Abstract. Investigation of patient dose from femoral angiography requires detailed logging of dose-area product (DAP) measurements, which is time-consuming ...
1996, The British Journal of Radiology, 69, 1159-1164
Patient dose measurements from femoral angiography 1
P R HOSKINS, PhD, FIPSM, 2 I GILLESPIE, FRCSE, FRCR and 2 H M IRELAND, MRCP, FRCR
Departments of Medical Physics and 2Medical Radiology, Royal Infirmary, Edinburgh EH3 9YW, UK Abstract Investigation of patient dose from femoral angiography requires detailed logging of dose-area product (DAP) measurements, which is time-consuming and only possible on small numbers of patients. A simple model for the femoral angiogram study consists of two regions; lower limbs and torso. The experimental phantom was the abdominal and pelvic sections of a Rando phantom, and two 10 cm diameter water filled cylinders to represent a lower limb. DAP rate during screening, and the DAP per spot film exposure were obtained in two rooms. Total DAP values were measured on 100 patients in each room. The median screening time and number of spot film exposures were divided between the two regions. The DAP from screening and spot films for each region was estimated by combining the phantom and patient data. The total DAP predicted by the model agreed to within 7% of the median DAP from the patient studies. Conversion to effective dose gave 9.0 mSv for the newer room compared with 2.8 mSv for the older room. In the newer room it was estimated that digital spot film exposure contributed 88% of the total effective dose. In the same room, exposure of the torso contributed 98% of the total effective dose. The model will enable interpretation of total DAP measurements made from femoral angiogram studies without the need for detailed DAP measurements on every patient. Attempts to reduce patient dose from femoral angiography must concentrate on reduction of the number and dose per exposure from abdominal and pelvic digital spot films.
The number of angiography studies has increased dramatically in recent years. This is largely associated with the ability to perform interventional procedures. Patient doses have been measured for a variety of angiographic studies including cerebral [ 1 - 3 ] , abdominal [1, 3], coronary [3-5] and femoral [ 1,3,6,7]. Thermoluminescent dosimetry (TLD) measurements on patients undergoing angiography are time-consuming and have only been used to provide estimates of the effective dose to specific organs such as the thyroid [ 2 ] and testes [ 7 ] . Effective dose equivalent [ 8 ] or, more recently, effective dose [ 9 ] have been estimated from measurements of dose-area product (DAP). If standard radiographic projections are used then published factors, based on Monte Carlo simulations, may be used to convert DAP to effective dose equivalent [10] or effective dose [11,12]. Femoral angiography is more complex as it involves exposure of the abdomen, in which there are a number of radiosensitive organs with high weighting factors, as well as exposure of the lower limbs which have little radiosensitive material and a low weighting factor. The first step which must be taken in estimation of effective dose from femoral angiography is partition of the DAP between different regions of the body. Full information on DAP would be ideally obtained by recording the cumulative DAP after each spot film exposure and each period of fluoroscopy. Recording of DAP values in this way is best done using a computerized
Received 14 June 1996 and in revised form 6 August 1996, accepted 13 August 1996.
Vol. 69, No. 828
system. However, it has been shown that the most commonly used DAP meter in the UK, the Diamentor M2, has limitations in its ability to assess accurately DAP from screening, and to count the number of spot films taken [13]. Another way to obtain these data is for the observer to note in a log book the cumulative DAP values at times throughout the procedure. This is extremely time-consuming and is only suitable for research studies on small numbers of patients. In this paper a technique is described whereby an assessment may be made of the contribution to the total DAP of screening, and of DAP values from different anatomical regions during femoral angiography, using a combination of phantom and patient measurements. The method is applied to femoral angiogram studies made in two vascular X-ray rooms. Methods Instrumentation Measurements were made on two dedicated vascular X-ray systems; an old system (IGE DF3000 installed in 1984) and a new system (Philips Integris 3000 installed in 1994). The older unit has a 22 cm intensifier and the newer unit a 38 cm intensifier. In the patient and phantom studies reported below, both units were used on the maximum field size. Both systems perform digital subtraction angiography (DSA) and, in addition, the older unit has a serial film changer. For both X-ray units DSA exposures are based on a peak detecting algorithm, with the peak level determined by initial manual exposures. For the older unit during screening the initial tube potential is selected manually, the spot film frame rate and 1159
P R Hoskins, I Gillespie and H M Ireland
tube potential are set manually, and mAs is under automatic brightness control. For the newer unit screening tube potential and tube current are automatically adjusted, and for DSA the user selects programmes; for example lower limb, abdominal aorta etc. These control the spot film frame rates, tube potential and delay. A DAP meter (Diamentor M2, PTW Freiberg, Germany) was used to obtain readings from the older X-ray unit. The newer X-ray unit calculates DAP rate from the calibrated output based on the tube potential and tube current using a knowledge of the collimator jaw position. Both DAP meters were calibrated according to a modification of the recommended procedure [14]. The procedure consisted of estimation of the true DAP from measurement of air kerma using a dosemeter (Radcal 2025, Monrovia, California, USA) and beam area from an exposed film at the same plane. In the National Radiological Protection Board (NRPB) procedure it is recommended that the ionization chamber be placed just above the couch top. In this study the ionization chamber was placed midway between the intensifier and the couch top in order to reduce backscatter to a minimum. Measurement of total dose-area product for patients There were sufficient femoral angiogram studies to provide meaningful comparison between the two X-ray units. Only patients in which there was examination of both the aorta and lower limb were included. Aortogram-only patients were excluded, as were any studies involving angioplasty. The study consisted of an initial digital aortogram. For the old X-ray unit this was followed by serial film exposures of the lower limb from the aortic bifurcation to the ankles involving four table positions, with subsequent digital subtraction used to fill in gaps, or to image in another plane for clarification as required. For the newer X-ray set the aortogram was followed by digital runs at specific overlapping sites down the leg. As part of a routine dose audit programme DAP readings were taken for patient studies performed on the older unit between January 1992 and April 1993. Details had been recorded of the number of spot film exposures taken, and of the total screening time as recorded from the X-ray console. Similar information was recorded from the newer unit between January and August 1995. Estimation of the dose-area product from screening and spotfilms A simple model was used to estimate the contribution to the total DAP from screening and spot film exposure, from lower limb and abdominal exposures. The model assumes that the examination consists of two regions: (i) lower limbs, and (ii) abdomen/pelvis. For example, to estimate the typical DAP from screening of the lower limb two inputs are required; the typical screening time, and the DAP rate (Gy cm2 s" 1 ) for a typical lower limb. For each X-ray set a typical study was drawn up on the basis of discussion with the radiographer and the median values for numbers of exposures and screening times from the patient studies. 1160
The experimental phantom consisted of a Rando phantom (Phantom Laboratory, New York) to simulate the torso, and two water filled polyethylene cylinders of 10 cm diameter and 40 cm length to simulate the lower limbs, and this was used with each X-ray unit. For the lower limb phantom a tapered aluminium filter was placed underneath the point where the two cylinders touched in order to reduce flaring. This is the same filter provided with the X-ray unit and used in patient femoral studies. With the assistance of an experienced radiographer the phantoms were positioned, appropriate collimation applied, and X-ray exposures were made. For the Rando phantom collimation was necessary only for the abdominal projection viewed on the newer unit, due to flaring beyond the lateral edges of the image. Lateral collimation was necessary for the lower limb phantom on both X-ray units. For the torso, measurements with the Rando phantom were made using two projections. The abdomen was imaged with the field of view centred on the fourth lumbar vertebrae; this projection being used to image the aorta. Secondly, the pelvis was imaged with the field of view centred on the pelvic inlet; this being the projection used for imaging of iliac vessels. For each projection DAP readings were taken during 1 min of fluoroscopy and during spot film exposure. The radiographer was asked to select the tube potential (for the older unit) and the programme (for the newer unit) which would normally be used for a patient whose anteroposterior (AP) thickness was the same as that of the phantom. The DAP readings from the two Rando phantom projections were averaged. By combining the experimental measurements with the median data obtained from patients a prediction is made of the breakdown of the total DAP. If the model is good it would be expected that the total DAP would be similar to the median DAP from the patient studies. Estimation of effective dose Conversion factors to estimate effective dose from DAP for particular radiographic projections have been provided [11,12], and the factors are dependent on tube potential and on the total filtration of the X-ray beam. Femoral angiograms comprise views of the abdomen, pelvis and lower limbs. For posteroanterior (PA) projections in the abdomen and pelvis the factors vary between 0.12 and 0.22 mSv (Gycm 2 )" 1 at 80 kV, and 0.10 and 0.20 mSv (Gy cm 2 )" 1 at 72 kV. The higher value is for the PA pelvis projection in which the testes are in the field of view. For the lower part of the leg it has been estimated that the conversion factor is approximately 0.01 mSv (Gycm 2 )" 1 . This is based on the estimated dose to bone, muscle and skin in the field of view, the scatter dose to the testes, and the proportion of the total mass of these tissues within the field of view. The conversion factor will be greater in the upper leg due to increased scatter to the testes up to a maximum of 0.03 mSv (Gy cm 2 )" 1 . For leg views which include the testes the conversion factor will be comparable with that for the pelvis.
The British Journal of Radiology, December 1996
Patient doses from femoral angiography
Conversion factors of 0.17 and 0.15 mSv (Gy cm2) 1 are applied to the DAP values from the abdomen/pelvis for the old and new X-ray units respectively. A factor of 0.01 mSv (Gy cm 2 )" 1 is applied to the lower limb DAP values. Results Patient measurements Table I shows the patient DAP readings for each unit, along with the total number of spot film exposures taken. For the standard protocol the screening time is comparable for each X-ray unit. However, for the new unit the DAP is higher by a factor of 3.0, and the total number of images taken is higher by a factor of 1.5. The median DAP is higher by a factor of 2.1 compared with the older. Table II shows details of typical studies for the two X-ray rooms. The numbers of cut film and digital exposures were adjusted in order that the total number equalled the median values in Table I. It can be seen that the tube potential for the cut film and for the digital exposures in the old room are typically 5-10 kV higher than for the digital exposures in the new room. The numbers of spot film exposures were partitioned between the two regions of the model. Aorta and pelvic exposures were allocated to the torso. The remaining leg exposures (upper leg, mid leg, lower leg and foot) were Table I. Measurements made on patients; median (minimum, maximum)
Number of patients Screening time (min) DAP (Gy cm2) No. digital exposures No. film exposures
Old unit
New unit
100 1.7 (0.4, 6.7) 24.4 (5.6, 99.8) 48(14, 182) 16 (0, 24)
100 2.3 (0.9, 13.7) 74(19.8, 184.0) 100 (36, 213)
Table II. Typical femoral angiography studies for the two rooms Projection
Cut-film
Screening
Digital
No.
kV
No.
kV
kV
Old room Aorta/pelvis Upper leg Mid leg Lower leg
3 3 4 6
80 80 80 80
16 16 — 16
80 70 — 65
80 80 80 80
New room Aorta AP Pelvis AP Pelvis oblique Upper leg Mid leg Lower leg Foot
— — — — — — —
— — — — — — —
18 18 18 11 11 11 11
70 72 75 60 60 55 50
71 73 75 65 62 58 55
Vol. 69, No. 828
allocated to the lower limb (Table III). Screening time was equally divided between the two regions. Phantom DAP measurements Table IV shows the phantom data which were measured, and used as input to the model. The DAP values are much higher for the pelvic projection compared with the lower limb projection, as would be expected from the greater tissue thickness and amount of bone in the pelvis. These data were combined with the patient data from Table III. Model predictions of DAP partition Tables V and VI show the predicted DAP breakdown into screening and spot film exposure, and abdomen/ pelvis and lower limb exposure. Total DAP is within 7% of the median DAP of the patient studies of Table I, Table III. Patient data input to model Old unit
New unit
Torso Fluoroscopy time (min) No. digital exposures No. film exposures
0.85 16 3
1.15 54 —
Lower limb Fluoroscopic time (min) No. digital exposure No. film exposures
0.85 32 13
1.15 46 —
Numbers of exposures are derived from the typical studies of Table II. The median screening times of Table I have been equally divided between the torso and the lower limb. Table IV. DAP measurements made on the abdominal phantom (this is the phantom data used as input to the model)
Torso Fluoroscopic DAP (Gy cm2 min" 1 ) Digital DAP (Gy cm2 exposure" 1 ) Film screen DAP (Gy cm2 exposure" 1 )
Old unit
New unit
3.51 0.36 2.27
6.02 0.96 —
0.14 0.09 0.57
1.45 0.39
Lower
Fluoroscopic DAP (Gy cm2 min 1) Digital DAP (Gy cm2 exposure" 1 ) Film screen DAP (Gy cm2 exposure ~1)
Table V. Predicted DAP breakdown for fluoroscopy and spot film exposure Old unit (Gy cm2)
New unit (Gy cm2)
Fluoroscopy Digital Film
3.1 (12%) 8.6 (33%) 14.2 (55%)
8.6(11%) 68.0 (89%)
Total
26.0(100%)
78.4 (100%)
1161
P R Hoskins, I Gillespie and H M Ireland Table VI. Predicted DAP breakdown for lower limb and abdomen/pelvis exposure
In a study investigating organ doses, Castellano et al [7] undertook detailed collection of the machine settings (beam area A, tube potential, mAs and number of expoOld unit New unit sures n or mA and screening time) during each procedure (Gy cm2) (Gy cm2) for 17 patients. Total DAP was partitioned between individual regions by assuming that the DAP was proAbdo/pelvis 15.6 (60%) 58.8 (75%) portional to (mAs kV2 A n). Thwaites et al [6] divided Lower limb 10.4 (40%) 19.6 (25%) the femoral angiogram examination into five regions 26.0(100%) 78.4(100%) Total from the abdomen to the feet in their study of 25 patients. They performed detailed collection of machine settings and DAP values during each procedure. In the present indicating that the model is a reasonable one. The study the lower limbs contributed an estimated 40% and fluoroscopic contribution is small (12% for the old unit 25% of the total DAP for the old and new X-ray units, and 11% for the new unit). For the old unit the film- respectively. These values are different from the 50% screen DAP contribution (55%) is more than that for used by Steele and Temperton [1] and highlight the the digital spot film contribution (33%), even though necessity of providing some estimate of the lower limb the number of exposures is much less. For the new X-ray contribution. Use of the simple model described in this room two-thirds of the DAP is associated with digital paper enables an estimate to be made of the lower limb exposures of the pelvis and abdomen. DAP contribution without the need for detailed measurements for each patient, which would be enorEstimation of effective dose Tables VII and VIII show the estimated effective dose mously time-consuming for the investigator. The second step which must be taken for estimation breakdown by screening and spot film exposure, and of effective dose from femoral angiography involves abdomen/pelvis and lower limb exposure. Effective dose is a factor of 3.3 higher for the new choice of conversion factors (to convert from DAP to room. Fluoroscopy contributes only 12-18% of the effective dose equivalent or effective dose) for the different effective dose, and lower limb exposure contributes only anatomical regions. Published studies use differing degrees of complexity. Steele and Temperton [ 1 ] effec2-4% of the total. Conversion factors for the abdomen and pelvis differ tively assume that the conversion factor is zero for the by + 30% from the average value and this will determine lower limb, and has a single value for the abdomen and pelvis. Thwaites et al [6] use different conversion factors the limits of accuracy of the effective dose values. for each of the five regions they considered. Castellano et al [7] derived correction factors for several different Discussion The first step which must be undertaken for estimation organs. This study involves selection of conversion facof the effective dose from femoral angiography is par- tors for two regions; the abdomen/pelvis and the lower tition of the DAP between different regions of the body. limb, which is a level of complexity between the first and Steele and Temperton [1] in a study of 31 patients the third of the studies above. The final step consists of summating the effective dose divided the total patient DAP by two, on the assumption that 50% of the DAP arose from lower limb exposure. (or effective dose equivalent) from different regions or organs. In the study [ 1 ] the mean effective dose equivalTable VII. Estimated effective dose breakdown for fluoroscopy ent was 4 mSv. The ratio of effective dose equivalent to effective dose for the abdomen and pelvis is 1.35 [12], and spot film exposure so that an effective dose equivalent of 4 mSv gives an Old unit New unit effective dose of 3 mSv. In studies [6] and [7] the mean (mSv) (mSv) effective dose was 4 mSv and 3.1 mSv, respectively. The effective dose of 2.7 mSv for the older X-ray unit of Fluoroscopy 0.51 (18%) 1.06(12%) this study is in reasonable agreement with these values; Digital 1.01 (37%) 7.95 (88%) however, the value of 9.0 mSv for the new X-ray unit Film 1.23 (45%) — is higher. Total 2.7(100%) 9.0(100%) In the study of Castellano et al [7] it is noted that the largest contribution to patient dose came from fluoroscopy during manipulation of the catheter; 62% of Table VIII. Predicted effective dose breakdown for lower limb DAP. This was not found in the current study. and abdomen/pelvis exposure Fluoroscopy, which would include catheter manipulation, accounted for 12% or less of the total DAP, and Old unit New unit less than 18% of the effective dose. (mSv) (mSv) The effective dose in the new vascular room is a factor of 3.3 higher than the old room. Factors which contribute Abdo/pelvis 2.65 (96%) 8.81 (98%) Lower limb 0.10 (4%) 0.20 (2%) to this are listed below. Number of exposures. For the new X-ray unit there is Total 2.7(100%) 9.0(100%) a clear tendency for more spot film exposures to be taken 1162
The British Journal of Radiology, December 1996
Patient doses from femoral angiography
(median value is 100 compared with 64 in the old room). There is increased use of oblique projections in the abdominal and pelvic regions, in addition to standard AP projections, and these will contribute significantly to patient dose. There is also increased use of foot and ankle projections. This is due to current surgical practice, whereby there is increased performance of arterial by-pass surgery to distal vessels. This particular change in procedure is unlikely to contribute significantly to increased patient dose due to the small amount of tissue in the field of view, and also due to the low weighting factor. The method for starting spot film exposures is also likely to have an impact on the number taken. On the old unit there is a programmed delay, whereas on the new unit the operator starts the exposure manually before the contrast arrives. The improved control which the manual method gives means that in all patients exposure is started before contrast arrives, which is not always the case for the old set, but at the expense of an increase in the number of exposures. Field of view. The useful field of view of the new unit is a factor of three more than that of the old unit. Table IV shows that the increase in DAP per digital exposure for the new set is a factor of 2.7 for the abdomen/pelvis projection and 4.3 for the lower limb projection. The increase in field size for the new X-ray unit is one factor which will contribute to this increase. The large field of view of the new unit means that, for digital exposures, fewer overlapping fields are needed than would be the case for a small field of view. In the old unit, sequential images down the leg were performed using cut film, which is 35 cm square. In practice, therefore, the number of overlapping digital fields for the new unit is comparable with the number of overlapping films for the old unit. Equipment factors. These will have a significant impact on patient dose. For fluoroscopy the relevant factors will be the input intensifier exposure rate and the tube potential and tube current selected by the automatic exposure control (AEC). For digital and film exposures the relevant factors are the tube potential selected by the radiographer, along with the mAs determined by the AEC. The input intensifier exposure rate with 1 mm of copper in the beam was 0.36 u G y s " 1 for the old unit and 0.39 uGy s" 1 for the new unit. Both values are below the action level of 0.44 u G y s " 1 used in our region. Skin dose rate, measured with 18 cm of Perspex attenuator, was 0.38 mGy s" 1 for the old unit and 0.27 mGy s" 1 for the new unit. These values are also below the action levels for the relevant field sizes (0.40 m G y s " 1 for the old set and 0.30 mGy s" 1 for the new set). It has been shown that alteration of machine settings can help reduce patient dose in barium studies and lumbar spine examinations [15, 16]. In this study 89% of the DAP from the new X-ray set arose from digital exposures. Reduction in patient dose from digital exposure could be performed by maximizing the delay before starting exposure, reducing the frame rate during exposure, and increasing the tube potential. Exact choice of tube potential is important as iodine contrast, and hence the ability to image
Vol. 69, No. 828
small vessels, is strongly related to the tube potential [17]. Pre-programmed tube potential values used during digital exposure by the new unit are less than those selected manually by the radiographer on the old unit by 5-10 kV. It is possible that there is scope for reduction in patient dose by increase in the pre-programmed tube potential values, but the impact on image quality would need to be critically assessed. Conclusion
A simple model has been described whereby estimates may be made of the contribution of screening and of abdominal/pelvic exposure to the total DAP during femoral angiography studies. The agreement to within 7% between the median patient DAP and the predicted total DAP indicates that the model is reasonable. Most of the radiation dose during femoral angiogram procedures arises from spot film exposure, and therefore attempts to decrease spot film exposure will have a significant impact on patient dose. Acknowledgments
Thanks to Joan Ritchie and all vascular room radiographers of the Edinburgh Royal Infirmary for assistance with this study. Thanks to Mr J R Williams for helpful comments concerning the manuscript. References 1. STEELE, H R and TEMPERTON, D H, Technical note: Patient doses received during digital subtraction angiography, Br. J. Radioi, 66,452-456 (1993). 2. MARSHALL, N W, NOBLE, J and FAULKNER, K, Patient and staff dosimetry in neuroradiological procedures, Br. J. Radioi, 68, 495-501 (1995). 3. VANO, E, GONZALEZ, L, FERNANDEZ, J M ET AL, Patient dose values in interventional radiology, Br. J. Radioi., 68, 1215-1220(1995). 4. FAULKNER, K, LOVE, H G, SWEENEY, J K ET AL, Radiation doses and somatic risk to patients during cardiac radiological procedures, Br. J. Radioi., 59, 359-363 (1986). 5. LEUNG, K C and MARTIN, C J, Effective doses for coronary angiography, Br. J. Radioi., 69, 426-431 (1996). 6. THWAITES, J H, RAFFERTY, M W, GRAY, N ET AL, A patient dose survey for femoral arteriogram diagnostic radiographic examinations using a dose-area product meter, Phys. Med. Bioi, 41, 899-907 (1996). 7. CASTELLANO, I A, McNEIL, J G, THORP, N C ET AL, Assessment of organ radiation doses and associated risk for digital bifemoral angiography, Br. J. Radioi., 68, 502-507 (1995). 8. INTERNATIONAL COMMISSION ON RADIOLOGICAL PROTECTION, 1977 Recommendations of the International Commission on Radiological Protection, ICRP Publication 26 (Pergammon Press, Oxford), Ann. ICRP, Vol. 1, No. 3(1977). 9. ICRP, 1990 Recommendations of the International Commission on Radiological Protection, Publication 60 (Pergammon Press, Oxford), Ann. ICRP, Vol. 21, No. 1-3 (1990). 10. SHRIMPTON, P C, WALL, B F, JONES, D G ET AL, A National Survey of Doses to Patients Undergoing a Selection of Routine X-ray Examinations in English Hospitals (NRPBR200) (HMSO, London) (1986).
1163
P R Hoskins, I Gillespie and H M Ireland 11. HART, D, JONES, D G and WALL, B F, Estimation of Effective Dose in Diagnostic Radiology From Entrance Surface Dose and Dose-area Product Measurements (NRPBR262) (HMSO, London) (1994). 12. HART, D, JONES, D G and WALL, B F, Normalised Organ Doses for Medical X-ray Examinations Calculated Using Monte Carlo Techniques (NRPB-SR262) (NRPB, Didcot) (1994). 13. HOSKINS, P R and ADAM, R D, An evaluation of a Diamentor based system for estimation of spot film dose-area product values, Br. J. Radioi., 68, 1106-1111 (1995). 14. NATIONAL RADIOLOGICAL PROTECTION BOARD, National Protocol For Patient Dose Measurements in Diagnostic Radiology (NRPB, Didcot) (1992).
1164
15. MARTIN, C J and HUNTER, S, Reduction of patient doses from barium meal and barium enema examinations through changes in equipment factors, Br. J. Radioi, 67, 1196-1205 (1994). 16. MARTIN, C J and HUNTER, S, Analysis of patient doses for myelogram and discogram examinations and their reduction through changes in equipment set-up, Br. J. Radioi., 68, 508-514 (1995). 17. TOSI, G, TORRESIN, A, RAIMONDI, G ET AL, Contrast media in neuroangiographical examinations: optimisation of X-ray beam quality and screen-film combination. In Optimisation of Image Quality and Patient Exposure in Diagnostic Radiology, ed. by B M Moores, B F Wall, H Eriksat ET AL (British Institute of Radiology, London), pp. 150-152 (1989).
The British Journal of Radiology, December 1996
