Aki Kangasmäki

Evaluation of four postprocessing methods for determination of cerebral blood volume and mean transit time by dynamic susceptibility contrast imaging

Four different postprocessing methods to determine cerebral blood volume (CBV) and contrast agent mean transit time (MTT) by dynamic susceptibility contrast (DSC) MRI were compared. CBV was determined by two different methods that integrate tracer concentration–time curves numerically and by two other methods that take recirculation into account. For the two methods that use numerical integration, one method cuts the integration after the first pass while the other method integrates over the whole time series. For the two methods that account for recirculation, one method uses a gamma‐variate fit, whereas the other method utilizes tissue impulse response. All four methods determine MTT as the ratio of CBV and cerebral blood flow (CBF). In each case, CBF was obtained as the height of the impulse response obtained by deconvolving the tissue concentration–time curves with a noninvasively determined arterial input function. Monte Carlo simulations were performed to determine the reliability of the methods and the validity of the simulations was supported by observation of similar trends in 13 acute stroke patients. The method of determining CBV and subsequently MTT was found to affect the measured value especially in areas where MTT is prolonged, but had no apparent effect on the visually determined hypoperfusion volumes. Magn Reson Med 47:973–981, 2002.

Scintigraphy in prediction of the salivary gland function after gland-sparing intensity modulated radiation therapy for head and neck cancer

BACKGROUND AND PURPOSE To evaluate salivary gland scintigraphy in prediction of salivary flow following radiation therapy. PATIENTS AND METHODS Twenty patients diagnosed with head and neck cancer were treated with intensity modulated radiation therapy with an intention to spare the salivary gland function. The total quantitative saliva secretion was measured prior to and 6 and 12 months after therapy, and the function of the major salivary glands was monitored using Tc-99m-pertechnetate scintigraphy. Two models were designed for prediction of the post-treatment salivary flow: an average model, based on the average proportions of saliva produced by each of the four major glands in healthy subjects, and an individual model, based on saliva produced by each gland as measured by scintigraphy prior to therapy. These models were compared with volume-based (Lyman) normal tissue complication probability models using two published sets of model parameters. RESULTS The D(50) for the parotid and the submandibular gland function assessed at 6 and 12 months after radiotherapy was approximately 39Gy. The scintigraphy-based individual model predicted well the measured post-treatment saliva flow rates. The correlation coefficient between the predicted stimulated and the measured saliva flow rate was 0.77 (p<0.0001) at 6 months and 0.55 (p=0.034) at 12 months after completion of radiotherapy. The relative changes in unstimulated and stimulated salivary flow rates showed similar dependency on the cumulative radiation dose. CONCLUSIONS Salivary gland function assessed by scintigraphy prior to radiotherapy is useful in prediction of the residual salivary flow after radiotherapy.

Study of the relative dose-response of BANG-3® polymer gel dosimeters in epithermal neutron irradiation

Polymer gels have been reported as a new, potential tool for dosimetry in mixed neutron-gamma radiation fields. In this work, BANG-3 (MGS Research Inc.) gel vials from three production batches were irradiated with 6 MV photons of a Varian Clinac 2100 C linear accelerator and with the epithermal neutron beam of the Finnish boron neutron capture therapy (BNCT) facility at the FiR 1 nuclear reactor. The gel is tissue equivalent in main elemental composition and density and its T2 relaxation time is dependent on the absorbed dose. The T2 relaxation time map of the irradiated gel vials was measured with a 1.5 T magnetic resonance (MR) scanner using spin echo sequence. The absorbed doses of neutron irradiation were calculated using DORT computer code, and the accuracy of the calculational model was verified by measuring gamma ray dose rate with thermoluminescent dosimeters and 55Mn(n,gamma) activation reaction rate with activation detectors. The response of the BANG-3 gel dosimeter for total absorbed dose in the neutron irradiation was linear, and the magnitude of the response relative to the response in the photon irradiation was observed to vary between different gel batches. The results support the potential of polymer gels in BNCT dosimetry, especially for the verification of two- or three-dimensional dose distributions.

Brain function during multi-trial learning in mild cognitive impairment: a PET activation study.

We explored functional brain changes with positron emission tomography (PET) in mild cognitive impairment (MCI) patients and elderly normal controls by employing an episodic memory task that included two successive encoding trials of semantically related word-pairs and final retrieval. Both groups demonstrated significant learning across the two trials. The control group showed predominantly left frontal activity during encoding, and right frontal plus left temporal activity during retrieval. However, the MCI patients recruited partly different brain regions. They failed to activate right frontal and left temporal areas during retrieval, and failed to show any different activation for encoding on the first and second trials, whereas the controls activated a region of posterior cingulate. There was indication of compensatory increases in rCBF of the occipital cortex during incremental learning and the left frontal lobe during retrieval in the patients. These results suggest different episodic memory processing in the MCI group, and a possible over-reliance on semantic processing. Subtle functional changes occur in the pre-Alzheimer brain before there are marked structural or behavioural abnormalities.

Radiation exposure in body computed tomography examinations of trauma patients

Multi-slice CT provides an efficient imaging modality for trauma imaging. The purpose of this study was to provide absorbed and effective dose data from CT taking into account the patient size and compare such doses with the standard CT dose quantities based on standard geometry. The CT examination data from abdominal and thoracic scan series were collected from 36 trauma patients. The CTDI(vol), DLP(w) and effective dose were determined, and the influence of patient size was applied as a correction factor to calculated doses. The patient size was estimated from the patient weight as the effective radius based on the analysis from the axial images of abdominal and thoracic regions. The calculated mean CTDI(vol), DLP(w) and effective dose were 15.2 mGy, 431 mGy cm and 6.5 mSv for the thorax scan, and 18.5 mGy, 893 mGy cm and 14.8 mSv for the abdomen scan, respectively. The doses in the thorax and abdomen scans taking the patient size into account were 34% and 9% larger than the standard dose quantities, respectively. The use of patient size in dose estimation is recommended in order to provide realistic data for evaluation of the radiation exposure in CT, especially for paediatric patients and smaller adults.

Non-linear model for the kinetics of 10B in blood after BPA?fructose complex infusion

A numerical model with a memory effect was created to describe the kinetics of 10B in blood after a single 4-dihydroxyborylphenylalanine-fructose complex (BPA-F) infusion in boron neutron capture therapy (BNCT). The model formulation was based on the averaged data from 10 glioma patients from the Brookhaven National Laboratory (BNL) BNCT-trials. These patients received a 2 h i.v. infusion of a BPA-fructose complex that delivered 290 mg BPA/kg body weight. The model was validated by fitting the original BNL patient data and new patient data from the Finnish BNCT-trials. The new 3-parameter non-linear model provided mean absolute differences between the measured and estimated 10B concentrations in blood that were less than 3.9% when used to simulate actual patient irradiations that comprised two irradiation fields separated by a break to reposition the patient. The flexibility of the model was successfully tested with two different infusion protocols. The patient data were modelled with a two-compartment model and a bi-exponential fit for comparison. The 3-parameter model is better than previously described models in predicting the time course of blood 10B concentration after cessation of intravenous infusion of BPA-fructose.

Abnormal Intravoxel Cerebral Blood Flow Heterogeneity in Human Ischemic Stroke Determined by Dynamic Susceptibility Contrast Magnetic Resonance Imaging

Background and Purpose— The determination of cerebral blood flow heterogeneity (FH) by dynamic susceptibility contrast (DSC) magnetic resonance imaging has recently been proposed as a tool to predict final infarct size in acute stroke. In this study, we describe the evolution of FH during the first week as well as its correlation to the patients’ clinical status. Methods— Ten patients with ischemic stroke were studied with DSC MRI and diffusion-weighted imaging in hyperacute (<6 hours) phase, at 24 hours, and 1 week after symptom onset. In addition to intravoxel FH, cerebral blood volume (CBV), cerebral blood flow (CBF), and contrast agent mean transit time (MTT) were determined from DSC MRI. All patients were evaluated neurologically with National Institute of Health Stroke Scale concurrently with the imaging sessions. Results— All patients showed infarct growth, judged by diffusion-weighted imaging, during the week with simultaneous decrease in the sizes of FH, CBV, CBF, and MTT abnormalities. The FH abnormality was shown to be larger than CBV and CBF abnormalities at the hyperacute phase and 24 hours, but smaller than MTT abnormality in all 3 imaging sessions. The sizes of hyperacute FH, CBV, CBF, and MTT abnormalities correlated well with infarct size at 24 hours and at 1 week. Additionally, FH was the only perfusion parameter that correlated with the clinical score. Conclusions— FH predicts infarct size equally well with the other perfusion parameters but is superior in correlation with the clinical score. FH can easily be incorporated to hyperacute stroke imaging without additional efforts.

1H MRS of a boron neutron capture therapy 10B-carrier, L-p-boronophenylalanine-fructose complex, BPA-F: phantom studies at 1.5 and 3.0 T.

The quantification of a BNCT 10B-carrier, L-p-boronophenylalanine-fructose complex (BPA-F), was evaluated using 1H magnetic resonance spectroscopy (1H MRS) with phantoms at 1.5 and 3.0 T. For proper quantification, relaxation times T1 and T2 are needed. While T1 is relatively easy to determine, the determination of T2 of a coupled spin system of aromatic protons of BPA is not straightforward with standard MRS sequences. In addition, an uncoupled concentration reference for aromatic protons of BPA must be used with caution. In order to determine T2, the response of an aromatic proton spin system to the MRS sequence PRESS with various echo times was calculated and the product of the response curve with exponential decay was fitted to the measured intensities. Furthermore, the response curve can be used to correct the intensities, when an uncoupled resonance is used as a concentration reference. BPA was quantified using both phantom replacement and internal water referencing methods with accuracies of +/- 5% and +/- 15%. Our phantom results suggest that in vivo studies on BPA concentration determination will be feasible.

4D SPECT/CT acquisition for 3D dose calculation and dose planning in (177)Lu-peptide receptor radionuclide therapy: applications for clinical routine.

Molecular radiotherapy combines the potential of a specific tracer (vector) targeting tumor cells with local radiotoxicity. Designing a specific tumor-targeting/killing combination is a tailoring process. Radionuclides with imaging capacity serve best in the selection of the targeting molecule. The potential of targeted therapy with radiolabeled peptides has been reported in many conditions; peptide receptor radionuclide therapy (PRRT) is already part of Scandinavian guidelines for treating neuroendocrine tumors. Lu-177- and Y-90-labeled somatostatin analogs, including DOTATOC, DOTANOC, and DOTATATE, are most the commonly used and have turned out to be effective. For routine use, an efficient, rapid, and reliable dose calculation tool is needed. In this chapter we describe how serial pre- and posttherapeutic scans can be used for dose calculation and for predicting therapy doses. Our software for radionuclide dose calculation is a three-dimensional, voxel-based system. The 3D dose calculation requires coregistered SPECT image sets from several time points after infusion to reconstruct time-activity curves for each voxel. Image registration is done directly by SPECT image registration using the first time point as a target. From the time-activity curves, initial activity and total half-life maps are calculated to produce a cumulated activity map. The cumulated activity map is then convoluted with a voxel-dose kernel to obtain a 3D dose map. We performed dose calculations similarly for both therapeutic and preplanning images. Preplanning dose was extrapolated to predict therapy dose using the ratio of administered activities. Our 3D dose calculation results are also compared with those of OLINDA. Our preliminary results indicate that dose planning using pretherapeutic scanning can predict critical organ and tumor doses. In some cases, the dose planning prediction resulted in slight, and slightly dose-dependent, overestimation of final therapy dose. Real tumor dose was similar in both pretherapeutic and posttherapeutic scans using our software. The OLINDA software and our program gave similar normal organ doses, whereas tumor doses could be calculated in a more detailed manner using the 3D program.

Evaluation of Alpha-Therapy with Radium-223-Dichloride in Castration Resistant Metastatic Prostate Cancer—the Role of Gamma Scintigraphy in Dosimetry and Pharmacokinetics

Radium-223-dichloride (223RaCl2) is a new bone-seeking calcium analogue alpha-emitter, which has obtained marketing authorization for the treatment skeletal metastases of hormone-refractory prostate cancer. The current treatment regimen is based on six consecutive doses of 223RaCl2 at 4 week intervals and the administered activity dose, 50 kBq/kg per cycle is based on patient weight. We analyzed two patients using quantitative serial gamma imaging to estimate dosimetry in tumors and see possible pharmacokinetic differences in the treatment cycles. The lesions were rather well visualized in gamma scintigraphy in spite of low gamma activity (<1.1% gamma radiation) at 0, 7 and 28 days using 30–60 min acquisition times. Both our patients analyzed in serial gamma imagings, had two lesions in the gamma imaging field, the mean counts of the relative intensity varied from 27.8 to 36.5 (patient 1), and from 37.4 to 82.2 (patient 2). The half-lives varied from 1.8 days to 4.5 days during the six cycles (patient 1), and from 1.5 days to 3.6 days (patient 2), respectively. In the lesion half-lives calculated from the imaging the maximum difference between the treatment cycles in the same lesion was 2.0-fold (1.8 vs. 3.6). Of these patients, patient 1 demonstrated a serum PSA response, whereas there was no PSA response in patient 2. From our data, there were maximally up to 4.0-fold differences (62.1 vs. 246.6 ) between the relative absorbed radiation doses between patients as calculated from the quantitative standardized imaging to be delivered in only two lesions, and in the same lesion the maximum difference in the cycles was up to 2.3-fold (107.4 vs. 246.6). Our recommendation based on statistical simulation analysis, is serial measurement at days 0–8 at least 3 times, this improve the accuracy significantly to study the lesion activities, half-lives or calculated relative absorbed radiation doses as calculated from the imaging. Both our patients had originally two metastatic sites in the imaging field; the former patient demonstrated a serum PSA response and the latter demonstrated no PSA response. In these two patients there was no significant difference in the lesion activities, half-lives or calculated relative absorbed radiation doses as calculated from the quantitative imaging. Our results, although preliminary, suggest that dose monitoring can be included as a part of this treatment modality. On the other hand, from the absorbed radiation doses, the response cannot be predicted because with very similar doses, only the former patient responded.