Peter Bloch

Automatic registration of portal images and volumetric CT for patient positioning in radiation therapy

The efficacy of radiation therapy treatment depends on the patient setup accuracy at each daily fraction. A significant problem is reproducing the patient position during treatment planning for every fraction of the treatment process. We propose and evaluate an intensity based automatic registration method using multiple portal images and the pre-treatment CT volume. We perform both geometric and radiometric calibrations to generate high quality digitally reconstructed radiographs (DRRs) that can be compared against portal images acquired right before treatment dose delivery. We use a graphics processing unit (GPU) to generate the DRRs in order to gain computational efficiency. We also perform a comparative study on various similarity measures and optimization procedures. Simple similarity measure such as local normalized correlation (LNC) performs best as long as the radiometric calibration is carefully done. Using the proposed method, we achieved better than 1mm average error in repositioning accuracy for a series of phantom studies using two open field (i.e., 41 cm2) portal images with 90 degrees vergence angle.

NEUROPHYSIOLOGICAL AND NEUROPSYCHOLOGICAL FUNCTION IN MERCURY-EXPOSED DENTISTS

In a study of the relation between cumulative exposure to mercury and chronic health impairment 298 dentists had their mercury levels measured by an X-ray fluorescence technique. Electrodiagnostic and neuropsychological findings in the dentists with more than 20 micrograms/g tissue mercury levels were compared with those of a control group consisting of dentists with no detectable mercury levels. 30% of the 23 high mercury dentists had polyneuropathies. No polyneuropathies were detected in the control group. The high mercury group had mild visuographic dysfunction; they also had more symptom-distress than did the control group. These findings suggest that the use of mercury as a restorative material is a health risk for dentists.

Initiation of multileaf collimator conformal radiation therapy

Abstract Clinical studies have been initiated in conformal radiotherapy using a computer controlled multileaf collimator. Quantitative dosimetry and treatment planning studies comparing field shaping by lead alloy blocks and the multileaf collimator demonstrate the clinical acceptability of the multileaf collimator. Sixteen patients with tumors in multiple sites have received some part of their treatments with both blocking systems. Studies of dosimetry and field shaping show that the multileaf collimator produces clinically acceptable blocking for most field shapes and disease sites. The 80 – 20% penumbra was characterized for a wide range of shaped beams. For straight edges perpendicular to the leaf travel, the penumbra of measured dose distributions from the multileaf collimator is equal to conventional divergent blocking. When the multileaf collimator leaves approach a contour at an angle, the penumbra increases. At forty-five degrees, the maximum angle of approach, the penumbra is approximately 4 mm wider than that for divergent blocks. Three-dimensional treatment planning demonstrates that equivalent dose distributions can be obtained from the two field shaping systems. The multileaf collimator can be used effectively and efficiently to treat a variety of disease sites. Its optimal utility may be in treating complex fields—five or more shaped coplanar or non-coplanar beams. It is well suited for conformal therapy applications.

3D dose calculations for electron and photon beams

A new method has been implemented to calculate rapidly and accurately the three-dimensional dose distribution induced in a heterogeneous volume by one or more electron and/or photon beams. The algorithm is based on the pre-calculation by Monte Carlo of one or more (pencil beam) kernels, where a kernel is the distribution of absorbed energy density within a uniform medium by a unit of beam energy incident normally on a small area of the air-medium surface. Once the intensity and energy distributions of a clinical beam are deduced from measurements, the appropriate kernels can be obtained for that beam. Kernels allowed one to estimate how each pencil of a beam deposits its energy in an inhomogeneous medium. When the patient contours and density distribution are known, and beam parameters are specified, the beam energy delivered to each element of the patient surface is calculated and then distributed within the patient volume in accord with the relevant kernels.

The use of magnetic resonance imaging and spectroscopy in the assessment of patients with head and neck and other superficial human malignancies

The proper demarcation of diseased tissue is important for radiation therapy planning and treatment. The volume to be irradiated is usually identified on radiographs or on x‐ray computed tomography (CT) sections. Magnetic resonance (MR)‐derived images of the proton T2 relaxation times in small pixel elements, typically 0.5 mm2 or less, provide significantly sharper differentiation between normal and diseased tissue. The T2 values in tissue depend on the tissue composition, histologic condition, and physiologic environment within the tumor. Furthermore, for many tumors the histogram of T2 values has a clear biphasic distribution suggesting that T2 maps may be useful for the identification of necrotic or hypoxic regions within tumors. The distribution of T2 values within the tumor bed shows the general pattern that the T2 values are elevated with a range greater than that seen in normal muscle. Elevated T2 values are not by themselves diagnostic of malignancy; however, they demonstrate the heterogeneity of the microenvironment present within a tumor. The spatial distribution of T2 values is being explored as a method for computer assistance in the delineation of the target volume for treatment planning. In addition, MR P‐31 spectroscopic examinations were performed on 30 patients with squamous cell carcinomas of the head and neck. Although hampered by muscle contamination in some P‐31 spectra obtained with surface coil profile localization techniques, significant trends can still be appreciated in our data. These trends include the following: (1) the P‐31 spectra from malignant tissue have well‐resolved spectral lines in the upfield region that correspond to Pi, phosphomonoester (PME), and phosphodiester (PDE) not usually seen in normal muscle; (2) the PDE/B‐ATP and PME/B‐ATP ratios are greater than unity in all cases; and (3) most of the tumors have higher PME peaks than PDE peaks. The P‐31 spectra from patients treated with ionizing radiation changed during and after therapy. Some of the changes could be associated with alteration of the tumor metabolic activity or synthesis and breakdown of lipoproteins. These studies suggest that magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) studies may be useful for both radiotherapy treatment planning and the noninvasive monitoring of patients both before and during treatment.

X-ray fluorescence analysis of lead in teeth of urban children in situ: correlation between the tooth lead level and the concentration of blood lead and free erythroporphyrins.

The tooth lead level of 30 lead-exposed children was measured in situ using an X-ray fluorescence technique. The tooth lead concentration divided by the childs age correlated with the free erythroporphyrin (FEP) (r = 0.51) and the blood lead levels (r = 0.31). The mean tooth lead concentrations of 10 Class IV children (FEP greater than or equal to 190 micrograms/100 ml; blood lead 30-80 micrograms/ml) was 14.5 +/- 5.5 ppm/year. Reexamination of the histories of children who had been classified as Class II or III, but who had tooth lead concentrations within 1 SD of the Class IV level, indicated that many of these children could also be considered to be Class IV children. When the x-ray fluorescence technique was used to screen an urban population of 300 children, the tooth lead values indicated that 72% of the children had been exposed to low levels of environmental lead. Six percent of the children were found to have tooth lead concentrations in excess of 9 ppm/year and within 1 SD of the mean value exhibited by Class IV children. These children were considered to have a high body lead burden. The percentage of children having an elevated tooth lead level is similar to the number previously reported using exfoliated deciduous teeth.

An x-ray fluorescence technique to measure the mercury burden of dentists in vivo.

Mercury exposure in dental offices may represent a health hazard. Previous studies evaluated mercury exposure through blood, urine or hair analysis. The mercury content of body fluids in hair depends on the time since mercury exposure, mode of intake and its excretion, and sequestration by different tissues. Utilizing an x-ray fluorescence technique, the mercury burden in bone, liver, lung, spleen, and kidney was measured in vitro. The x-ray fluorescence, XRF, and chemical assays of the mercury content of these samples correlated. The XRF technique was used to assay the mercury content of tissues irradiated in vivo. It was found that the mercury signal arising at a depth in tissue was reduced as the thickness of overlying material increased. Thus, the XRF assay for tissues in vivo is confined to organs near the surface. The mercury content of the superficial layers of the head and wrist were determined in situ for a dental population. Concentrations as low as 20 micrograms/g could be detected with an exposure of 60 mR. Of the 298 dentists evaluated, 28% had greater than 20 micrograms/g of mercury in the head.

Measurement of lead content of children's teeth in situ by X-ray fluorescence

The central nervous system of young children is particularly sensitive to neurological damage due to excessive lead exposure. The lead stored in a tooth is permanent and related to the quantity of ingested lead. The lead is assayed by measuring the characterisitc radiation from lead in a tooth in situ irradiated with gamma rays from a cobalt-57 source. The sensitivity of the method is adequate to measure lead levels commonly associated with environmental background. The X-ray exposure to the tooth necessary for this determination is approximately one tenth of that required for a routine dental examination.

Preliminary Investigation of the Two‐Photon Photoelectric Effect

The photoelectric current due to the absorption of two photons has been calculated. Its magnitude is dependent upon the square of the incident radiation intensity, whereas, for the photoelectric effect due to the absorption of one photon, the magnitude of the current is linearly dependent upon the intensity.The magnitude of the photoelectric current due to the absorption of two photons is small. However, under intense radiation, such as is obtainable from a ruby laser, a photoelectric current due to two‐photon absorption should be observed.

Extraction of the photon spectra from measured beam parameters

Knowledge of the photon spectrum of a radiotherapy beam is often needed for three-dimensional (3-D) dose calculations using Monte Carlo methods and/or algorithms employing energy deposition kernels. Direct measurement of the x-ray energy fluence spectrum is not feasible for the high-energy photon beams used clinically. In this paper, the spectrum is extracted from basic beam data that are readily obtained for a clinical beam. We describe the photon spectrum using just two parameters. One parameter, which determines the high-energy part of the spectrum, is obtained using the measured dose in the buildup region for a small field, where electron contamination of the beam can be neglected. The other parameter is extracted from the photon beam attenuation in water. The results compare favorably to spectra generated from Monte Carlo simulations.