259

I I i ~

Evaluation of Prototype Equipment for Digital Subtraction Angiography in Diagnosing Intracranial Lesions

Mutsumasa Taka~ashi , 1 :oshihisa Hirota,1 Hiromasa Bussaka,1 Tadatoshi Tsuchigame ,1 Masayuki Miyawaki,1 Sukeyoshl Ueno, Yasuhlko Matsukado,2 and Yoshibumi Hirata2

Two prototype units for digital subtraction angiography have Subjects and Methods been developed by Japanese manufacturers. Their performance in demonstrating various intracranial lesions in 121 patients was DSA was perform ed on 121 patients with known or suspected evaluated. Comparison with conventional selective catheter an­ intracran ial lesions. In 77 cases conve ntional cerebral angiography giography was possible in 77 patients. Digital subtraction an­ was performed within 2 weeks of DSA. Most cerebral angiograms giography proved useful in preoperative evaluation of pituitary were obtain ed usin g transfemoral catheter techniques with biplane tumors when transsphenoidal surgery was being considered; stereoscopi c magnifi cati on. demonstration of faint tumor stains which might be missed on Patients were denied food and water before the procedure. In conventional angiography; demonstration of larger aneurysms, most cases, a 16 gauge, 15-cm-long Tefl on needle was inserted in arteriovenous malformations, and tumors as a screening tech­ the antecubital ve in . Wh en the thoracic artery and its branches nique; demonstration of abnormalities involving the dural si­ were examined together or when placement of the needle in the nuses; and follow-up of patients after surgery or embolization. antecubital vein was un successful, an angiographic catheter (5.5 Fren ch) was in serted in th e femoral ve in using the Seldinger tech­ nique and advanced under flu oroscopy into the inferior ve na cava The usefulness of digital subtraction angiography (DSA) has been near the ri ght atrium . Forty ml of 76% meglumine and sodiu m established for evaluation of the aorta and its primary branches, diatrizoate solution was injected mechani call y at a fl ow rate of 12- especially for delineation of the extracranial carotid and vertebral 15 ml /sec, usin g a pressure injector. A 20- 25 ml solution of 5% arteries [1-6]. Several recent reports have demonstrated the value glucose or saline was layered over the contrast medium. At 5-8 of DSA in the intracranial vascular system [7 -10]. Japanese man­ sec after initial injection , seri al im ages were obtain ed at a rate of ufacturers have developed two prototype units for DSA . We have one/sec for 20 sec or two/ sec for 5 sec, foll owed by one / sec for evaluated their clinical usefulness in diagnosing intracranial dis­ 10 sec. The beginning of the exposure varied according to heart eases. rate. Filming of the first seri es of images in each patient generall y began sli ghtly earli er than filmings of later seri es, which were ad justed after review of the first se ri es. Equipment and Technique Straight anteroposterior and lateral projecti ons were obtained routinely. Caldwell projections were frequently useful for delineati on Two DSA imaging systems (Hitachi and Shimadzu, respectively) of the horizontal portion of th e anterior and middle cerebral arteries. were used in this study [11]. The major components and capabilities Towne projections were used for posterior fossa lesions. Sli ghtly of each unit are outlined in table 1 . Both units have U-arm assem­ oblique lateral projecti ons were obtained to avoid overl apping or blies th at can position the x-ray tube and im age intensifier for any superposition of th e carotid siphon. The patient with normal renal desired projection while maintaining the patient in a comfortable function can withstand three or four injecti ons. position on a floating table top. The results of DS A were graded as: (1) excell ent, when findings The output signal of the image intensifier-television camera sys­ were comparable or superi or to th ose obtained by conventional tem is digitized by an analog-digital converter in which digital angiog raphy; (2) good, when useful informati on was obtained, but information is formed on 512 x 512 pi xels, 8-9 bits deep. Digital with sign ificant chance of error; (3) fair, when no diagnostic infor­ information is fed into the im age processing assembly after loga­ mati on was obtained, although intracranial vessels were visualized ; rithmic amplification, where two to eight frames are combined and and (4) poor, when the study was un successful or uninterpretable. su btracted from the mask images to form the final digital im ages. Up to 32 serial digital images can be stored in real time in the frame memories, while additional images are transferred to other record­ ing units as necessary (Hitachi) or recorded on a video disk record er Results (Shimadzu). Using the host computer, image manipulations such as remasking, edge enhancement, image smoothing, image mag nifi­ In 18 (14.9%) of 121 patients, DSA was judged excell ent or as cation, r-adjustment, and changing of window level and width can diagnosti c as conventional angiog raphy for evaluati on of intracra­ be carri ed out. ni al lesions. In 17 cases (14%), studies were technicall y un suc-

, Departm ent of Radiology, Kumamoto University Sc hool of Medicine, Hon jo 1-1-1 , Kumamoto, Japan 860. Address reprint requ ests to M. Takahashi . 2 Departm ent of Neurosurg ery, Kumamoto Unive rsity School of Medicine, Kumamoto, Japan 860. AJNR 4:259-262, May / June 1983 0195- 6108/ 83 / 0403- 0259 $00.00 © American Roentgen Ray Society 260 DIGITAL RADIOGRAPHY AND DSA AJNR:4, May/ June 1983

cessful or uninterpretable. In th e great majority (70.8%) of cases, Cerebral Aneurysms r es ult ~ of DSA fell between th ese two extremes (table 2). Most Larger cerebral aneu rysms and their necks were clearly dem­ " poor" examinations were secondary to misregistration artifacts onstrated by DSA, which was as diagnostic as conventional angiog­ from pati ent motion and technical difficulties encountered during raphy (fig. 3). Th ere was considerable difficulty in diagnosing th e early phase of clinical trials. smaller aneurysms, particularly when the diameter of the aneurysm was less th an twice th e diameter of th e parent artery. Arteri al spasm second ary to subarachnoid hemorrhage was ob­ Brain Tumors served with some difficulty, although th eoretically such physiologic Larg er tumors and vascular tumors were diagnosed easily by activity should be more readily demonstrated by DSA. DSA could DSA, whereas smaller tumors (diameter <3 cm) and avascular be repeated easily to monitor such conditions. tumors were difficult to diagnose. Tum or stains were clearly demonstrated by DSA. Meningiomas in th e suprasellar area, olfactory groove, and planum sphenoidale Arteriovenous Malformations were better delineated by DSA th an by conventional angiography Arteriovenous malformations were demonstrated to good advan­ (fig. 1). This was due partly to higher contrast sensitivity and partly tage. The feeding arteri es and draining veins were visualized with to bilateral blood supply to th e tumors, both sides being opacified DSA as well as by conventional angiography (fig. 4). Eve n a 1 cm during DSA. arteriovenous malformation was clearly shown. The superior sagittal sinus was clearly visualized with DSA and in vasion of th e superi or sagittal sinus was demonstrated in two convexity meningiomas (fig. 2). Arterial Occlusive Diseases In patients with intrasell ar tumors, DSA provided enough infor­ mation to recommend transsphenoidal hypophysectomy. Arterial occlusions and stenoses of more than 50% were dem­ onstrated in larger vessels such as the internal carotid arteries, horizontal parts of the anterior and middle cerebral arteries, and basilar and proximal posterior cerebral a rt e ri e~ (fig. 5). Occlusions TABLE 1: DSA Prototype System Components and Capabilities in small er vessels and stenoses of less than 50% were not demon­ strated with certainty. Coll ateral vessels were often difficult to M anufacturer Component observe and only th e recanalized parts of th e arteries were visual­ Hitachi Shimadzu ized. Generator 150 kVp, 1,300 mA 150 kVp, 1,000 mA X-ray tu be 450 kH, 0 .8 / 1.3 400 kH , 1.0 mm fo- mm focus cus Other Intracranial Lesions Image intensifier 22.9 cm, 17.8 cm 22.9 cm Csi Csi A case of carotid-cavernous sinus fistula was clearly demon­ Gri d 8:1 , 28 line pairs/ 10:1, 32 line pairs/ strated with visualization of th e site of th e fistula and enlarged cm, wood cm, glass fiber superior ophthalmic vein. Vascular anomalies such as azygous TV camera Plumbicon, 525 Plumbicon, 525 anterior cerebral artery, persistent primitive arteries, and ven ous lines, 5 MHz, lines, 5 MHz, sin us anomali es were also demonstrated. A large hypertensive SNR 650:1 SNR 450:1 intracerebral hemorrhage was diagnosed. Amplification Logarithmic or lin- Logarithmic or lin- ear ear A/ D converter 9 bit, 20 MHz 8 bit, 10 MHz Postoperative Evaluation Di gital memory 512 x 512 x 16 512 x 512 x 13 bits ( x2) bits ( X2) Patients with brain tumors, arteriovenous malformations, and 512x512x10 cerebral aneurysm were evaluated by postoperative DSA. TI e bits (18-30) procedure was also used to evaluate anastomosis between the Nole.-Csi = cesium iodide: SNR = signal-Io-noise ralio: AI D = analog-la-digital. superficial temporal artery and middle cerebral artery. Patency or

TABLE 2: Diagnostic Usefulness of DSA in Demonstrating Intracranial Disease

No. (% ) Pathotogy Excellent Good Fair Poor Totat;; Brain tumor 10 16 7 9 42 Cerebral aneurysm 4 2 7 Arteriovenous malformation 2 1 3 Arterial occlu sive disease 2 15 11 3 3 1 Cerebrovascular anomaly 1 4 5 Hypertensive intracerebral 3 2 5 hemorrhage STA-MCA anastomosis 1 1 3 Other intracranial lesions 15 6 4 25 Totals . . 18 (14.9) 54 (44.4) 32 (26.4) 17(14.0) 12 1

Note.-STA-MCA = superfi cial lemporal artery to midd le cerebral arlery. AJNR: 4, May/ June 1983 DIGITAL RADIOGRAPHY AND DSA 261

Fig. 1.-Subfrontal meningioma, lateral DSA. Cath eter angiography demonstrated enlargement of ophthalmic and ethmoidal arteries but failed to show IUmor stain . A, Arterial phase. Proximal part of an­ terior cerebral artery is displaced superoposteriorly. Ophthalmic artery is not shown. B, Venous phase. Dense tumor stain in anterior fossa with drainage into cavern ous sinus via uncal vein .

A 8

Fig. 2. - Smali frontal meningioma, oblique lat­ eral DSA. Catheter angiography revealed tumor in superior left frontal lobe near coronal suture, where radiating vessels were supplied by left middle men­ ingeal artery; no definite tumor stain was noted. DSA sh ows occlusion of superior sag ittal sinus with for­ mation of venous coliaterals inferi orl y over frontal lobe (arrow).

Fig. 3. - Cerebral aneurysm, anteroposterior DSA. Selective carotid angiography was successful only on left. DSA shows saccular aneurysm overl y­ ing anterior communicating artery and basilar tip. Ve rtebrobasilar system and right carotid artery are without aneurysms. 2 3

4A 48 5

Fig. 4.-Arteriovenous malformation, lateral DSA. A, Arterial phase. Extensive arteriovenous malfor­ Fi g. 5. - Cerebral infarction , anteroposterior malion in right parietal area, fed by pericaliosal and caliosomarginal arteries. Middle cerebral artery also DSA. Occlusion of ri ghl middle cerebral artery from con tributes to malformation. B , Venous phase. Drain age via vein s of Trolard and Labbe toward superior its origin (arrow). Leplomeningeal coli aterals from sagittal and lateral sinuses. posterior cerebral and anterior cerebral arleries. 262 DIGITAL RADIOGRAPHY AND DSA AJNR:4, May/ June 1983

stenosis of the anastomosis was vi sualized, but sli ght stenosis or DSA can be performed expeditiously and noninvasively. With slightly compromised flow could not be demonstrated. DSA was further technical development, it will become an important diagnos­ often useful in evaluating patients who had undergone embolization. ti c tool in the pathologic evaluation of the central nervous system.

Discussion REFERENCES The advantages of DSA have been emphasized by many authors (2-4, 7, 12]. DSA is safer and less invasive than conventional 1. Chilcote WA, Modic MT, Pavlicek WA, et al. Digital subtraction cerebral angiography since arteri al catheterization or direct arterial angiography of the carotid arteries: a comparative study in 100 puncture is not necessary. The procedure can be performed easily patients. Radiology 1981 ; 139: 287 -295 on patients who are in poor condition or of advanced age. In 2. Christenson PC, Ovitt TW, Fisher HD, Frost MM, Nudelman S, addition, it can be performed on an outpatient basis. The cost of Roehrig H. Intravenous angiography using digital video sub­ DSA is lower than that of conventional cerebral angiography. Real­ traction: intravenous cervicocerebrovascular angiography time images can be obtained, and repeat and / or follow-up exarni­ AJNR 1980;1 : 3 79-386, AJR 1980; 135: 1145-1152 nati ons are easily performed. 3. Meaney TF, Weinstein MA, Buonocore E, et al. Digital subtrac­ Because of better low-contrast sensitivity, faint tumor stains are tion angiography of the human cardiovascular system. AJR better demonstrated by DSA than by conventional angiography (8, 1980; 1 35: 11 53-11 60 10]. Tumor stains of meningioma and metastasis were well delin­ 4. Ovitt TW, Christenson PC, Fisher HD III , et al. Intrave n ou ~ eated in this study. The dural sinuses are better visualized by DSA angiography using digital video subtraction: x-ray imagin £1 because opacified blood drains into the sinus bilaterally. For this system. AJNR 1980;1 :387-390, AJR 1980;135: 1153-11 6(, reason, DSA may be th e procedure of choice for imaging sinus 5. Strother CM, Sackett JF, Crummy AB, et al. Clinical applica occlusion or stenosis (7,8, 10]. tions of computerized fluoroscopy: the extracranial caroticl DSA has more physiologic potential than conventional angiog­ arteries. Radiology 1980; 1 36: 78 1 - 7 83 raphy, since direct arteri al injection of contrast media and catheter 6. Turski PA, Strother CM , Turnipseed WD, et al. Evaluation 1 placement in the artery are unnecessary. Because all intracranial extracranial occlusive disease by digital subtraction angiogra vessels are simultaneously opacified, more physiologic observation phy. Surg Neuro/1981 ;16: 394- 398 of cross-filling and coll ateral flow can be made in occlusive diseases 7. Carmody RF, Smith JRL, Seeger JF, Ovitt TW, Capp MP (8]. Intracranial applications of digital intravenous subtraction an­ Although DSA may not provide as mu ch detail as conventional giography. Radiology 1982; 144: 529-534 angiography, the information obtained may be sufficient for deter­ 8 . Modic MT, Weinstein MA, Chilcote WA, et al. Digital subtraction mining the extent and vascularity of cerebral tumors, especially angiography of th e intracranial vascular system: co mp a r a ti v '~ when DSA data are correlated with computed tomographic (CT) study in 55 patients. AJNR 1981 ;2: 527 -534, AJR 1982; findings (11]. For example, intrasellar masses can be evaluated 138:299-306 with DSA and CT, eliminating the need for conventional angiogra­ 9 . Seeger JF, Weinstein PR, Carmody RF, Ovitt TW, Fisher HD, phy, since the primary role of angiography in these cases is to Capp MP. Digital video subtraction angiography of the cervical exclude the possibility of existing aneurysm in the parasellar area. and cerebral vasculature. J Neurosurg 1982;56: 173-179 DSA has certain limitations (3, 4, 11]. Limited spatial resolution 10. Takahashi M , Bussaka H, Nonaka N, Miura Y, Hirata Y, Ma - and misregistration artifacts may produce undiagnostic or uninter­ sukado Y. Digital fluoroscopic angiography in central nervous pretable images. The procedure cannot be used successfully for system (in Japanese). Neurol Med Chir (Tokyo) (in press) evaluation or detection of small lesions. With technical advance­ 11 . Takahashi M, Hirota Y, Tsuchigame T, et al. Clinical value of ments, however, these difficulties are likely to be overcome to some digital fluoroscopic angiography (in Japanese). J Med Imagi(,g extent. The overlapping or superposition of vessels on DSA images 1982;2 : 660-672 is also a disadvantage. Slightly oblique lateral projections are 12. Strother CM , Sac kett JF, Crummy AB, et al. Intravenous vi dpo frequently useful in overcoming this handicap. The amount of con­ arteriography of the intracranial vasculature: early experienc') . trast medium currently required for DSA is large, but biplane DSA AJNR 1981 ;2: 215-218 may soon solve this problem.