Mean Pixel Value Research Articles

SU-E-I-45: Feasibility for Using Iodine Quantification to Assist Diagnosis in Dual Energy Contrast-Enhanced Digital Mammography

Purpose: The objective of this study is to develop quantitative calibration between image quality indexes and iodine concentration with dual-energy (DE) contrast-enhanced digital mammography (CEDM) techniques and further serve as the assistance for diagnosis. Methods: Custom-made acrylic phantom with dimensions of 24×30 cm{sup 2} simulated breast thickness from 2 to 6 cm was used in the calibration. The phantom contained matrix of four times four holes of 3 mm deep with a diameter of 15 mm for filling contrast agent with area density ranged from 0.1 to 10 mg/cm{sup 2}. All the image acquisitions were performed on a full-field digital mammography system (Senographe Essential, GE) with dual energy acquisitions. Mean pixel value (MPV), and contrast-to-noise ratio (CNR) were used for evaluating the relationship between image quality indexes and iodine concentration. Iodine map and CNR map could further be constructed with these calibration curves applied pixel by pixel utilized MATLAB software. Minimum iodine concentration could also be calculated with the visibility threshold of CNR=5 according the Rose model. Results: When evaluating the DE subtraction images, MPV increased linearly as the iodine concentration increased with all the phantom thickness surveyed (R{sup 2} between 0.989 and 0.992). Lesions with increased iodine uptake couldmore » thus be enhanced in the color-encoded iodine maps, and the mean iodine concentration could be obtained through the ROI measurements. As for investigating CNR performance, linear relationships were also shown between the iodine concentration and CNR (R{sup 2} between 0.983 and 0.990). Minimum iodine area density of 1.45, 1.73, 1.80, 1.73 and 1.72 mg/cm{sup 2} for phantom thickness of 2, 3, 4, 5, 6 cm were calculated based on Rose’s visualization criteria. Conclusion: Quantitative calibration between image quality indexes and iodine concentrations may further serving as the assistance for analyzing contrast enhancement for patient participating the dual energy CEDM procedures.« less

TU-CD-207-03: Time Evolution of Texture Parameters of Subtracted Images Obtained by Contrast-Enhanced Digital Mammography (CEDM)

Purpose: To evaluate the time evolution of texture parameters, based on the gray level co-occurrence matrix (GLCM), in subtracted images of 17 patients (10 malignant and 7 benign) subjected to contrast-enhanced digital mammography (CEDM). The goal is to determine the sensitivity of texture to iodine uptake at the lesion, and its correlation (or lack of) with mean-pixel-value (MPV). Methods: Acquisition of clinical images followed a single-energy CEDM protocol using Rh/Rh/48 kV plus external 0.5 cm Al from a Senographe DS unit. Prior to the iodine-based contrast medium (CM) administration a mask image was acquired; four CM images were obtained 1, 2, 3, and 5 minutes after CM injection. Temporal series were obtained by logarithmic subtraction of registered CM minus mask images.Regions of interest (ROI) for the lesion were drawn by a radiologist and the texture was analyzed. GLCM was evaluated at a 3 pixel distance, 0° angle, and 64 gray-levels. Pixels identified as registration errors were excluded from the computation. 17 texture parameters were chosen, classified according to similarity into 7 groups, and analyzed. Results: In all cases the texture parameters within a group have similar dynamic behavior. Two texture groups (associated to cluster and sum mean) show a strong correlation with MPV; their average correlation coefficient (ACC) is r^2=0.90. Other two groups (contrast, homogeneity) remain constant with time, that is, a low-sensitivity to CM uptake. Three groups (regularity, lacunarity and diagonal moment) are sensitive to CM uptake but less correlated with MPV; their ACC is r^2=0.78. Conclusion: This analysis has shown that, at least groups associated to regularity, lacunarity and diagonal moment offer dynamical information additional to the mean pixel value due to the presence of CM at the lesion. The next step will be the analysis in terms of the lesion pathology. Authors thank PAPIIT-IN105813 for support. Consejo Nacional de Ciencia Y Tecnologia, PAPIIT-IN105813

SU-F-207-11: On the Relationship of Minimum Required Dose to Low Contrast Detectability and Blending Fractions With Adaptive Statistical Iterative Reconstruction (ASIR) in Abdominal CT

Purpose: It is of practical use to obtain the dose for a specified low contrast lesion detectability and a given ASIR fraction before the scan. To achieve the goal, we investigated the relationship of the minimum required dose to low contrast detectability and ASIR blending fraction in abdominal CT scans. Methods: A realistically shaped abdomen phantom (CIRS TE-07) with an embedded vertebral bone rod was utilized. Helical CT scans were performed on a GE 750 HD at two dose levels of 8 mGy and 18 mGy. The images were reconstructed using different ASIR fractions (0–100%). The noise images, obtained using subtraction from adjacent slices, were partitioned into different scales using matrices of square cells sized differently. The noise at each cell size can be quantified by the standard deviation (sd) of the cell mean pixel values distribution. Upon verifying that the distributions follow Gaussian, we proposed that the statistically defined minimum detectable contrast (MDC) can be applied to ASIR reconstructed images, i.e., MDC = 3.29sd. We then defined the normalized MDC (N_MDC) as the ratio of the MDC with ASIR to the MDC with filtered backprojection and postulated that N_MDC is dose independent. N_MDC was obtained at each cell size from different ASIR blending fractions at different dose levels. Results: The matrix cell means distribution was verified following Gaussian. N_MDC was found to linearly decrease as the ASIR blending fraction increases. The relationship is the same with two distinct doses. For the specified MDC at lesion size d and ASIR blending fraction (s), the minimum required dose D = [14.05(1-bs)/(MDC*d^1.12)]^2.03, where b = 0.005/d^0.784. Conclusion: The relationship of the minimum required dose to low contrast detectability and ASIR blending fraction was obtained. It can be used to adjust the scan parameters corresponding to the calculated minimum required dose.

Treatment of Ovarian Cysts in Buffaloes with Emphasis to Echotexture Analysis

The present study was conducted on 31 buffalo-cows aged 3-10 years and diagnosed rectally as they had follicular cyst. The previous diagnosis of the follicular cyst was confirmed by ultrasonography and serum progesterone assay. Ultrasound images of the ovarian follicles, both mature and cystic follicles, were individually identified and monitored and were subsequently digitized for computer analysis of echotexture [mean pixel value (MPV) and pixel heterogeneity (PH)] of the follicle antrum and wall. Animals were classified into two groups (Treated, n=20 and control, n=11). The treated buffalo-cows injected i.m with 100mg progesterone oily in nature followed by i.m injection of 5000 IU HCG at days 5. Out of the twenty treated animals 14 buffaloes responded and diagnosed as they had luteinized cyst. Out of the 14 animals that had luteinized cyst 9 buffaloes injected with 25 mg of PGF2α at day 14 and the other 5 buffaloes left as control. The 6 non responded buffaloes were classified into two subgroups each (n=3), one of them injected with HCG and the other left as control without any treatment. The time elapsed till luteinization of the follicular treated group was significantly different (32.67 ± 1.45 days and 15.71 ± 0.75 days respectively). The recovery rate was 27.27 % in control animals and 70% in treated ones. Days elapsed till appearance of estrus was significantly shorter in treated buffaloes (22.20 ± 1.19 days) in comparison to the control ones (56.33 ± 2.03 days). The conception rates were 33.33% and 66.67% in control and treated group respectively. Days elapsed from the treated animals (36.50 ± 1.47 days) compared to the control ones (64 days). Serum progesterone level was significantly different in treated animals comparing to the control ones. Changes in echotexture of bovine ovarian follicles assessed by computer analysis of ultrasound images has revealed a significant increase in echotexture parameters (MPV & PH) in cystic follicles compared with preovulatory follicles. PH of the antrum and wall of luteinized cystic follicles increased significantly compared to preovulatory and cystic follicles. Ultrasonography examination and echotexture analysis has revealed a presence of positive correlation between the cyst wall thickness and progesterone level which consequently aids in the treatment protocol. From the present study, it could be recommended that accurate diagnosis of the ovarian cyst was done by ultrasonography in combination with echotexture characteristics of ultrasound images together with progesterone profile 2 weeks postpartum early as soon as possible; and the animals should be re-examined with 7-10 days if possible for re-treatment of non responded cases if necessary.

Wavelength optimized cross-polarized wide-field imaging for noninvasive and rapid evaluation of dermal structures.

Changes in the morphology of dermal collagen may indicate aging or pathological processes. At present, there is no technology for in vivo real-time assessment of collagen structures. Our goal was to introduce and validate polarization optical imaging for noninvasive quantitative evaluation of dermal collagen. Seventeen volunteers participated in the study. Cross-polarized 440nm images were acquired noninvasively from facial skin of the study subjects. Collagen content and intensity histogram were computed from the optical images. Quantitative results showed a decrease in the collagen content with increasing age of the subjects. Analysis of the collagen image histogram parameters demonstrated decreasing mean pixel value and increasing full width at half maximum (FWHM) with increasing age. Polarization optical imaging has the potential for rapid noninvasive in vivo evaluation of human dermis. Cross-polarized imaging at different wavelengths emphasizes different features of human skin. (A) Skin photograph. (B) In vivo cross-polarized image at 440 nm. (C) In vivo cross-polarized image at 570nm. (D) In vivo cross-polarized image at 690nm.

Feasibility of percutaneous contrast ultrasound-guided cholecystography in dogs.

Differentiating hepatocellular disease versus biliary obstruction can be challenging in dogs presented for icterus. The purpose of this prospective study was to determine the feasibility of percutaneous contrast ultrasound-guided cholecystography in dogs. Ten normal dogs weighing 7.6-13.0 kg (median 9.8 kg) were recruited. All dogs were considered normal based on complete blood count, serum chemistry profile, ultrasound examination, and percutaneous radiographic cholecystography. Percutaneous contrast ultrasound-guided cholecystography was performed using 0.5 ml of commercially available contrast agent and two conventional ultrasound machines for simultaneous scanning at two different locations. Two observers independently evaluated the time to initial detection of contrast in the proximal duodenum and duration of contrast enhancement via visual monitoring. Dynamic contrast enhancement was calculated using time-intensity curves. Mean (± SD) and median (range) of time to initial detection were 8.60 s (± 3.35) and 8.0 s (2.0-11.0), respectively, and mean and median duration were 50.45 s (± 23.24) and 53.0 s (20.0 - 70.0), respectively. Mean, median, and range of peak intensity were 114.1 mean pixel value (MPV) (SD ± 30.7), 109.2 MPV, and 79.7-166.7, respectively, and mean, median, and range of time to peak intensity were 26.1 s (SD ± 7.1 s), 24.0 s, and 19.0-41.0 s, respectively. Findings indicated that percutaneous contrast ultrasound-guided cholecystography is a feasible technique for detecting and quantifying patency of the bile duct in normal dogs. Future studies are needed to assess the diagnostic utility of this technique for dogs with biliary obstruction.

SU‐E‐T‐497: Semi‐Automated in Vivo Radiochromic Film Dosimetry Using a Novel Image Processing Algorithm

Purpose:To validate an automated image processing algorithm designed to detect the center of radiochromic film used for in vivo film dosimetry against the current gold standard of manual selection.Methods:An image processing algorithm was developed to automatically select the region of interest (ROI) in *.tiff images that contain multiple pieces of radiochromic film (0.5×1.3cm2). After a user has linked a calibration file to the processing algorithm and selected a *.tiff file for processing, an ROI is automatically detected for all films by a combination of thresholding and erosion, which removes edges and any additional markings for orientation. Calibration is applied to the mean pixel values from the ROIs and a *.tiff image is output displaying the original image with an overlay of the ROIs and the measured doses. Validation of the algorithm was determined by comparing in vivo dose determined using the current gold standard (manually drawn ROIs) versus automated ROIs for n=420 scanned films. Bland‐Altman analysis, paired t‐test, and linear regression were performed to demonstrate agreement between the processes.Results:The measured doses ranged from 0.2‐886.6cGy. Bland‐Altman analysis of the two techniques (automatic minus manual) revealed a bias of ‐0.28cGy and a 95% confidence interval of (5.5cGy,‐6.1cGy). These values demonstrate excellent agreement between the two techniques. Paired t‐test results showed no statistical differences between the two techniques, p=0.98. Linear regression with a forced zero intercept demonstrated that Automatic=0.997*Manual, with a Pearson correlation coefficient of 0.999. The minimal differences between the two techniques may be explained by the fact that the hand drawn ROIs were not identical to the automatically selected ones. The average processing time was 6.7seconds in Matlab on an IntelCore2Duo processor.Conclusion:An automated image processing algorithm has been developed and validated, which will help minimize user interaction and processing time of radiochromic film used for in vivo dosimetry.

Landsat TM 영상과 현장조사를 이용한 잣나무림 재적 추정

본 연구는 강원대학교 학술림을 대상으로 조사한 42개 표본점의 재적정보와 Landsat TM 영상으로 추출한 Remote Sensing(RS)정보를 이용하여 잣나무임분의 재적을 추정하는 것을 목표로 한다. 실험 대상 학술림 잣나무림의 ha당 평균재적은 <TEX>$307.7m^3/ha$</TEX>, 표준편차는 <TEX>$168.4m^3/ha$</TEX>이며 산출된 잣나무림 재적을 등급화하였다. TM 영상에 3 by 3 majority filtering을 수행하기 전과 후에 각각 11개의 지수를 추출하였으며, 지수별 평균 화소 값을 이용하여 선형 회귀식 도출에 필요한 독립변수를 선정하였다. 11개의 지수는 6개의 DN(밴드값, 열감지밴드인 Band6을 제외), NDVI(정규식생지수), Band Ratio(BR1:Band4/Band3, BR2:Band5/Band4, BR3:Band7/Band4), Tasseled Cap-Greeness(TC G) 1개로 구성하였다. 그 결과, 필터링 전과 후 모두 NDVI와 TC G가 회귀식에 가장 적합한 지수로 선정되었으며, <TEX>$R^2$</TEX>는 필터링 전과 후가 각각 0.736, 0.753로 모두 높았다. 또한, 정확도 비교를 위하여 오차검증을 실시한 결과, RMSE는 필터링 전과 후가 각각 약 <TEX>$69.1m^3/ha$</TEX>, 약 <TEX>$67.5m^3/ha$</TEX>으로 필터링 후가 낮았으며, bias는 각각 약 <TEX>$-12.8m^3/ha$</TEX>, 약 <TEX>$9.7m^3/ha$</TEX>으로 필터링 후의 편차가 적어 필터링을 실시한 회귀식이 적합한 모형으로 선정되었다. 필터링 후의 회귀식을 적용하여 추정한 임반별 재적은 총 재적이 약 <TEX>$160,947m^3$</TEX>이며, 평균 재적은 약 <TEX>$315m^3/ha$</TEX>로 실제 잣나무림의 재적보다 약 1.2배 높게 추정되었다. The objective of this research is to estimate the stand volume of Pinus koraiensis, by using the investigated volume and the information of remote sensing(RS), in the research forest of Kangwon National University. The average volume of the research forest per hectare was <TEX>$307.7m^3/ha$</TEX> and standard deviation was <TEX>$168.4m^3/ha$</TEX>. Before and after carrying out 3 by 3 majority filtering on TM image, eleven indices were extracted each time. Independent variables needed for linear regression equation were selected using mean pixel values by indices. The number of indices were eleven: six Bands(except for thermal Band), NDVI, Band Ratio(BR1:Band4/Band3, BR2:Band5/Band4, BR3:Band7/Band4), Tasseled Cap-Greeness. As a result, NDVI and TC G were chosen as the most suitable indices for regression before and after filtering, and R-squared was high: 0.736 before filtering, 0.753 after filtering. As a result of error verification for an exact comparison, RMSE before and after filtering was about <TEX>$69.1m^3/ha$</TEX>, <TEX>$67.5m^3/ha$</TEX>, respectively, and bias was <TEX>$-12.8m^3/ha$</TEX>, <TEX>$9.7m^3/ha$</TEX>, respectively. Therefore, the regression conducted with filtering was selected as an appropriate model because of low RMSE and bias. The estimated stand volume applying the regression was <TEX>$160,758m^3$</TEX>, and the average volume was <TEX>$314m^3/ha$</TEX>. This estimation was 1.2 times higher than the actual stand volume of Pinus koraiensis.

Deep brain stimulation surgery under general anaesthesia with microelectrode recording: the best of both worlds or a little bit of everything?

<b>2639</b> <h3><b>Objectives</b></h3> Most aged people will experience brain atrophy and decrease of cerebral blood flow. In order to avoid missing brain lesions, we apply brain processing software NeuroGam with SPECT to evaluate and analyze the brain cerebral cortical blood flow dynamic data in elderly patients. <h3><b>Methods</b></h3> Siemens Symbia T16 SPECT/CT, patients were injected 99Tcm-ECD 1480Mbq. 10 male patients aged 70 years or older who underwent SPECT/CT in period from May to July 2013. Group A of five cases who were group control without cerebral infarction, cerebral hemorrhage, senile dementia, Parkinson9s disease, or any organic or functional impairments. Group B of 5 cases who had unilateral old cerebral infarction or cerebral hemorrhage. <h3><b>Results</b></h3> NeuroGam software offers each corresponding region of interest expressed as a percentage of the mean pixel value in the entire volume, including the anterior cerebral, middle cerebral, and vertebral terrain. The data of Group A, the left and right side of the mean blood flow did not exceed more than 1.8%. The data of Group B, according to the lesion location in different regions, left and right side of the mean blood flow blood vessels were more than 2.5%. Except for the temporal lobe and the cerebellar cortex. The data of Group A, the difference between left and right side of the mean blood flow did not exceed more than 1.2%. The data of Group B, according to the lesion location in different regions , the difference between left and right side of the mean changes in blood flow were more than 2%. Some data of Group A in temporal lobe and cerebellar cortex appears more than 2% difference, but there is no clinical evidence of other test results. Compared with traditional nuclear medicine images, when the mean blood flow changes more than 4%, it can show the difference. When the mean blood flow changes more than 6%, it can show a significant difference in visual form. <h3><b>Conclusions</b></h3> NeuroGam software can quickly reflect the changes in the cerebral cortex between left and right side. However, when it analyzes the changes in the temporal lobe and cerebellum regions, it may be false positives.

A morphology-stitching method to improve Landsat SLC-off images with stripes

Abstract: Stripes are artifacts in satellite images caused by various factors such as hardware defects. In some cases, these artifacts are introduced by some mitigating algorithms like Landsat SLC-off (Scan Line Corrector) gap-filling methods of LLHM (Local Linear Histogram Matching) and AWLHM (Adaptive Window Linear Histogram Matching), which leave stripes as a byproduct. To improve Landsat SLC-off images with stripes, we propose an algorithm involving some hypothetical stripe-crossing stitch lines using the mean pixel value of the stitch lines.

SU‐E‐T‐185: Time‐Dependent Calibration Correction for EBT Film Dosimetry

Purpose: The aim of this study is to simplify EBT film‐based dosimetry calibration. In addition, two different approaches are proposed for film and scanner non‐uniformity reduction. Methods: EBT films, EBT3 and EBT2 were exposed under a calibration plan using 6X flattening‐filter‐free (FFF) photon beams. In method 1, pre‐irradiation scans were carried out and subtracted from post‐irradiation scans to calculate net‐optical density (net OD), removing non‐uniformity from both scanner and film. In method 2, two unexposed films were independently scanned for each irradiated film scan and subtracted from post‐irradiation scans. Calibration curves were then generated using the calibration plan dose and the film net OD values. Finally, a set of time‐dependent correction factors were derived from the calibration curves. The per‐batch average correction factors were applied in order to test the feasibility of using a per‐batch generic calibration curve. All converted film doses were verified against the planned dose. The scanner light effect on the film dose was estimated by measuring the changes in mean pixel value. Results: Up to 7% pixel value change was noted four days after exposure. A correction was necessary if a generic calibration curve was utilized. The method to remove non‐uniformity using the pre‐exposure film (method 1) was more stable than using two unexposed films (method 2). The batch average correction factors reduced calibration curve differences to less than 0.6%. The dose effect from each scan was minimal (∼0.1 cGy). Conclusion: It is feasible to use one generic calibration curve per film batch with a time‐dependent correction factor table without compromising overall EBT film dosimetry accuracy. This method may alleviate the current practice of the typical labor intensive film‐based dosimetry for patient specific QA as well as physics commissioning tasks.

WE-G-103-01: Dual-Energy Temporal Study of Contrast-Enhanced Digital Mammography in 20 Patients: The Use of a Pixel-By-Pixel Weighting Formalism for the Subtraction

Purpose: To study the possible correlation between iodine uptake in contrast-enhanced digital mammography images and microvessel density (MVD) in breast lesions. Methods: 20 BIRADS 4–5 patients were included. Mask images were acquired with low-and high-energy spectra from a SenographeDS. Iodine-based contrast medium (CM) was mechanically injected and high-energy CM images acquired between 1 and 5 minutes after injection. A biopsy was obtained after the images and specific biomarkers for newly-formed blood and lymphatic microvessels were applied; MVD was later evaluated in microscope fields. The subtraction combined dual-energy and temporal modalities. LE masks were subtracted from weighted CM images, and weight factor was a matrix obtained from the masks, containing pixel-by-pixel anatomical and radiological information. Iodine uptake in subtracted temporal series was quantified by contrast between lesion and normal glandular tissue. Alternative metrics for contrast quantification were evaluated. Contrast was transformed into iodine mass-thickness using calibrated samples. Results: 12 cases were malignant and 8, benign. The matrix-based subtraction formalism severely reduced anatomic noise in resulting images, compared with alternative techniques based on mean pixel values within regions-of-interest. Five types of time-intensity curves were identified, qualitatively similar to MRI. If iodine uptake in the lesion was evaluated, most frequent curve was plateau type, for cancer and malignant cases. The normalized Weber contrast indicated a frequency distribution with zero-contrast cases for benign lesions and most-frequent curve type washout for cancer. No correlation was found between contrast indicators and MVD. Blood and lymphatic MVD were correlated (r=0.94, p<0.05) and mean blood MVD values in cancer were about twice those in benign cases. Conclusion: This subtraction formalism, previously validated in non-homogeneous phantom images, adds quantitative features to the resulting clinical images. The statistical distribution of contrast curves for cancer and benign cases seems to be the most relevant parameter with a possible diagnostic value. Funding: DGAPA-UNAM Grants 102610 and 105813; Conacyt Grant Salud 2009-01-112374. No conflict of interest.

Metal Artifacts From Posterior Mandibular Implants as Seen in CBCT

Metal artifacts from dental implants have not been clarified in cone-beam computed tomography (CBCT) images in previous studies. Hence, only mesiodistal metal artifacts from dental implants in the posterior mandibular region were assessed using CBCT. Twenty-two patients with a total of 61 dental implants in posterior mandibular sites were enrolled. The pixel values surrounding dental implants were measured using CBCT and rotational panoramic radiography, and then the means were calculated. The mean pixel values between dental implants, between dental implants and neighboring teeth, and posterior to dental implants were significantly lower than between neighboring teeth in CBCT. In addition, the mean pixel value between dental implants did not significantly differ from that between neighboring teeth in rotational panoramic radiography. It was concluded that mesiodistal metal artifacts surrounding dental implants are demonstrated on CBCT.

A SPATIAL DOMAIN OBJECT SEPARABILITY BASED NO-REFERENCE IMAGE QUALITY MEASURE USING MEAN AND VARIANCE

In many modern image processing applications determining quality of the image is one of the most challenging tasks. Researchers working in the field of image quality assessment design algorithms for measuring and quantifying image quality. The human eye can identify the difference between a good quality image and a noisy image by simply looking at the image, but designing a computer algorithm to automatically determine the quality of an image is a very challenging task. In this paper, we propose an image quality measure using the concept of object separability. We define object separability using variance. Two objects are very well separated if variance of individual object is less and mean pixel values of neighboring objects are very different. Degradation in images can be due to a number of reasons like additive noises, quantization defects, sampling defects, etc. The proposed no-reference image quality measure will determine quality of degraded images and differentiate between good and degraded images.

Preliminary assessment of the quantitative relationships between testicular tissue composition and ultrasonographic image attributes in the ram

Abstract This study examined the relationships between gross chemical composition and ultrasonographic characteristics of the ram testes. Ten testes from sexually mature Karakul rams were scanned ex situ with an 8-MHz linear-array transducer, in a transverse and longitudinal plane. All ultrasonograms were saved as digital images and subjected to computerized analyses. Crude protein content was determined by the Kjeldahl method, moisture was determined with an oven-drying method, and fat was measured by the Soxhlet extraction of dried samples. Mean pixel values (r = −0.64, P = 0.04), pixel heterogeneity (standard deviation of pixel values; r = −0.64, P = 0.04) and maximum pixel intensity (r = −0.76, P = 0.01) were all negatively correlated with parenchymal protein content. Pixel heterogeneity correlated directly with extractable lipids (r = 0.66, P = 0.02). The quantitative correlations between echotextural and biochemical parameters found in the present experiment confirm the utility of ultrasonographic imaging combined with computer-assisted image analysis for determining changes in testicular histophysiology.

Monitoring of Drusen and Geographic Atrophy Area Size after Cataract Surgery Using the MD3RI Tool for Computer-Aided Contour Drawing

Background/Aims: To monitor possible changes in the cumulated drusen or geographic atrophy area size (CDGAS) of nonexudative age-related macular degeneration (AMD) in patients before and after cataract surgery, using a new tool for computer-aided image quantification. Methods: Randomized, prospective, clinical trial. 54 patients with cataract and nonexudative AMD were randomly assigned into an early surgery group (ES = 28) and a control group (CO = 26) with a 6-month delay of surgery. CDGAS was determined with the MD3RI tool for contour drawing in a central region of digitized fundus photographs, measuring 3,000 µm in diameter. To evaluate CDGAS progression, differences in pixels and square millimeters were calculated by equivalent tests. Results: Forty-nine patients completed the visits over the 12-month period (ES = 27 and CO = 22). Mean pixel values increased from 201.5 (11.33 × 10^–3 mm²) to 202.7 (11.39 × 10^–3 mm²) in the ES group and from 191.6 (10.77 × 10^–3 mm²) to 194.6 (10.94 × 10^–3 mm²) in the CO group. Finally, equivalence of CDGAS differences between ES and CO could be demonstrated. No exudative AMD was recorded during the study period. Conclusion: In our cohorts, no significant changes were found in CDGAS 12 months after cataract surgery. The MD3RI software could serve as an efficient, precise and objective tool for AMD quantification and monitoring in future trials.

174 ULTRASOUND BIOMICROSCOPIC IMAGE ATTRIBUTES OF OVARIAN FOLLICLES IN HEIFERS

High-resolution ultrasound biomicroscopy (UBM) with high-frequency transducers (20 to 50 MHz) has enabled imaging of small antral follicles (&lt;1 mm) and acoustic characterisation of the follicular antrum and wall. Quantitative echotextural analysis of the images generated by UBM provides detailed functional information about the granulosa and theca layers of follicle walls. The objectives of our study were to compare biomicroscope image echotextures between dominant and largest subordinate follicles during the first follicular wave and to characterise UBM attributes of the ovulatory follicle. Hereford crossbred heifers (14 to 16 months old, n = 12) were used. Ovarian examinations were performed transvaginally from Day –4 or –3 (Day 0 = ovulation) to Day 10 using a UBM imaging instrument (Vevo 660; Visual Sonics Inc., Toronto, Canada) equipped with a 25-MHz end-fire transducer. Echotextural assessment was performed by spot metrics of the follicle antrum and follicle wall by using a series of custom-developed computer algorithms optimized for ultrasonography (Synergyne©, Version 2.8, WHIRL, Saskatoon, SK, Canada). Although subordinate and dominant follicles development were tracked from Day –2 (around 1 mm in diameter) to Day 10, the ovulatory follicle was tracked only from Day –4 to ovulation. All data were analysed by analysis of variance for repeated measures using the Mixed procedure (Littell et al. 1998) in the Statistical Analysis System software package (SAS version 8.2 for MS Windows; SAS Institute Inc., Cary, NC, USA). No differences were detected between dominant and largest subordinate follicle in the mean pixel value (MPV) or pixel heterogeneity (PH) of the antrum. However, the MPV of the wall of dominant follicles was greater and PH tended to be greater than that of the largest subordinate follicles (P = 0.05 and P = 0.1, respectively). This result could be evidence that the dominant follicle would be healthier. The PH of the wall the ovulatory follicle decreased (P = 0.03) from Day –4 to Day –1; however, no changes in MPV were observed. Similarly, no changes were detected in MPV or PH of the follicular antrum. The loss of granulosa cells from the follicular wall can explain, at least in part, the decrease in mean pixel value and pixel heterogeneity of the follicular wall detected in ovulatory follicles. Moreover, morphologic changes indicative of atresia are typically represented by high values of number of pixel values. In conclusion, although no differences in antral characteristics of dominant v. subordinate follicles were detected, mean pixel value and pixel heterogeneity of the follicular wall were higher in dominant follicles, and pixel heterogeneity of follicular wall of ovulatory follicles decreased over days. Quantitative pixel analysis from follicular wall images using UBM could be applied to evaluate follicles in live animals. However, more studies using this technique are necessary. Research was supported by NSERC and CIHR. Luiz Pfeifer was supported by CAPES from Ministry of Education, Brazil.

Multi-sensor derivation of regional vegetation fractional cover in Africa

A spatially continuous field of landscape fractional covers of tree, grass and bare soil is required at regional and continental scales for earth system modeling and environmental monitoring. Climate and its variability drive vegetation fractional cover over time and space. For savanna ecosystems, precipitation plays the main role in shaping vegetation composition. In this study, we estimate land cover fraction at a satellite pixel scale by employing an existing ‘Mean-Sensitivity Unmixing Algorithm’ (MSUA), which is based on a state space defined by two key variables: (1) mean pixel values (referring to mean vegetation states), and (2) inter-annual sensitivity of pixel values to precipitation (referring to vegetation sensitivity to precipitation). We define these two variables through a multi-sensor assessment of three vegetation remote sensing datasets, namely (i) Normalized Difference Vegetation Index (NDVI), based on the visible and near-infrared bands from the Advanced Very High Resolution Radiometer (AVHRR); (ii) backscatter coefficients (dB) from the NASA QuikSCAT active-microwave scatterometer; and (iii) Vegetation Optical Depth (VOD) based on NASA Advanced Microwave Scanning Radiometer on EOS (AMSR-E) passive-microwave radiometry measurements. A merged satellite-gauge precipitation dataset from the Tropical Rainfall Measuring Mission (TRMM) version 3B42V6 is used. The three remote sensing datasets show generally similar but distinctive performances in characterizing the two key variables over various land cover types. NDVI and VOD perform better than dB in characterizing land cover variation based on mean pixel values; while dB represents more reliable and robust vegetation sensitivity to precipitation. By using NDVI for mean vegetation states and dB for inter-annual variability of vegetation to precipitation, we develop an improved fractional cover product. We find that our product agrees well with the tree fraction derived from high-resolution images for natural vegetation regions, and can reproduce the distinctive land cover pattern of grass and bare soil in the Moderate Resolution Imaging Spectroradiometer (MODIS) land cover product. For cropland-mixed regions, our tree fraction is overestimated since human impacts (e.g. irrigation) have not been accounted for in the MSUA. The improved performance from our approach is achieved by the synergistic use of the three vegetation remote sensing datasets, and their physical interpretations have been discussed to support the validity of this approach.

TH‐A‐213CD‐10: A Preliminary Evaluation of An Industrial X‐Ray Backscatter System for Medical Applications

Purpose: To evaluate the advantages of Compton backscatter radiography system as a medical imaging tool by comparing images obtained to traditional transmission radiography methods. Methods: Five phantoms were scanned with a scatter x‐ray imaging (SXI) system and state of the art transmission imaging systems. Four phantoms contained biological tissue equivalent samples used to exemplify objects found in medical imaging scenarios. For a complex biological setting a euthanized 60 lb pig was scanned. Parameters such as the x‐ray tube voltage, x‐ray tube current, pixel integration time, stand‐off distance, detector type and detector collimation were varied. Images acquired with the scatter and transmission modalities were compared by observing the differences between the mean pixel value and standard deviation of regions within the images. Visual comparisons between images were also used. Results: The contrast differences between SXI and transmission images varied based on phantom composition and placement of objects within them. As a result, neither system showed better overall object contrast across all phantom scans. SXI images of bone at 1.6 cm depth and 7.8 cm depth revealed a 30.3% change in contrast while transmission images showed a 0.1% difference in contrast. Standard deviation of the mean pixel values (noise) in the SXI images ranged from 13.7% – 21.1% whereas the standard deviation for the transmission images ranged from 3.1% to 6.6%. Detector collimation prevented the detection of surface scatter and better represented objects deeper within a phantom. Conclusions: Clinical applicability of an industrial SXI system for medical applications was demonstrated. Exposures with the SXI system required access from only one side of a patient, giving it a smaller form factor and eliminating the spatial constraints imposed by traditional transmission x‐ray imaging systems. Additionally, depth information was observed in SXI images, potentially reducing the number views necessary to properly diagnose patients.

SU-E-T-96: Energy Dependence of the New GafChromic- EBT3 Film's Dose Response-Curve.

To study and compare the dose response curves of the new GafChromic EBT3 film for megavoltage and kilovoltage x-ray beams, with different spatial resolution. Two sets of EBT3 films (lot#A101711-02) were exposed to each x-ray beam (6MV, 15MV and 50kV) at 8 dose values (50-3200cGy). The megavoltage beams were calibrated per AAPM TG-51 protocol while the kilovoltage beam was calibrated following the TG-61 using an ionization chamber calibrated at NIST. Each film piece was scanned three consecutive times in the center of Epson 10000XL flatbed scanner in transmission mode, landscape orientation, 48-bit color at two separate spatial resolutions of 75 and 300 dpi. The data were analyzed using ImageJ and, for each scanned image, a region of interest (ROI) of 2×2cm2 at the field center was selected to obtain the mean pixel value with its standard deviation in the ROI. For each energy, dose value and spatial resolution, the average netOD and its associated uncertainty were determined. The Student's t-test was performed to evaluate the statistical differences between the netOD/dose values of the three energy modalities, with different color channels and spatial resolutions. The dose response curves for the three energy modalities were compared in three color channels with 75 and 300dpi. Weak energy dependence was found. For doses above 100cGy, no statistical differences were observed between 6 and 15MV beams, regardless of spatial resolution. However, statistical differences were observed between 50kV and the megavoltage beams. The degree of energy dependence (from MV to 50kV) was found to be function of color channel, dose level and spatial resolution. The dose response curves for GafChromic EBT3 films were found to be weakly dependent on the energy of the photon beams from 6MV to 50kV. The degree of energy dependence varies with color channel, dose and spatial resolution. GafChromic EBT3 films were supplied by Ashland Corp. This work was partially supported by DGAPA-UNAM grant IN102610 and Conacyt Mexico grant 127409.