from the near-side echo to the midline echo includes the thickness of scalp and bone on the side of sounding and hence is greater than the midline-far-side echo distance. Therefore the M echo is analyzed from each side of the head. A photographic record (Fig. 1) is made from which the midlinefar-side echo distances can be determined. With the transducer on the right side of the head the M echo-far-side-echo measurement represents the distance between the midline structures and the inner table of the left side of the skull. With
NEUROLOGIC DISEASES
421
Fig. 1. Normal echoencephalogram. This figure is composed of near-side, midline, and far-side echo complexes. By convention the upper trace is obtained with the transducer on the right side of the skull, and the lower inverted trace is obtained with the transducer on the left side of the skull. The farside echoes for both traces are aligned and located on the reader's right side. The midline echoes from both sides of the head coincide, indicating no midline shift.
the transducer on the left side the distance between the midline and the inner table of the right side is obtained. The photographic record of the echoencephalogram must meet certain criteria to be considered diagnostic.~° The far-side echoes, representing the inner tables on the distal side of the skull, must be aligned within 2 mm of each other on the two tracings. If the midline echo has a double-spike configuration, then measurement of the width of the Complex should not vary more than 2 mm when comparing the tracings from the two sides. Any shift of the midline complex from the theoretical midline must show identical deviation on both the right-to-left and left-to-right scans if it is to be considered verified. There is some controversy as to what constitutes an abnormal M echo displacement. Several large series 9.'.~ suggest that when the center of the midline echo complex is displaced from the theoretical midline by more than 2 mm there is pathologic displacement. A 3-mm shift is an even more definite indication of diencephalic displacement, and many authorities ~3-~5 consider a shift of less than 3 mm insignificant. With the above somewhat simplified summary of A-mode encephalography in mind, we will consider where the technique can be utilized alone or in conjunction with radionuclide techniques to deduce neurologic diagnoses. There are four major areas of interest where correlation is appropriate: (I) in the evaluation of
422
WHELAN, PASSALAQUA, AND BRAUNSTEIN
trauma and post-traumatic disorders, (2) in the detection of tumors and in the evaluation of tumor growth, (3) in the characterization of intracranial lesions, and (4) in the evaluation of hydrocephalus and the analysis of ventricular shunting procedures. HEAD INJURIES
Echoencephalography has achieved its widest application and acceptance in the clinical setting of head trauma. Subdural hematomas, epidural hematomas, intracerebral hematomas, and the edema associated with cerebral trauma have all been shown to produce sufficient mass effect to create a recognizable shift of the midline echo complex. The degree of shift is largely dependent on the location of the lesion, with collections in the same plane as the pineal gland and posterior recesses of the third ventricle producing the greatest shift. Anterior and posterior masses, even when large, often do not produce shifts of the diencephalic structures, x8 The potential of echoencephalography in the detection of cerebral midline displacement was recognized early. Both European ~.~5,1r and American 9.~8 investigators reported diagnostic accuracy rates of over 90%. Yet others xg.~° were unable to achieve such high accuracy. White ~4recognized the primary flaw in the procedure. He showed that the examination was inherently subjective and that the validity of the results was dependent on the experience of the sonographer3 x Clinical prejudice.may influence the selection of the M echo. The lack of anatomic landmarks in the A scans prevents definitive recognition of improper echo recordings on the video display. A valid screening test should be objective, and an examination that is vulnerable to recorder and interpreter bias does not meet this fundamental criterion. Brylski22 in 1965 cited the reluctance of the medical community to accept sonography as a reliable procedure. A similar statement could be made today, and the extent of utilization of A-mode echoencephalography varies considerably from one institution to another. Two ways to eliminate subjectiviiy recently have been investigated. An automatic midline computer apparatus that renders gated histograms rather than simple echo spikes can reduce operator bias. White has recorded his experience with over 3300 automatic midline examinations.23 The histograms produced are quite reliable in excluding a midline shift at the level of the third ventricle. The results are adequately objective provided that the transducer is properly positioned. False positive results (15%-20%) are a bothersome problem with the automatic midline device. Since false negative examinations are quite rare, this device may well make a contribution as a screening examination in cases of head trauma: To the extent that this technique permits an objective evaluation of one part of the cerebral midline, the automatic midline device may provide the clinician with a certain degree of confidence when there is no shift demonstrated. It is possible with more sensitive equipment to increase the amount of information available on the A scan. Recently Rothman and Gershowitz24 reported their experience with a Hoffrel 21A ultrasonoscope that possesses improved near-side resolution and obtains actual echoes from structures at or adjacent to the lateral surface of the brain. By measuring the distance from the inner dural echo to the gyral surface echo they were able to localize 13of 17 subdural hema-
NEUROLOGIC DISEASES
423
.,,
4
.
f
B Fig. 2. Bilateral subdural hematomas. (A) Echoencephalogram showing a 5-mm shift of the midline echo toward the left side. (B} Anger-camera brain scintiphoto revealing a bilateral increased peripheral activity compatible with bilateral subdural hematomas. (C) Cerebral angiogram confirming bilateral subdural hematomas with a midline shift. Note that subdural on the right side is larger than subdural on the left side.
tomas. An additional benefit to be derived from equipment rendering greater anatomic detail is recognition of bilateral subdural'hematomas. Routine midline A scans may be normal in the presence of bilateral subdurals, since there may be no significant diencephalic shift. Figui'e 2A is an example of a relatively small displacement of the midline in a patient with a large right subdural hematoma because of another smaller subdural hematoma on the left side.
,
424
WHELAN, PASSALAQUA, AND BRAUNSTEIN
Specific techniques and procedures have been well established for the evaluation of cerebral trauma in nuclear medicine. Dynamic arterial perfusion studies permit reliable detection of displaced peripheral flow in the presence of a subdural hematoma. 25-~r Although earlier series suggested that brain scans were about 80% accurate ~9in the detection of subdural collections, the routine use of delayed scans and a complementary flow study may well result in greater sensitivity,s° Two recent investigations have compared the accuracy of A-mode echoencel~halography and brain scanning in the detection of chronic subdural hematomas. Raskind et al. 3t found the echoencephalogram positive in 20 of 32 patients (63%) as compared to 35 positive brain scans in 41 patients (85%) studied. Hurwitz et al. x6showed an even greater disparity in favor of the nuclear medicine procedure, with 13 of 14 patients studied having positive scans, whereas the echo was only 44% accurate in the same gi'oup. The particularly unsatisfactory result with echoencephalography in this particular study is partially attributable to the fact that 5 of the hematomas were eccentrically located and therefore did not in fact produce any diencephalic shift. Under certain circumstances echoencephalography may make a positive contribution in evaluating patients with cerebral trauma. The information gained may complement the results of the brain scan. Brain scanning is most appropriate and reliable with chronic subdural hematomas, whereas the echoencephalogram is most often helpful in acute situations and may be a valuable tool in the emergency.ward. Echoencephalography is a more appropriate study than the radionuclide scan in a patient suspected of having an-epidural or acute subdural hematoma, since these are not likely to be detected by radionuclide scanning. When the brain scan shows a "crescent sign ''3s and the analysis of its significance is complicated by ipsilateral scalp laceration or other superficial injury, then the demonstration of coexisting diencephalie displacement will help clarify the situation. When the .clinical situation involves an adult in a nonemergent situation and an obscure diagnosis, then the greater sensitivity and objectivity of the radionuclide angiogram with delayed static scans merit utilization as the screening examination of choice. 33 But here, too, there may be occasional value in supplementary echoencephalography if the brain scan appears normal. Activity in the tempo.ral muscles hinders radionuclide recognition of a subdural collection near the temporal lobe. This type ofhematoma is most likely to produce a sizable diencephalic displacement and may result in a dramatically abnormal A scan or histogram. The echoencephalogram also may substitute for serial brain scanning in certain situations. If surgical intervention is not undertaken for a subdural hematoma, then serial echography may be utilized to evaluate increasing or decreasing diencephalic displacement. Resolution of a cerebral contusion also can be followed by means of serial echography.
TUMORS In our experience, recourse to the sonogram for the detection of tumor is virtually unknown. By collating the information obtained from brain scanning with computerized transmission tomography we have found that it is possible to detect accurately over 95% of brain tumors, 34 and given such sensitivity echoen-
NEUROLOGIC DISEASES
425
i
•
•
i
Fig. 3A. Right-sided cerebral lesion. Echoencephalogramshowing a 5 - m m shift of the midline toward the left side. Fig. 3B. Anterior (left) and lateral (right) Anger-camera brain scintlphotos with99mTc-pertech netate showing a localized area of increased activity in the right cerebral hemisphere.
cephalography has little to add. Other investigators, however, have employed echoencephalography profitably, and we shall refer briefly to their endeavors. Space-occupying lesions in certain areas of the cerebral hemispheres may be expected to cause a shift of the midline echo complex to the contralateral side (Fig. 3). However, if only midline displacement is analyzed the echo will be positive in only about 50% of brain tumors. To be effective as a screening device
426
WHELAN, PASSALAQUA, AND BRAUNSTEIN
Fig. 3C. Cerebral angioconfirming the gram presence of a right cerebral neoplasm and a midline shift toward the left side.
for brain tumors, the echoencephalogram must display additional signs. This demands an experienced sonographer and clinical competence. Tumor impingement on the normal pathway of cerebrospinal flow may result in obstructive hYdrocephalus. Evaluating ventricular size by means of echoencephalography will be discussed in greater detail below in the section dealing with hydrocephalus. The mass effect of tumors affecting the brain stem has also been described. Tenner et al. 36 outlined a technique for localizing the aqueduct of Sylvius and quadrigeminal plate when the ultrasound beam is directed from a point midway between the occipital protuberance and the vertex toward the level of the lower orbital rim. They have been able to identify anterior and posterior displacement of the aqueduct in this fashion. The tumor itself may have a sufficiently distinct composition to provide abnormal echoes as a result of the particular acoustic impedance of the tumor contents. The velocity of sound in calcified tumors is particularly rapid, and such masses may produce high-amplitude echoes. It is difficult for most investigators tO recognize a particular echo as one emanating from a tumor. Often the echoes from the tumor appear as just one part of a dazzling array of confusing reflections. This is particularly true with A scanning. When immersion techniques ~ or other sophisticated B scans are obtained the tumor echoes may be more reproducible and more easily identified. Lombroso 37 reported visualizing tumor echoes in 34% (18/44)of supratentorial tumors evaluated with intensity-modulated twodimensional B scans.
NEUROLOGIC DISEASES
427
Fig. 3D. CTT scan revealing right cerebral mass with midline shift and ventricular dilatation.
CHARACTERIZATION OF BRAIN MASSES
The potential for distinguishing solid from cystic masses by sonographic techniques has been evaluated in several other articles in this Seminar. The reliability of such ultrasonic differentiation has been established in renal masses, and the rewards in pancreatic, hepatic, and thyroid sonography may be equally great. Unfortunately no similar success has been obtained with echoencephalography. It is, however, occasionally possible to differentiate solid from cystic lesions. The mass must first be localized with other neurologic diagnostic procedures. Transonic or echo-free zones may then be detected by sounding from the opposite side of the head. The reliability of the A scan in this regard is questionable. The difficulties of obtaining transcalvarial B scans have been pointed out above. When the skull is thin or has a defect, as was the case with the patient whose scan is shown in Fig. 4A, it is possible to obtain reliable B scans. Direct sounding of the brain through areas of limited acoustic impedance has been reported by Heimburger. 38 In this situation the resolution and attenuation problems inherent in transcalvarial scanning are absent, and it is feasible to determine cystic components and tumor growth rates, as well as ventricular outlines. Similarly, resolution is considerably better when scans are obtained through the very thin skulls of newborns. Brain scans are positive in approximately 85% of primary and metastatic brain tumors. 34.39.4° Dynamic flow studies and static radionuclide scanning permit
428
WHELAN, PASSALAQUA, AND BRAUNSTEIN
L
Fig. 4A. Hydrocephalus. B scan (gray tone) taken through a right temporal cranlotomy defect shows the contour of the enlarged lateral ventricles with thinning of the cortex and no shift of the midline.
!A
limited characterization of abnormal masses in the brain. Similarly, some information about the neoplastic nature of brain abnormalities can be obtained by studying masse.~ with different radiopharmaceuticals. Except in rare instances the echoencephalogram is not reliably helpful in characterizing lesions seen by radionuclide imaging. HYDROCEPHALUS
Prior to the introduction of computerized axial tomography, the echoencephalogram was virtually the only effective noninvasive technique for evaluating ventricular size. Radionuclide brain imaging is neither informative nor appropriate in the evaluation of ventricular size. Radionuclide cisternography does provide useful information about ventricular patterns and is particularly informative in the presence of communicating hydrocephalus. Cisternography, however, does involve a lumbar puncture and is not a noninvasive screening procedure. The detection of ventricular size by A-mode echoencephalography has been
Fig. 4B. Hydrocephalus. A-mode study (reversed to correlate with B scan} shows echoes from the midline and the lateral wall of the left lateral ventricle illustrating marked ventricular enlargement.
NEUROLOGIC DISEASES
429
Fig. 4C. H y d r o c e p h a lus, C1-1"scan taken through s a m e level as ultrasound studies. Note the craniotomy defect in the right temporal area.
proposed by many investigators.9.3s.a6,"-4a Careful examination of midline echo complexes usually shows that the relatively high M-echo peak is surrounded by two smaller echo patterns. These represent the posterior part of the third ventricle and may be used to evaluate size. When the midline tracings are being sought, lateral echoes are often detected approximately halfway between the Mecho complex and the echoes emanating from the inner table. The source of these echoes may be the walls of the lateral ventricles, although the Sylvian fissure echoes occasionally cause confusion. The atrial region and the temporal horns of the lateral ventricles both produce echoes, but the two usually can be distinguished by the experienced sonographer~ The atrial echo will be detected when the beam is directed horizontally, whereas the temporal-horn reflection is noted when the beam is directed downward and somewhat posteriorly from a position 2 cm above the root of the ear. A-mode scanning can be used to measure ventricular size (Fig. 4B), but an experienced examiner must perform such an examination. The ventricles may actually be measured from the A scan or an index relating the position of the temporal horn to the midline. When the ventricles are dilated the echos are often more easily recognized, since their echo amplitudes are increased. It may be difficult to obtain good reflections from the ventricular interfaces where a tumor itself displaces the wall. This may distort the normal perpendicular arrangement necessary for sharp echo display patterns. Therefore A-mode sc~/nning has been utilized as a reliable means of detecting and following children with hydrocephalus because of the thinness of the skull mentioned previously. In the adult, hydrocephalus is probably present when the third ventricle diameter exceeds 10 •
430
WHELAN, PASSALAQUA, AND BRAUNSTEIN
m m and the lateral ventricles m e a s u r e greater than 2.8 cm. S t a n d a r d reference planes have been established for obtaining echoes from the ventricular walls. 3s Patients with ventricular shunts m a y be followed with either serial echoenc e p h a l o g r a m s or nuclear medicine studies. Serial echoes obtained f r o m identical external l a n d m a r k s will indicate progressive ventricular dilatation in situations o f shunt malfunction. Intraventricular clearance studies offer an objective radionuclide test of C S F shunt patency, R u d d et al. 4s have found that in 25 patients with n o r m a l l y functioning shunts the Tt/2 o f 99mTc-pertechnetate injected into the central reservoir (flushing device) averaged 3 min. CONCLUSION
The role o f echoencephalography in neurologic diagnosis depends to some extent upon the availability o f computerized transmission t o m o g r a p h y . W h e r e C T T is available, its superior ability to objectively evaluate the midline and ventricular size has largely eliminated the use o f A - m o d e echoencephalography. C T T is capable o f detecting acute subdural h e m a t o m a s , and the relative absorption coefficients provid e reliable characterization of the nature o f a mass lesion. 4~ The combination o f radionuclide brain scanning and C T T provides c o m plementary information with r e m a r k a b l e sensitivity, to which A - m o d e echoencephalography has little to add. When C T T is not available and in situations where the appropriate information can be obtained utilizing ultrasound, A - m o d e echoencephalography employed by an experienced u l t r a s o n o g r a p h e r may be of clinical value in detecting midline shifts, evaluating hydrocephalus in children, and following patients with C S F shunts. Finally, echoeneephalography m a y provide reassuring information to the clinician dealing with patients with acute head t r a u m a . REFERENCES
h Leskell L: Echoencephalography. I. Detection of intracranial complications following head injury. Acta Chit Seand 110:301, 1956 " 2. Makow DM, McRae DL: Two-dimensional echoencephalography using immersion scanning: Recent results, in Kazner E, Schiefer W, Zulch KJ (eds): Proceedings in Echoencephalography. New York, Springer-Verlag, 1968 3. White DN, Clark JM, Chesebrough JN, et al: The effect of the skull in degrading resolution in echo-encephalographie B- and C-scans, in Kazner E, Schiefer W, Zulch KJ (eds): Proceedings in Echoencephalography. New York, Springer-Verlag, 1968 4. Fry WJ: lntracranial anatomy visualized in-vivo by ultrasound. Invest Radiol 3:243, 1968 5. Erba G, Lombroso CT: Detection of ventricular landmarks by two dimensional ultrasonography. J Neurol Neurosurg Psychiatry 31:232, 1968 6. Lombroso CT, Erba G, YogoT, et al: Two dimensional sonar scanning for detection of intracranial lesions. Arch Neurol 23:518, 1970 7. Schiefer W, Kazner E, Kunze S: Clinical
Echoencephalography. New York, SpringerVerlag, 1968, p 38 8 . Leskell L: Echoencephalography. 11. Midline echo from the pineal body as an index of pineal displacement. Acta Chir Scand 115:255, 1958 9. Brinker RD, King DL, Taveras JM: Echoeneephalography. Am J Roentgenol 93:781, 1965 10. Sandok BA, Henson TE, Skaggs H: Analysis of echoencephalograms. Neurology 20:933, 1970 II. Schiefer W, Kazner E, Kunze S: Clinical Echoencephalography. New York, SpringerVerlag, 1968, p.49 12. Jeppson S: Echoencephalography. IV. The midline echo; an evaluation of its usefulness for .diagnosing intracranial expansivities and investigation into its sources. Acta Chir Scand [Suppl] 1961, p 272 13. Elizondo-Martel E, Gersham-Cohen J: Medical ultrasonics. Am J Roentgenol 93:79J, 1965 14. White DN, Kraus AS, Clark JM, Camp-
NEUROLOGIC DISEASES
bell JK: Interpreter error in echoencephalography. Neurology 19:775, 1969 15. Ford R, Ambrose J: Echoencephalography. Brain 86:189, 1963 16. Hurwitz SR, Halpern SE, Leopold G: Brain scans and echoencephalography in the diagnosis of chronic subdural hematoma. J Neurosurg 40:347, 1974 17. Kazner E: Recognition and differential diagnosis of intracranial complications following head injuries by means of echoencephalography, in Kazner E, Schiefer W, Zulch KJ (eds): Proceedings in Echoencephalography. New York, Springer-Verlag, 1968 18. Rothman J, Shatsky S, Kricheff I, Chase N: Ultrasonic diagnosis of subdural hematomas. Am J Roentgenol 105:413, 1969 19. Drydale LN: Echoencephalography. The use of pulsed ultra-sound in neurologic diagnosis. Audt Radiol 10:207, 1969 20. Hanser LH, tlolmes HL: EchoencephaIography in a large city hospital. Rocky Mt Med J 63:34, 1966 21. White DN: The limitations of echoencephalography. Ultrasonics 5:88, 1967 22. Brylski MD, Ibenstark JL: Anatomic localization of midline echo in s°n°grams of the brain. Am J Roentgeno193:81 I, 1965 23. White DN, Hanna LF: Automatic midline echoencephalography. Neurology 24:80, 1974 24. Rothman J, Gershowitz PS: Ultrasonic diagnosis of subdural hematomas, epidural hematomas, brain edema and brain atrophy. Am J Roentgenol 122:531, 1974 25. Fish NB, Koch RL, Pollycove M, et al: Basis for accurate scintophotographic detection of subdural hematoma. J Nucl Med 9:316, 1968 26. Rosenthal L: Intravenous and intracarotid radionuclide cerebral angiography. Semin Nucl Med 1:70, 1971 27. Beiker A: Displacement of anterior cerebral vessels in cerebral dynamic studies in cases of chronic subdural hematomas. J Nucl Med 16;86, 1975 28. Apfelbaum RI, Newman SA, Fingesser LH: Dynamics of technetium scanning of subdural hematomas. Radiology 107:571, 1973 29. Cowan R J, Maynard CD, Lassiter KR: Technetium 99m pertechnetate brain scan in the detection of subdural hematomas. A study of the lesion as related to the development of a positive scan. J Neurosurg 32:30, 1970 30. Cowan R J, Maynard CD: Trauma to the brain .and extracranial structures. Semin Nucl Med 4:319, 1974 31. Raskind R, Glover MB, Weiss SR: Chronic subdural hematoma in the elderly: A
431
challenge in diagnosis and treatment. J Am Geriatr Soc 20:330, 1972 32. Hersen WJ, Quinn JL III, Mellihan WV: The crescent pattern of increased radioactivity in brain scanning. Radiology 87:483, 1966 33. Gilday DL, Coates G, Goldenbar D: Subdural bematoma--What is the role of brain scanning in its diagnosis. J Nucl Med 14:283, 1973 34. Passalaqua AM, Braunstein P, Kricheff I, et al: Clinical comparison of radionuclide brain imaging and computerized transmission tomography. Symposium on Past, Present and Future of Non-invasive Brain Imaging. (in press) 35. Ford RM, McRae DL; Echoencephalography--A standardized technique for the measurement of the width of the third and lateral ventricles, in Grossman CC (ed): Diagnostic Ultrasound, Proceedings of First International Conference. New York, Plenum, 1966, p 117 36. Tenner MS, Wodrasha G, Adapon BD: Newer ultrasound techniques in the evaluation of neurologic disorders. Radiol. Clin North Am 12:283, 1974 37. Lombroso CT, Erba G, Yogo T, et al: Two dimensional sonar scanning for detection of intracranial lesions. Arch Neurol 23:518, 1970 38. Heimburger RF, Eggleton RC, Fry FJ: Ultrasonic visualization in detection of tumor growth rate. JAMA 224:497, 1973 39. O'Mara RE, Mozley JN: Current status of brain scans. Semin Nucl Meal 1:7, 1971 40. Gottschalk A: Brain scanning--Is it becoming unnecessarily complicated? Am J Roentgenol I 11:864, 1971 41. Schiefer W, Kazner E, Kunze S: Clinical Echoencephalography. New York, SpringerVerlag, 1968, p 174 42. Kirkpatrick J: Echoencephalography in the evaluation of hydrocephalics. Radiology 86:1052, 1966 43. Brahme F, Traghard B: Echoencephalographic estimation of size of the lateral ventricles in normal children. Radiology 92:60, 1969 44. Erba G, Lombroso CT: Detection of ventricular landmarks by two dimensional ultrasonography. J Neurol Neurosurg Psychiatry 31:232, 1968 45. Rudd TG, Shurtleff DB, Loeser JD, Neph WB: Radionuclide assessment of cerebrospinal fluid shunt function in children. J Nuci Med 14:683; 1973 . 46. Ambrose J: Computerized transverse axial scanning (tomography). Part 2: Clinical application. Br J Radio146:1023, 1973 47. New PF, Scott WR, Schnur JA, et al: Computerized axial tomography with the EMI scanner. Radiology 110:109, 1974
