Quantitative performance evaluation of 124I PET/MRI lesion dosimetry in differentiated thyroid cancer.

The aim was to investigate the quantitative performance of 124I PET/MRI for pre-therapy lesion dosimetry in differentiated thyroid cancer (DTC). Phantom measurements were performed on a PET/MRI system (Biograph mMR, Siemens Healthcare) using 124I and 18F. The PET calibration factor and the influence of radiofrequency coil attenuation were determined using a cylindrical phantom homogeneously filled with radioactivity. The calibration factor was 1.00 ± 0.02 for 18F and 0.88 ± 0.02 for 124I. Near the radiofrequency surface coil an underestimation of less than 5% in radioactivity concentration was observed. Soft-tissue sphere recovery coefficients were determined using the NEMA IEC body phantom. Recovery coefficients were systematically higher for 18F than for 124I. In addition, the six spheres of the phantom were segmented using a PET-based iterative segmentation algorithm. For all 124I measurements, the deviations in segmented lesion volume and mean radioactivity concentration relative to the actual values were smaller than 15% and 25%, respectively. The effect of MR-based attenuation correction (three- and four-segment µ-maps) on bone lesion quantification was assessed using radioactive spheres filled with a K2HPO4 solution mimicking bone lesions. The four-segment µ-map resulted in an underestimation of the imaged radioactivity concentration of up to 15%, whereas the three-segment µ-map resulted in an overestimation of up to 10%. For twenty lesions identified in six patients, a comparison of 124I PET/MRI to PET/CT was performed with respect to segmented lesion volume and radioactivity concentration. The interclass correlation coefficients showed excellent agreement in segmented lesion volume and radioactivity concentration (0.999 and 0.95, respectively). In conclusion, it is feasible that accurate quantitative 124I PET/MRI could be used to perform radioiodine pre-therapy lesion dosimetry in DTC.

Hard beta and gamma emissions of 124I. Impact on occupational dose in PET/CT.

AIM: The hard beta and gamma radiation of 124I can cause high doses to PET/CT workers. In this study we tried to quantify this occupational exposure and to optimize radioprotection. METHODS: Thin MCP-Ns thermoluminescent dosimeters suitable for measuring beta and gamma radiation were used for extremity dosimetry, active personal dosimeters for whole-body dosimetry. Extremity doses were determined during dispensing of 124I and oral administration of the activity to the patient, the body dose during all phases of the PET/CT procedure. In addition, dose rates of vials and syringes as used in clinical practice were measured. The procedure for dispensing 124I was optimized using newly developed shielding. RESULTS: Skin dose rates up to 100 mSv/min were measured when in contact with the manufacturer's vial containing 370 MBq of 124I. For an unshielded 5 ml syringe the positron skin dose was about seven times the gamma dose. Before optimization of the preparation of 124I, using an already reasonably safe technique, the highest mean skin dose caused by handling 370 MBq was 1.9 mSv (max. 4.4 mSv). After optimization the skin dose was below 0.2 mSv. CONCLUSION: The highly energetic positrons emitted by 124I can cause high skin doses if radioprotection is poor. Under optimized conditions occupational doses are acceptable. Education of workers is of paramount importance.

Blocking the saphenofemoral junction during ultrasound-guided foam sclerotherapy-- assessment of a presumed safety-measure procedure.

OBJECTIVES: Ultrasound-guided foam sclerotherapy (UGFS) is a technique in which a mixture of sclerosing drug and gas is used to treat varicose veins. Several authors have demonstrated transient systemic effects after UGFS. These effects are not well understood but probably originate from a systemic distribution of the sclerosing foam. Therefore, safety measures have been developed to prevent foam from flowing into the deep venous system. The aim of the study is to evaluate whether blockage of the saphenofemoral (SF) junction by either manual compression or surgical ligation prevents microbubbles from leaking into the deep venous circulation. METHODS: To detect the distribution of microbubbles, radioactive pertechnetate (99mTcO4-) was added to the foam solution. Initially, in vitro trials were performed in the laboratory to investigate the effect of 99mTc on foam stability. The time taken for foam to liquefy was measured for foam alone and for the mixture with 99mTc. In subsequent research, eight varicose great saphenous veins (GSVs) were treated by UGFS. In three patients, this treatment was preceded by surgical ligation of the SF junction. In three patients, the groin was manually compressed during UGFS. In two patients, UGFS was performed without compression of the groin. RESULTS: In vitro, 99mTc did not influence foam stability; after 2.6 min all foam had reduced to liquid, regardless of whether 99mTc had been added or not. In vivo trials showed that all patients showed a decrease in the cumulative amount of 99mTc detected in the GSV following polidocanol-99mTc mixture injection. However, the decrease of radioactivity was slightly reduced when compression or ligation of the SF junction was performed. CONCLUSIONS: Blocking the SF junction during UGFS using either manual compression or ligation does not prevent, but may reduce the flow of foam into the femoral vein.

The effect of equipment set up on patient radiation dose in conventional and CT angiography of the renal arteries.

Patient radiation dose in angiography of the renal arteries was assessed and optimized after installing new radiological equipment. In three separate studies (n=50, 25 and 20) patient exposure was monitored in detail. For the first study default factory settings were used, for the second the number of digital subtraction angiography (DSA) images was halved and the X-ray beam filtering during fluoroscopy was increased, and for the third study filtering during DSA was increased as well. Standard projections were derived and used in Monte Carlo simulations to derive dose conversion coefficients to calculate effective dose from the dose-area product (DAP). Dose conversion coefficients were also calculated for CT angiography (CTA). Using default factory settings on the new angiography system, DAP, number of images and effective dose were much higher than on the replaced unit. For the studies given above, DAP was reduced from 144 Gy cm(2) to 65 Gy cm(2) to 32 Gy cm(2), and effective dose from 22 mSv to 11 mSv to 9.1 mSv, respectively. Effective dose due to CTA was 5.2 mSv. It is concluded that modern angiography systems, resulting in high customer satisfaction, may readily cause much higher patient exposure than older systems. These doses may also be much higher than necessary. Optimization before putting such systems into use is absolutely essential. Internationally accepted recommendations for image quality and technique factors in angiography would be of great help.

Comparison between human pharmacokinetics and imaging properties of two conjugation methods for 99mTc-annexin A5.

Annexin A5 (AnxA5) is a protein with high affinity for phosphatidyl serine, a phospholipid exposed on the cell surface during apoptosis. This phenomenon has been used for determination of cell death after myocardial infarction. To evaluate the potential of (99m)Tc-AnxA5 for in vivo scintigraphy of apoptotic cells, the pharmacokinetics and imaging properties of two radiopharmaceuticals, (99m)Tc-(n-1-imino-4-mercaptobutyl)-AnxA5 (I-AnxA5) and (99m)Tc-(4,5-bis(thioacetamido)pentanoyl)-AnxA5 (B-AnxA5), were studied. I-AnxA5 was administered intravenously to seven patients and one healthy volunteer, and B-AnxA5 was administered to 12 patients. All patients in the pharmacokinetic study had myocardial disease. Additionally, imaging was performed in a patient with acute myocardial infarction, as well as in three patients with different malignancies. The plasma concentration, excretion and biodistribution of (99m)Tc-AnxA5 were measured, as well as levels of AnxA5 antigen. The kinetic data of both radiopharmaceuticals in plasma fitted a two-compartment model. Both preparations had similar half-lives, but a different distribution over the two compartments. Plasma levels of AnxA5 antigen showed a broad variation. Both radiopharmaceuticals accumulated in the kidney, liver and gut. B-AnxA5 was excreted significantly faster than I-AnxA5. Both compounds can be used for imaging of the head/neck region, the thorax and the extremities. B-AnxA5 has a faster clearance and a lower radiation dose. Imaging of apoptosis in the abdomen will be difficult with both radiopharmaceuticals, and especially with B-AnxA5 because of its faster appearance in the gut.

Patient and occupational dose in neurointerventional procedures.

Neurointerventional procedures can involve very high doses of radiation to the patient. Our purpose was to quantify the exposure of patients and workers during such procedures, and to use the data for optimisation. We monitored the coiling of 27 aneurysms, and embolisation of four arteriovenous malformations. We measured entrance doses at the skull of the patient using thermoluminescent dosemeters. An observer logged the dose-area product (DAP), fluoroscopy time and characteristics of the digital angiographic and fluoroscopic projections. We also measured entrance doses to the workers at the glabella, neck, arms, hands and legs. The highest patient entrance dose was 2.3 Gy, the average maximum entrance dose 0.9+/-0.5 Gy. The effective dose to the patient was estimated as 14.0+/-8.1 mSv. Other average values were: DAP 228+/-131 Gy cm(2), fluoroscopy time 34.8+/-12.6 min, number of angiographic series 19.3+/-9.4 and number of frames 267+/-143. The highest operator entrance dose was observed on the left leg (235+/-174 microGy). The effective dose to the operator, wearing a 0.35 mm lead equivalent apron, was 6.7+/-4.6 microSv. Thus, even the highest patient entrance dose was in the lower part of the range in which nonstochastic effects might arise. Nevertheless, we are trying to reduce patient exposure by optimising machine settings and clinical protocols, and by informing the operator when the total DAP reaches a defined threshold. The contribution of neurointerventional procedures to occupational dose was very small.

Biodistribution and dosimetry of 99mTc-BTAP-annexin-V in humans.

The purpose of this study was to determine the biodistribution and the associated radiation dose of technetium-99m 4,5-bis(thioacetamido)pentanoyl-annexin-V (99mTc-Apomate), a tracer proposed for the study of apoptosis. Eight patients (including two females) with normal kidney and liver functions were included in the study. An activity of 580 +/- 90 MBq of 99mTc-Apomate was injected intravenously, immediately followed by a dynamic study of 30 frames of 1 min each. At about 1 h, 4 h and 20 h p.i., whole-body scans were acquired. All activity distributions were measured using a dual-head gamma camera. Before injection of activity, a transmission scan with a cobalt-57 flood source had been performed to determine patient attenuation. Blood samples were taken every 10 min during the first hour after injection, and at about 4 and 20 h. Urine and faeces were collected during the first 20 h. Organ uptake was estimated after correction for body background activity, attenuation and scatter. Residence times were calculated from the dynamic and whole-body studies and used as input in the Mirdose 3.1 program to obtain organ doses and effective dose. It was found that radioactivity strongly accumulated in the kidneys and the liver [at 70 min p.i., 28% +/- 8% and 20% +/- 4% of the injected dose (ID), respectively]. Uptake in the target tissues (lymphomas or heart) was negligible from a dosimetric point of view. Extrapolating data from the first 20 h, one finds that approximately 73% of the ID will be excreted in the urine, and 27% in the faeces. The biological half-life of the activity in the total body was 16 +/- 7 h. Some organ doses +/- standard deviation (SD) in microGy/MBq were: kidneys 63 +/- 22, urinary bladder 20 +/- 6, spleen 15 +/- 3, liver 13 +/- 3, upper large intestine 12 +/- 6, lower large intestine 8 +/- 4, testes 6 +/- 2 and red bone marrow 4 +/- 0.7. The effective dose was 7.6 +/- 0.5 microSv/MBq, corresponding to a total effective dose of 4.6 +/- 0.3 mSv for a nominal injected activity of 600 MBq. In conclusion, 99mTc-Apomate has a high uptake in the kidneys and liver--in fact a factor of 1.3-1.6 higher than that found for the previously studied 99mTc-(n-1-imino-4-mercaptobutyl)-annexin-V. The biological half-life is shorter, however, but still long compared with the physical half-life of 99mTc. The faster appearance of activity in the intestines may preclude imaging of apoptosis in the abdomen. The effective dose is within the lower range of values reported for typical 99mTc compounds.

Patient and occupational dosimetry in double contrast barium enema examinations.

A new and relatively simple method is presented to distribute total dose-area product (DAP) over a number of projections that model exposure during double contrast barium enema (DCBE) examinations. In addition, hitherto unavailable entrance and effective doses to the physician performing the DCBE examination have been determined. DAP, fluoroscopy time, number of images as well as some patient data were collected for 150 DCBE examinations. For a subset of 50 examinations, the distribution of DAP over 12 hypothetical but representative projections was estimated by measuring the entrance dose in the centre of each of these projections during the complete procedure. Effective dose to the patient was obtained using DAP to effective dose conversion coefficients calculated for each of the 12 projections. Exposure of the worker was quantified by measuring the entrance dose at the forehead, neck, arms, right hand and legs. The sex-averaged effective dose to the patient per examination was 6.4+/-2.1 mSv (mean+/-SD; n=50) and the corresponding DAP was 44+/-22 Gy cm(2). The effective dose to the worker per examination was 0.52 microGy (n=50), whereas the highest entrance dose of 30+/-25 microGy was found for the right arm. The proposed method for deriving the distribution of total DAP over a set of representative projections is much less time consuming than visual observation of patient exposure, whilst accuracy seems acceptable. Entrance and effective doses per examination for workers in DCBE examinations are very low. For a normal workload, doses remain far below the legally established dose limits.

Patient dosimetry of intravenously administered 99mTc-annexin V.

UNLABELLED: Annexin V labeled with 99mTc is evaluated as a potential in vivo marker for tissue with increased apoptosis. Promising results in patients have been obtained with 99mTc-(n-1-imino-4-mercaptobutyl)-annexin V (99mTc-i-AnxV). Because information on biodistribution and radiation burden is desired for the application of any radiopharmaceutical, a dosimetric study of 99mTc-i-AnxV was undertaken. METHODS: Eight persons with normal kidney and liver functions were included in this study: six patients with myocardial infarction, one with Crohn's disease, and one healthy volunteer. Approximately 600 MBq 99mTc-i-AnxV were injected intravenously immediately before a dynamic study with a dual-head gamma camera in conjugate view mode. In the next 24 h, two to four whole-body scans were acquired. Patient thickness was determined from a transmission scan with a 57Co flood source. Organ uptake was estimated after correction for background, attenuation, and scatter, using a depth-independent buildup factor and an organ-size-dependent attenuation correction. Residence times were calculated from the dynamic and whole-body studies and used as input for the MIRDOSE 3.1 program to obtain organ-absorbed doses and effective dose. RESULTS: Activity strongly accumulated in the kidneys (21% +/- 6% of the injected dose at 4 h postinjection) and the liver (12.8% +/- 2.2%). Uptake in the target tissues (myocardium or colon) was limited and negligible from a dosimetric point of view. The biologic half-life of activity registered over the total body was 62 +/- 13 h. Of the excreted activity, approximately 75% went to the urine and 25% to the feces. The absorbed dose for the more strongly exposed organs was (in microGy/MBq): kidneys, 93 +/- 24; spleen, 22 +/- 6; liver, 17 +/- 2; testes, 15 +/- 3; thyroid, 10 +/- 6; urinary bladder wall, 7.5 +/- 2.6; and red bone marrow, 5.5 +/- 0.8. The effective dose was 9.7 +/- 1.0 microSv/MBq, corresponding to a total effective dose of 5.8 +/- 0.6 mSv for a nominally injected activity of 600 MBq. CONCLUSION: 99mTc-i-AnxV strongly accumulates in the kidneys and to a lesser degree in the liver. The associated effective dose per MBq is in the midrange of values found for routine 99mTc-labeled compounds. From a dosimetric point of view 99mTc-i-AnxV is therefore well suited for the study of apoptosis in patients.

Motion of the proximal renal artery during the cardiac cycle.

In 48 hypertensive patients, the motion of the proximal renal artery during the cardiac cycle was quantified using two-dimensional quantitative flow (QF) measurements and automatic contour detection. Substantial translational motion was observed with an amplitude ranging from 1 to 4 mm. Since motion effectively reduces spatial resolution, the use of motion suppression techniques should be strongly considered for renal MR angiography. J. Magn. Reson. Imaging 2000;12:924-928.

Visualisation of cell death in vivo in patients with acute myocardial infarction.

BACKGROUND: In-vivo visualisation and quantification of the extent and time-frame of cell death after acute myocardial infarction would be of great interest. We studied in-vivo cell death in the hearts of patients with an acute myocardial infarction using imaging with technetium-99m-labelled annexin-V-a protein that binds to cells undergoing apoptosis. METHODS: Seven patients with an acute myocardial infarction and one control were studied. All patients were treated by percutaneous transluminal coronary angioplasty (six primary and one rescue), resulting in thrombolysis in myocardial infarction (TIMI) III flow of the infarct-related artery. 2 h after reperfusion, 1 mg annexin-V labelled with 584 MBq Tc-99m was injected intravenously. Early (mean 3.4 h) and late (mean 20.5 h) single-photon-emission computed tomographic (SPECT) images of the heart were obtained. Routine myocardial resting-perfusion imaging was also done to verify infarct localisation. FINDINGS: In six of the seven patients, increased uptake of Tc-99m-labelled annexin-V was seen in the infarct area of the heart on early and late SPECT images. No increased uptake was seen in the heart outside the infarct area. All patients with increased Tc-99m-labelled annexin-V uptake in the infarct area showed a matching perfusion defect. In a control individual, no increased uptake in the heart was seen. INTERPRETATION: Increased uptake of Tc-99m-labelled annexin-V is present in the infarct area of patients with an acute myocardial infarction, suggesting that programmed cell death occurs in that area. The annexin-V imaging protocol might allow us to study the dynamics of reperfusion-induced cell death in the area at risk and may help to assess interventions that inhibit cell death in patients with an acute myocardial infarction.

Plasma FFA utilization and fatty acid-binding protein content are diminished in type 2 diabetic muscle.

In this study, we investigated the hypothesis that impairments in forearm skeletal muscle free fatty acid (FFA) metabolism are present in patients with type 2 diabetes both in the overnight fasted state and during beta-adrenergic stimulation. Eight obese subjects with type 2 diabetes and eight nonobese controls (Con) were studied using the forearm balance technique and indirect calorimetry during infusion of the stable isotope tracer [U-(13)C]palmitate after an overnight fast and during infusion of the nonselective beta-agonist isoprenaline (Iso, 20 ng. kg lean body mass(-1) x min(-1)). Additionally, activities of mitochondrial enzymes and of cytoplasmatic fatty acid-binding protein (FABP) were determined in biopsies from the vastus lateralis muscle. Both during fasting and Iso infusion, the tracer balance data showed that forearm muscle FFA uptake (Con vs. type 2: fast 449+/-69 vs. 258 +/-42 and Iso 715+/-129 vs. 398+/-70 nmol. 100 ml tissue(-1) x min(-1), P<0.05) and FFA release were lower in type 2 diabetes compared with Con. Also, the oxidation of plasma FFA by skeletal muscle was blunted during Iso infusion in type 2 diabetes (Con vs. type 2: Iso 446 +/- 274 vs. 16+/-70 nmol. 100 ml tissue(-1) x min(-1), P<0.05). The net forearm glycerol release was increased in type 2 diabetic subjects (P< 0.05), which points to an increased forearm lipolysis. Additionally, skeletal muscle cytoplasmatic FABP content and the activity of muscle oxidative enzymes were lowered in type 2 diabetes. We conclude that the uptake and oxidation of plasma FFA are impaired in the forearm muscles of type 2 diabetic subjects in the overnight fasted state with and without Iso stimulation.

Value of radiological diagnosis of skull fracture in the management of mild head injury: meta-analysis.

OBJECTIVES: Head injury is a common event. Most patients sustain a mild head injury (MHI), and management depends on the risk of an intracranial haemorrhage (ICH). The value of a plain skull radiograph as a screening tool for ICH is controversial. The aim of this meta-analysis was to estimate and explain differences in reported sensitivity and specificity of the finding of a skull fracture for the diagnosis of ICH, in order to assess the value of the plain skull radiograph in the investigation of patients with MHI, and to estimate the prevalence of ICH in these patients. METHOD: After a systematic literature search 20 studies were selected that reported data on the prevalence of ICH after MHI and/or data on the diagnostic value of skull fracture for the diagnosis of ICH. The mean prevalence of ICH weighted for the sample size was determined. The sensitivity and specificity of different studies were combined using a summary receiver operator characteristic curve. Correlation analysis was used to determine factors that could explain the reported differences between studies. RESULTS: The weighted mean prevalence of ICH after MHI is 0.083. The potential for verification bias and the percentage of patients who had suffered loss of consciousness or post-traumatic amnesia were the most significant factors explaining interstudy differences in sensitivity and specificity. Based on studies wherein at least 50% of patients had a CT study of the brain, the estimated sensitivity of a radiographic finding of skull fracture for the diagnosis of ICH is 0.38 with a corresponding specificity of 0.95. CONCLUSION: The plain skull radiograph is of little value in the initial assessment of MHI patients.

Quantitative analysis of magnetization transfer images of the brain: effect of closed head injury, age and sex on white matter.

Magnetization transfer (MT) imaging has an application in quantitative assessment of cerebral white matter. Previously published postprocessing methods have inherent problems, and therefore a new analysis technique is presented. This technique was found to be more sensitive for white matter changes in patients with a postconcussional syndrome, compared to other methods previously described. Because of the potential application of this technique in longitudinal and group studies, age and sex dependence of the MT ratio (MTR) of white matter were studied. In a group of 51 healthy subjects, a decrease in the mean MTR as well as an increasing distribution width of the MTR was found with increasing age. The mean MTR in males was higher than in females. These results stress the need to take age and sex into account when interpreting MTR data. Magn Reson Med 42:803-806, 1999.

Microdialysis assessment of local adipose tissue lipolysis during beta-adrenergic stimulation in upper-body-obese subjects with type II diabetes.

The present study was designed to investigate indicators of abdominal adipose tissue lipolysis (microdialysis), and subcutaneous adipose tissue blood flow and whole-body lipolysis, in obesity-associated type II diabetes during overnight-fasted conditions (baseline) and during intravenous infusion of the non-selective beta-agonist isoprenaline. Basal subcutaneous adipose tissue blood flow and isoprenaline-induced increases in adipose tissue blood flow were not significantly different between subjects with type II diabetes and non-obese, non-diabetic controls. Adipose tissue interstitial glycerol concentrations were significantly higher in subjects with type II diabetes compared with controls (P<0. 01), and during isoprenaline infusion there was a decrease in interstitial glycerol in both groups (P<0.001). Arterial glycerol concentrations were higher in subjects with type II diabetes compared with controls (P<0.05), whereas the increases in arterial glycerol concentration in response to isoprenaline infusion were of a similar magnitude in the two groups. Estimated subcutaneous adipose tissue glycerol release was not significantly different between the groups (controls and subjects with type II diabetes: baseline, -129+/-32 and -97+/-72 micromol.min(-1).100 g(-1) adipose tissue respectively; isoprenaline, -231+/-76 and -286+/-98 micromol. min(-1).100 g(-1) respectively). Values for fat oxidation were not significantly different between groups, whereas the isoprenaline-induced increase in fat oxidation tended to be less pronounced in subjects with type II diabetes compared with controls (0.022+/-0.008 and 0.038+/-0.003 g/min respectively; P=0.058). Thus estimated basal subcutaneous adipose tissue glycerol release, expressed per unit of fat mass, is not different in controls and in subjects with type II diabetes. Additionally, the isoprenaline-induced increases in indicators of local abdominal subcutaneous adipose tissue, systemic lipolysis and abdominal adipose tissue blood flow responses were comparable in obese subjects with type II diabetes and in controls. The last two findings contrast with previous data from obese subjects, indicating that the regulation of lipolysis may differ in obesity and obesity-associated type II diabetes.

Patient dosimetry in abdominal arteriography.

This study aims at accurate quantification of x-ray exposure and effective dose to the patient in abdominal arteriography. Using an automatic monitoring system, all relevant exposure parameters were determined during 172 abdominal arteriographies. Common projections were extracted for a 'normal' reference group of procedures and used in Monte Carlo calculations of dose-area product to organ dose conversion coefficients. Dose-area product, organ doses and effective dose were quantified for intravenous and intra-arterial procedures. The large data sets describing exposure could be condensed to a set of 28 common views. New coefficients to convert dose area product to organ equivalent dose and effective dose were calculated for nine views contributing approximately 80% to the total dose-area product. The average dose-area product was 32 Gy cm2 in intravenous procedures and 47 Gy cm2 in intra-arterial procedures. The corresponding average effective doses to the patient were 4 mSv and 6 mSv respectively (range 2-12 mSv, actual value depending on procedure type and gender). It is concluded that automatic monitoring of x ray exposure parameters, complemented by the calculation of Monte Carlo organ dose conversion coefficients, is a feasible and promising approach to accurate dosimetry of complex arteriographic procedures.

Improving contrast and tracking of tags in cardiac magnetic resonance images.

Cardiac tagging permits non-invasive study of myocardial motion with high accuracy. Unfortunately, tagging contrast is impaired at later heart phases due to longitudinal relaxation. Histogram modification is presented as a method for improving contrast in later, faded images of a tagging series by altering these images such that their intensity histograms approximate the shape of the first, unfaded image of the series. This technique greatly improves the contrast and facilitates automatic detection of tags. Furthermore, a method is described for automatically tracking tag positions in short-axis images of the left ventricle modulated with a tagging grid. The method differs from previously reported methods in that, in one single filtering process in the Fourier domain, both the grid crossings as well as the grid centers are detected, and thus increased sampling resolution is obtained. The method was validated by applying a mathematical model of left ventricular motion to tagged images of the thigh muscle. The average discrepancy between theoretically predicted and automatically detected tag positions was 0.04 +/- 0.17 mm (mean +/- SD).

Reproducibility of spirometrically controlled CT lung densitometry in a clinical setting.

The aim of this study was to assess the reproducibility of quantitative, spirometrically gated computed tomographic (CT) lung densitometry at defined levels of inspiration in hospitalized patients. On two consecutive days, spirometrically gated CT sections were obtained from 20 hospitalized patients at 5 cm above and 5 cm below the carina, and at 90 and 10% of the vital capacity (VC). The mean, modal and median lung densities were calculated, the cut-off points of the frequency distribution of Hounsfield units (HU) defining the lowest and the highest 10th percentile, as well as the histogram full width at half maximum. The lung density parameters of corresponding CT sections of both studies were compared. Reproducibility was expressed as the standard deviation of the signed difference between the results of Day 1 and Day 2 divided by the square root of 2. Reproducibility data were correlated with results of airflow limitation. At 90% VC, reproducibility was of the order of 3-14 HU in both lung zones. At 10% VC, reproducibility was worse by approximately a factor of three. No relationship was found between reproducibility and results of airflow limitation. In conclusion, objective measurement of lung density at spirometrically controlled levels of inspiration is a reproducible method in assessing pulmonary density. Reproducibility of lung density measurements is not influenced by severe respiratory insufficiency. The most reproducible computed tomographic lung density measurements can be obtained at 90% vital capacity.

On segmentation of lung parenchyma in quantitative computed tomography of the lung.

Our purpose in this study was to investigate the influence of segmentation threshold and number of erosions on parameters used in quantitative computed tomography (CT) of the lung (erosions are shrink operations on the segmented area). Parameters assessed were mean lung density, area of the segmented lung, two percentiles, and the pixel index, which is the relative area of the histogram below -905 Hounsfield Units (HU). We analyzed images of ten emphysematous and ten nonemphysematous patients, that had been scanned at carina level in inspiration and expiration, using sections of 1, 2, 3, 5, and 10 mm in combination with a standard, a smooth, and an ultrasmooth reconstruction kernel. The lungs were segmented using pixel tracing at thresholds of -200, -400, and -600 HU with 0-4 erosions, followed by histogram analysis. The area of the segmented lungs decreased with 0.9%-3.2% per 100 HU decrease in threshold and with 2.2%-3.1% per erosion, dependent on patient group and respiratory status. Estimated mean lung density changed up to 30% by changing the threshold and the number of erosions. The pixel index and the 10th percentile depended only slightly on threshold and number of erosions, but the 90th percentile showed a strong dependence of up to 40%. It is concluded that the segmentation protocol can have a large impact on densitometric parameters and that standardization is mandatory for obtaining comparable results. Ideally a threshold equal to the average of the densities of lung and soft tissue should be used, but -400 HU will do in a limited but common density range (-910 to -790 HU). For densitometry two erosions are recommended, for volumetry zero erosions should be used.

CT lung densitometry: dependence of CT number histograms on sample volume and consequences for scan protocol comparability.

PURPOSE: Our goals were to determine the dependence of CT number histograms of the lung on section thickness and reconstruction filter and to evaluate the consequences for scan protocol conformity required for universally comparable densitometry of the lungs. METHOD: The effects of section thickness and reconstruction filter were parameterized with the CT's sample volume [V approximately (section thickness x in-plane resolution2)]. In a study of 31 patients, we determined as a function of V the following CT number histogram parameters: percentiles P(10) and P(90), pixel indexes PI(-905) and PI(-950), and standard deviation. RESULTS: Patient histogram parameters depended strongly on sample volume. Large differences were found between protocols using 1 and 10 mm sections. For small variations in somewhat larger sample volumes (> 8 mm3), discrepancies were much smaller. CONCLUSION: To obtain comparable histogram parameters, nearly identical sample volumes (> or = 8 mm3) should be used. When this condition is satisfied, available data suggest that universally comparable densitometry is feasible.