Portable Gamma Camera for Clinical use in Nuclear Medicine

Transcription

1 Portable Gamma Camera for Clinical use in Nuclear Medicine R.Pani, R.Pellegrini, F.Scopinaro, F.de Notaristefani2, A.Pergola,G.De Vincentis,A.Soluri3, F.Iacopi, A.Grammatico4, A.Del Guerra Dept of Experimental Medicine and PathologyUniv. of Rome La Sapienza Italy, 1 Dept of Physics Univ. of Rome La Sapienza ITBM CNR Rome Italy, 4POL.M.TEC. Carsoli L Aquila Italy, 5Dept of Physics Univ. of Ferrara Italy 1 Abstract Up today Hamamatsu R3292 is the Position Sensitive Photo Multiplier Tube ( PSPMT) with the largest sensitive area ( 1 cm of diameter). At the same time it has the minimum size for clinical application in Nuclear Medicine. A portable gamma camera was realized, based on 5 PSPMT coupled to a scintillating array. The head has a light weight (15 Kg.) spatial resolution resulted better than that of Anger Camera with good linearity response, good energy resolution and FOV coincident with intrinsic one of PSPMT. To optimize gamma camera response two different scintillating arrays were tested: YAP:Ce and CsI (Tl). Their overall size cover all photochatode active area, and crystal pixel size was 2 mm x 2 mm. The detection efficiency resulted comparable to that of Anger Camera. The best result was obtained by CsI (Tl) scintillating : an intrinsic spatial resolution of 1.6 mm FWHM and a relative energy resolution of 17% FWHM. With a standard general purpose collimator a spatial resolution of about 2 mm resulted. Some preliminary results were also obtained in breast scintigraphy. I. INTRODUCTION Large area 5 Position Sensitive PMT [l] is the minimum size for clinical use in Nuclear Medicine. It allows to arrange a portable gamma camera with light weight head that can be very useful for a number of examinations such as: thyroid, bone, scintimammography, radioimmunosurgery. Recently scintimammography is showing the capability to detect breast cancer with very high specificity [21,[3,1,[41,[51. A dedicated gamma camera with small FOV [6],[7],[8] placed in the same geomehy of conventional Xray mammography, i.e. with breast compression can improve the detection of subcentimeter cancer and can allow integration imaging between functional and morphological diagnostics. Although advantages offered by PSPMT size first results obtained were not encouraging [9], [lo]. It was mainly due to glass photochatode window thicker and to intrinsic spread of charge wider than 3 PSPMT [ll], [12], [13]. They contribute to make worse spatial resolution in particular when planar scintillation crystal are coupled to PSPMT. The use of thick planar crystal involves a good detection efficiency but at the same time the larger spread of scintillation light distribution produces a spatial resolution worse than that in Anger camera. Furthermore a wide light distribution reduces the position linearity range of camera increasing the dead zone ring on PSPMT boundary and makes worse the energy / IEEE 117 resolution in the case of poor uniformity response of photocathode. Scintillating array maximizes the intrinsic characteristics of PSPWs. Strongly reducing the spread of light distribution it allows spatial resolution values better than 1 mm when coupled to 3 PSPMT [14], [151, [161, [171, [18], [19], [2], [21]. Regarding 5 PSPMT some preliminary results [6], [7], obtained coupling 5 PSPMT to YWCe scintillating array shown spatial resolution values of about 2 mm and a first interesting application in scintimammography. The use of scintillating crystal array represents a technological progress in the gamma ray imaging in competition with pixellated solid state detectors. Scintillation light output and spot depend on a number of factors: crystal pixel size and length, light yield, emission wavelength, surface treatment, reflective material, etc. It is very difficult to foresee the final response of energy resolution and spatial resolution from scintillating array and extensive studies are needed. With the aim to realize a portable gamma camera with better imaging characteristics than that of Anger Camera, two scintillating arrays were purposely made. Thickness of scintillation crystal and crystal pixel size were chosen on the basis of clinical application, the first one to obtain a detection efficiency comparable with that of Anger Camera and the second one to obtain a good sensitivity by matching a standard collimator hole. 11. EQUIPMENT AND METHOD The portable gamma camera consists of mechanical system with an arm supporting the head for its positioning as required by the specific application. It is equipped also with a compressor for scintimammographic use. The camera head is screened by lead 1 cm thick and in one face by tungsten.5 cm thick to minimize dead zone in the case of application in scintimammography. It has a weight of 15 Kg. Two removable parallel hole collimators are used with hexagonal hole and the following characteristics: 1.3 mm of hole aperture.2 mm of septa, 22 mm and 35 mm length for general purpose and high resolution collimator respectively. 5 PSPMT Hamamatsu is equipped with two resistive chains connecting 56 wires of anode. Four preamplifiers housed close together PSPMT with 2 ps of time constant and charge sensitivity of.5 V/pC connect the four ends of resistive chains. Four amplifiers for gain adjustment are then connected to ADC s for the digitalization of pulses. The puises were acquired in list mode through a multiparameter FAST MPA/PC consisting of a card installed on a Pc

2 ~ Pentium. The PC transfers the data to a DEC station for image processing in real time. PSPMT was selected at factory having a pulse height uniformity response of 15.5% calculated as standard deviation of mean value. The maximum pulse height variation ranged between f 3% of mean value. Two Scintillation crystal were chosen for array assembling: YmCe produced by Preciosa Crytur (Czech Republic) and CsI (Tl) produced by Hilger Analytical (Great Britain).YAP.Ce scintillation detector consists of 5 x 5 array with square shape and size of loxloxl cm3. The pixel size is 2 mm x 2 mm. The element of array is covered by reflective layer and optically separated from each other. The total dead layer between crystal elements is 1 pm. The main detection characteristics of YAPCe are [22]: absence of hygroscopicity, high atomic number (Z= 39), high density and a light yield of about 4% of NaI(T1) with a fast decay time (25 ns). Due to its hardness and the lack of cleavage plane YmCe scintillating array can be realized with element cross section down to.3 x.3 mm2 and length up to 3 mm [13]. CsI (Tl) scintillating array has a circular shape, 11 cm diameter and 3 mm thick with an epoxy ring of 5 mm. The pixel size is 2 mm x 2 mm. It has 5 crystal pixels along the diameter. The element of array is covered by diffusive white reflector (epoxy), a total dead zone of 25 pm allows a complete optical insulation between neighboring crystals. Scintillation and detection properties of this crystal are well known. Its main characteristic for this application consist of ductility that easily allows to arrange arrays with small pixel size ( down to.5 mm x.5 mm). All tests on the gamma camera were performed by a mechanical scanning system with an accuracy better than 1 pn. To analyze the intrinsic characteristics of the camera a collimated CO line source was used. 12 measurements were performed with a scanning step of 3 mm along the two crystal axes defining an array of 33 x 33 irradiation spots. Pulse height uniformity response of gamma camera was also performed by flood field measurements and 57C free source at 4 m distance from detector. The uniformity matrix is able to equalize point by point differences in pulse height introduced by scintillating array, by photocathode quantum efficiency and by charge multiplication. The list mode file produces a list of positions and energy channel, for each event. Applying the pulse height uniformity matrix during image processing is possible to select the same energy window in all image points, taking into account different pulse height response. Storing all digitized pulses of flood field image it was possible to process data, calculating the mean energy value of spectrum detected on each pitch. The size of pitch was fixed comparable with crystal pixel size. Furthermore the flood field was utilized to correct images for spatial uniformity response introduced by different spatial resolution values and non linearity response RBULTS For a better understanding of different imaging behavior of two scintillating arrays, the results are listed and discussed following the main gamma camera characteristics. A. Position Linearity It has a good position linearity response along the whole crystal dimension (1 cm). The slope of position linearity is constant within 1% in a range of 8 mm. A changing of slope of about 4% resulted on PSPMT boundary. To test the constancy of position linearity slope in the whole active area, all rows and columns of the image obtained from irradiation spots were analyzed. Within a circle of 6 mm of diameter, the values of position slope: resulted constant within 3%.In the remaining active area they disagreed within 15%. 2) Csl(T1) Due to spatial resolution less then pixel size, The resulted image by flood field irradiation was the Same obtainable by a light spots scanning of PSlPMT with a pitch of 2.25 mm. It was well visible the intrinsic position linearity of PSPMT where 42 crystal pixels on 44 (defining the active diameter) were recognizable. A strong position compression was visible, in the ring outside active area. Fig. 1 Flood field image by s7co irradiation obtained from CsI(T1) scintillating array B. Spatial Resolution The intrinsic spatial resolution values ranged meanly between 2. mm and 2.7 mm. 2) Csl (TI) In fig 1 is shown a particular of flood field image. Crystal pixels are well visible. Thle intrinsic spatial resolution was calculated as FWHM of position peak, taking into account distance between two pixel axes (2,25 mm). The intrinsic spatial resolution ranged between 1.6 mm and 1.8 mm.. C. Energy Resolution The normalized energy resolution of a single crystal array element resulted between 3% and 45%. The result obtained by pulse height correction is shown in fig. 2 where the energy spectrum is relative to a flood field irradiation. The normalized energy resolution, of detected spectrum, resulted 1171

3 87%. A strong improving was obtained by pulse height correction (57%), however energy resolution resulted still affected by PSPMT non uniformity response. Position indetermination affects spectrum unfolding, it is mainly due to spatial resolution and to energy transport following the Compton interaction process. 2) CsI (Tl) Light output resulted 4 times higher then YAP:Ce. The energy resolution (FWHM) of single spot irradiation resulted about 17% as shown in fig. 3. It is an excellent result in comparison with literature data obtained by NaI(T1) flat crystal coupled to PSPMT (13% ~16% at 122 kev) [9], [lo],[23], [24]. Final result obtained after pulse height correction of flood field spectrum was also very good, producing relative energy resolution of 2% FWHM. These results is mainly due to the very good uniformity response of array, to the good energy resolution of all pixels and to the distribution of position events very close to pixel size that makes more accurate the correction of pulse height. D. Detection EfJiciency In table I, the detection efficiencies of scintillating arrays in comparison with NaI(T1) crystal of Anger Camera are shown. Detection efficiency were calculated by a Monte Carlo code. The detection efficiencies calculated on the basis of Gaussian distribution of position events, generated by a spot irradiation, resulted comparable for both scintillating arrays. The difference between full energy peak efficiency an position peak efficiency is due to a background produced on the image by Compton reabsorption events in YmCe detector ( 1% of full energy peak). $7. 26 e m ;4c 5% Fig. 2 CO spectra detected by YAPCe array A) Flood field. B) Flood field corrected by pulse height uniformity response Figd YAPCe scintillating array : =Tc wire phantom image 1,oo,9,8 A.z,7 c =I,6 e 9,5 5 v) v,4 E,3 g,2,lo,oo Fig. 3 C57 spectra detected by CsI(TI) array A) Flood field. B) Flood field corrected by pulse height uniformity response. C) 1 mm diameter collimated spot Fig.5 CsI(TI) scintillating array : =Tc wire phantom image E. Clinical Application In fig. 4 and fig5 are shown the images, obtained by YAPCe and CsI(Tl) arrays respectively, of a wire phantom. They were obtained by a parallel hole collimator with hexagonal hole 1.3 mm size and 22 mm length with.2 mm septum. The phantom consists of 2 cotton wires, 5 mm apart imbibed with %Tc and 1 cm length. In fig. 6A and fig. 6B are shown relative image crosssections. Image from CsI(Tl) array has less contrast due to the lack of alignment between collimator hexagonal holes and square pixels of scintillating array. In fact, each collimator hole can irradiate a number crystal elements of array. Due to the spatial resolution less than pixel size many events produced in the boundaries of neighboring crystals resulted shifted in position determination. This effect is less evident 1172

4 in YAPCe array because of spatial resolution greater than pixel size resulted Finally a clinical image of breast carcinoma was obtained by YAPCe array. Briefly, 2 hours after i.v administration of 37 MBq of wmt~ MIBI, a patient affected by breast masses previously assessed by mammography, was submitted to a scintimammography by means of the described detector that permitted us to obtain the same radiological views. To give an idea of the influence of energy resolution and pulse height uniformity on scintimammographic image, in fig.7 are shown the spectra obtained by 'u'ap:ce scintillating array from breast analysis with and without pulse height correction. The little peak visible in the unhlded spectrum represents the true gamma emission of breast OC 1 l n g g 5 4 V g 3 c a, oozo Ch an ne1 s Fig. 7 YAl?Ce scintillating array :Uncorrected spectra of breast image and corrected 4 a I I I I, Fig 6 Image cross section of wire phantom obtained by CsI(TI) (A) and YmCe (B) scintillating arrays. Table I Detection Efficiencies Scintillation Photofraction Thickness Pixel size Detection urn Detection efficiency Spatial Resolution Image Full energy peak (FVWM) mm background < 1% 72% 2.3 < 2% I IV. CONCLUSION CsI(TI) scintillating array seems to be the best coupling to 5" PSPMT. The final detection characteristics of a portable gamma camera were: good position linearity, very good spatial resolution (less than 2 mm FWHM), good energy resolution ( 2% FWHM).l%rthermore FOV coincides with intrinsic active area of PSIPMT minimizing the dead zone ring. This Camera,can open a new field of clinical application, in in the scintimammographic imaging. 1173

5 V. ACKNOWLEDGMENTS [12] H. Kume, S. Suzuchi, J. Taleuchi, and K. Oba, Newly developed photomultiplier tubes with position This work was sup~rted the National Institute for sensitivity capability, IEEE Trans. on Nucl. Sei., Vol. Nuclear Physics (1 ) in Italy, the Italian Ministry for NS32, , (1985) Research (MURST) and the Italian Association for Cancer E131 HAMAMATSU Technical Data, Position sensitive Research (ARC) photomultiplier tubes with a crossedwire anode Hilger Analytical collaboration and support is gratefully acknowledged REFERENCES [l] HAMAMATSU Technical Data Sheets R3292 Oct 1988, tentative data, Printed in Japan. [2] Khalkali I, Mena I, Jouanna E. et al. Prone scintimmmography in patients with suspicion of carcinoma of the breast. J.Am.Coll.Surg 1994; 178: [3] Taillefer R, Robidoux A, Lambert R, Twin S, LaperriCre J. Technetium99m sestamibi prone scintimammography to detect primary breast cancer and axillary lymph node involvement. J.Nucl.Med 1995; 36: [4] Scopinaro F, Schillaci, Ussov W et al. Accuracy prone scintimammography (SM): a threecenter study on 35 patients. Eur.J.Nucl.Med. 1996; 23 (9): 191 (abstr) [5] Carvalho PA, Chiu ML, Kronauge JF. Subcellular distribution and analysis of technetium99m.mibi in isolated perfused rat hearts. J.Nucl.Med. 1992; 33 (8): [6]. R. Pani, F. de Notaristefani, F. Scopinaro et al. Single Photon Emission mammography (SPEM): a new scintigraphic technique. Eur.J.Nucl.Med. 22(8) 876 (1995) [7] F. Scopinaro, R. Pani, R. Pellegrini. SPEM: a dedicated camera for scintimammography. Quarterly Journal of Nuclear Medicine Vol. 2 (1996) [8] R.Pani, R.Pellegrini, F.Scopinaro et al. Scintillating array gamma camera for clinical use. 4th. International conference on Position Sensitive Detectors. University of Manchester, 913 September (1996) [9] ZHe, AJ. Bird, D.Ramsden, Y.Meng. A 5 inch diameter positionsensitive scintillation counter. IEEE Trans. Nucl.Sci., Vol4 No 4) (1993) [ 11 A.Truman, AJ.Bird, D.Ramsden, Z.He. Pixellated CsI(Tl) arrays with positionsesnsitive PMT readout. Conf. Proc. for Eight symposium on Radiation Measurements and Applications, May 1619 (1994) [ll] H. Kume, S. Muramatsu, H. Iida, Position sensitive photomultiplier tubes for scintillation imaging, IEEE Trans. Nucl. Sci., Vol. 33, No. 1,359363, (1986) R2486 series, Aug Rev., Supersedes Oct 1987, CR 2, Printed in Japan 141 K. Blazek, F. De Notaristefani, P. Maly et al YAP.Ce Multicrystal Gamma Camera Prototype,IEEE Trans Nucl. Sci., Vol. 42, No. 5,October F. de Notaristefani, R. Pani, L. M. Barone, et al. Light Yield and response function on YmCe Multicrystal detectors. Proceedings of the International Conference on Inorganic Scintillators and their applications Scint 95. Delft,The Netherlands, (1995) [16] Pani R., F.Scopinaro, G.Depaola et al Very High resolution gamma camera based on position sensitive photomultiplier tube, Physica Medica Vol.IX N. 23, (1993). [17] R.Pani, K. Blazek, F. De Notaristefani et al. Multicrystal YAPCe Detector System For Position Sensitive Measurements. Nucl. Instr. & Meth. A348), 551, (1994) [U] L.M.Barone, K.Blazek, D.Bollini et al. Toward a Nuclear Medicine with submillimiter spatial resolution. Nucl. Ins@. & Meth. Vol. A (1995) [19] L.M. Barone, K. Blazek, D. Bollini et al. A Detector for Submillimiter Gamma Cameras. Nuclear Physics B (Proc.Suppl.) (1995) [2] F. de Notaristefani, R. Pani, F. Scopinaro et al. First results from YmCe gamma camera for small animal studies. IEEE Trans. Nucl. Science 43 5 (1996) [21] R. Pani, F. de Notaristefani, L.M. Barone et al. Performance studies of YmCe crystal pillars for imaging application in Nuclear Medicine. IEEE Conference Record (1995) [22] S. Baccaro, K. Blazek, F. De Notaristefaniet et al, Scintillation properties of YAP:Ce, Nucl. Instr. and Meth., Vol. A361,29215, (1995). [23] JM. Poulsen, R.Verbeni, F.Frontera. Positionsensitive scintillation detector for hard Xrays. Nucl. Instr. Meth. VolA3 1,39842, (1991) [24] JS. Gordon, RH. Redus, V.Nagarkar and MR. Squillante. New uses of position sensitive photomultiplier tubes. SPlE Vol GammaRay Detectors, 187 (1992) 1174