A computer simulation study comparing lesion detection accuracy with digital mammography, breast tomosynthesis, and cone-beam CT breast imaging

Although conventional mammography is currently the best modality to detect early breast cancer, it is limited in that the recorded image represents the superposition of a three-dimensional (3D) object onto a 2D plane. Recently, two promising approaches for 3D volumetric breast imaging have been proposed, breast tomosynthesis (BT) and CT breast imaging (CTBI). To investigate possible improvements in lesion detection accuracy with either breast tomosynthesis or CT breast imaging as compared to digital mammography (DM), a computer simulation study was conducted using simulated lesions embedded into a structured 3D breast model. The computer simulation realistically modeled x-ray transport through a breast model, as well as the signal and noise propagation through a CsI based flat-panel imager. Polyenergetic x-ray spectra of Mo/Mo 28 kVp for digital mammography, Mo/Rh 28 kVp for BT, and W/Ce 50 kVp for CTBI were modeled. For the CTBI simulation, the intensity of the x-ray spectra for each projection view was determined so as to provide a total average glandular dose of 4 mGy, which is approximately equivalent to that given in conventional two-view screening mammography. The same total dose was modeled for both the DM and BT simulations. Irregular lesions were simulated by using a stochastic growth algorithm providing lesions with an effective diameter of 5 mm. Breast tissue was simulated by generating an ensemble of backgrounds with a power law spectrum, with the composition of 50% fibroglandular and 50% adipose tissue. To evaluate lesion detection accuracy, a receiver operating characteristic (ROC) study was performed with five observers reading an ensemble of images for each case. The average area under the ROC curves (Az) was 0.76 for DM, 0.93 for BT, and 0.94 for CTBI. Results indicated that for the same dose, a 5 mm lesion embedded in a structured breast phantom was detected by the two volumetric breast imaging systems, BT and CTBI, with statistically significant higher confidence than with planar digital mammography, while the difference in lesion detection between BT and CTBI was not statistically significant.     

An assessment of an automated fluorimetric blood phenylalanine technique for phenylketonuria screening and for accurate estimations

An assessment of the automated fluorimetric technique for the estimation of phenylalanine, using blood collected and stored on filter paper, indicated that its accuracy and precision compared favourably with those of other methods, including the bacteriological inhibition assay. The method appeared to offer advantages, both in the detection of phenylketonuria and for the more accurate determinations required in the further investigation and control of known cases.     

Design and evaluation of an external filter technique for exposure equalization in mammography.

We are developing an external filter method for equalizing x-ray exposure in the peripheral region of the breast. This method requires the use of only a limited number of custom-built filters for different breast shapes in a given view. This paper describes the design methodology for these external filters. The filter effectiveness was evaluated through a simulation study on 171 mediolateral and 196 craniocaudal view digitized mammograms and through imaging of a breast phantom. The degree of match between the simulated filter and the individual 3-D exposure profiles at the breast periphery was quantified. An analysis was performed to investigate the effect of filter misalignment. The simulation study indicates that the filter is effective in equalizing exposures for more than 80% of the breast images in our database. The tolerance in filter misalignment was estimated to be about +/- 2 mm for the CC view and +/- 1 mm for the MLO view at the image plane. Some misalignment artifacts were demonstrated with simulated filtered mammograms.     

An automated technique for identifying associations between medications, laboratory results and problems

BackgroundThe patient problem list is an important component of clinical medicine. The problem list enables decision support and quality measurement, and evidence suggests that patients with accurate and complete problem lists may have better outcomes. However, the problem list is often incomplete.ObjectiveTo determine whether association rule mining, a data mining technique, has utility for identifying associations between medications, laboratory results and problems. Such associations may be useful for identifying probable gaps in the problem list.DesignAssociation rule mining was performed on structured electronic health record data for a sample of 100,000 patients receiving care at the Brigham and Women’s Hospital, Boston, MA. The dataset included 272,749 coded problems, 442,658 medications and 11,801,068 laboratory results.MeasurementsCandidate medication-problem and laboratory-problem associations were generated using support, confidence, chi square, interest, and conviction statistics. High-scoring candidate pairs were compared to a gold standard: the Lexi-Comp drug reference database for medications and Mosby’s Diagnostic and Laboratory Test Reference for laboratory results.ResultsWe were able to successfully identify a large number of clinically accurate associations. A high proportion of high-scoring associations were adjudged clinically accurate when evaluated against the gold standard (89.2% for medications with the best-performing statistic, chi square, and 55.6% for laboratory results using interest).ConclusionAssociation rule mining appears to be a useful tool for identifying clinically accurate associations between medications, laboratory results and problems and has several important advantages over alternative knowledge-based approaches.     

Validation and implementation of an algorithm for reporting the automated absolute neutrophil count from selected flagged specimens

Samples that are flagged by automated cell counters and, therefore, require a time-consuming microscopic review cause unacceptable wait times for patients in hematology/oncology clinics. We used a set of 518 samples to validate that 5 flags on Sysmex XE/XT instruments (Sysmex, Kobe, Japan) could safely be ignored when the absolute neutrophil count (ANC) was the primary clinical question. The R(2) between automated and manual ANCs was 96.9% for samples triggering non-ANC flags, with 1 clinically significant discrepant sample out of 296 (0.3%). A new test code allowing clinicians to specify "ANC-only" was implemented, and these non-ANC specific flags were disregarded. The new order set was used in 16.3% of patients. Automated reporting of the ANC in selected flagged samples reduced the review rate by 60% and decreased the turnaround time by 100 minutes. This approach to automatically report the ANC in selected flagged specimens in which the ANC is the primary clinical interest safely decreases the turnaround time for many ANC samples.     

An automated iterative algorithm for water and fat decomposition in three-point Dixon magnetic resonance imaging

An iterative, outlier exclusion, second-order surface fitting algorithm has been developed to solve the well-known phase wraparound problem associated with in vivo applications of the three-point Dixon magnetic resonance imaging method. The technique was optimized for speed by reducing the problem to a pair of planar fits. The spatial misalignment between water and fat components due to the chemical shift was handled on a subpixel level by invoking the shift theorem of Fourier transformation. From the chemical shift corrected water and fat images, high quality recombined MR images were generated. The algorithm was validated in both phantom and patient studies. In vivo breast images and pelvic images are provided as a demonstration of the method.     

An investigation of interpretive bias in insomnia: an analog study comparing normal and poor sleepers

STUDY OBJECTIVES: Cognitive theories state that psychological disorders are associated with, and are possibly maintained by, interpretive biases, which are tendencies to make threatening interpretations of ambiguous stimuli. Recent models of insomnia have highlighted the importance of cognitive processes. The aim of this study was to empirically evaluate whether an interpretive bias is present in poor sleepers. DESIGN: A mixed-design analysis of covariance was employed with group (normal sleepers vs poor sleepers) as a between-subjects variable and sentence type (insomnia-related vs anxiety related) as a within-subjects variable. The dependent variables were the extent to which participants interpreted insomnia-related and anxiety-related sentences as having a threatening meaning. Sleepiness was used as a covariate. SETTING: Treatment and research clinic at a university department of psychiatry. PARTICIPANTS: Forty-one normal and 34 poor sleepers. MEASUREMENTS AND RESULTS: A set of ambiguous scenarios were administered to participants who gave open-ended and forced-choice interpretations of the scenarios. Each scenario could be interpreted in a threat (insomnia or anxiety)-related or neutral manner. Even after controlling for sleepiness, poor sleepers were found to make significantly more threat-related interpretations of ambiguous scenarios than did normal sleepers. CONCLUSIONS: These findings suggest that there is a bias toward threat-related interpretations among poor sleepers and that the exploration of biased interpretations may be an important avenue for future research among individuals who meet full diagnostic criteria for insomnia.     

Tomographic bioluminescence imaging by use of a combined optical-PET (OPET) system: a computer simulation feasibility study

The feasibility and limits in performing tomographic bioluminescence imaging with a combined optical-PET (OPET) system were explored by simulating its image formation process. A micro-MRI based virtual mouse phantom was assigned appropriate tissue optical properties to each of its segmented internal organs at wavelengths spanning the emission spectrum of the firefly luciferase at 37 degrees C. The TOAST finite-element code was employed to simulate the diffuse transport of photons emitted from bioluminescence sources in the mouse. OPET measurements were simulated for single-point, two-point and distributed bioluminescence sources located in different organs such as the liver, the kidneys and the gut. An expectation maximization code was employed to recover the intensity and location of these simulated sources. It was found that spectrally resolved measurements were necessary in order to perform tomographic bioluminescence imaging. The true location of emission sources could be recovered if the mouse background optical properties were known a priori. The assumption of a homogeneous optical property background proved inadequate for describing photon transport in optically heterogeneous tissues and led to inaccurate source localization in the reconstructed images. The simulation results pointed out specific methodological challenges that need to be addressed before a practical implementation of OPET-based bioluminescence tomography is achieved.     

Optimization of technique factors for a silicon diode array full-field digital mammography system and comparison to screen-film mammography with matched average glandular dose

Contrast-detail experiments were performed to optimize technique factors for the detection of low-contrast lesions using a silicon diode array full-field digital mammography (FFDM) system under the conditions of a matched average glandular dose (AGD) for different techniques. Optimization was performed for compressed breast thickness from 2 to 8 cm. FFDM results were compared to screen-film mammography (SFM) at each breast thickness. Four contrast-detail (CD) images were acquired on a SFM unit with optimal techniques at 2, 4, 6, and 8 cm breast thicknesses. The AGD for each breast thickness was calculated based on half-value layer (HVL) and entrance exposure measurements on the SFM unit. A computer algorithm was developed and used to determine FFDM beam current (mAs) that matched AGD between FFDM and SFM at each thickness, while varying target, filter, and peak kilovoltage (kVp) across the full range available for the FFDM unit. CD images were then acquired on FFDM for kVp values from 23-35 for a molybdenum-molybdenum (Mo-Mo), 23-40 for a molybdenum-rhodium (Mo-Rh), and 25-49 for a rhodium-rhodium (Rh-Rh) target-filter under the constraint of matching the AGD from screen-film for each breast thickness (2, 4, 6, and 8 cm). CD images were scored independently for SFM and each FFDM technique by six readers. CD scores were analyzed to assess trends as a function of target-filter and kVp and were compared to SFM at each breast thickness. For 2 cm thick breasts, optimal FFDM CD scores occurred at the lowest possible kVp setting for each target-filter, with significant decreases in FFDM CD scores as kVp was increased under the constraint of matched AGD. For 2 cm breasts, optimal FFDM CD scores were not significantly different from SFM CD scores. For 4-8 cm breasts, optimum FFDM CD scores were superior to SFM CD scores. For 4 cm breasts, FFDM CD scores decreased as kVp increased for each target-filter combination. For 6 cm breasts, CD scores decreased slightly as kVp increased for Mo-Mo, but did not change significantly as a function of kVp for either Mo-Rh or Rh-Rh. For 8 cm breasts, Rh/Rh FFDM CD scores were superior to other target-filter combinations and increased significantly as kVp increased. These results indicate that low-contrast lesion detection was optimized for FFDM by using a softer x-ray beam for thin breasts and a harder x-ray beam for thick breasts, when AGD was kept constant for a given breast thickness. Under this constraint, optimum low-contrast lesion detection with FFDM was superior to that for SFM for all but the thinnest breasts.     

Rapid identification of fungi by sequencing the ITS1 and ITS2 regions using an automated capillary electrophoresis system.

We developed a standardized DNA sequence-based approach for the accurate and timely identification of medically important fungi by sequencing polymerase chain reaction (PCR) products with a rapid automated capillary electrophoresis system. A simple DNA extraction method and PCR amplification using universal fungal primers was used to amplify ribosomal DNA from a range of clinical isolates and reference strains. The entire internal transcribed spacer (ITS) 1-5.8S-ITS2 ribosomal DNA region was sequenced using automated dye termination sequencing for 89 clinical isolates. These had previously been identified by traditional methods and included 12 ascomycetous yeast species, three basidiomycetous yeast species, eight dermatophyte species and two thermally dimorphic fungi, Scedosporium prolificans and S. apiospermum. Furthermore, 21 reference strains representing 19 different Candida species, Geotrichum candidum and Malassezia furfur were also sequenced as part of this study and were used either as standards for sequence-based comparisons, or as assay controls. Sequence-based identification was compared to traditional identification in a blinded manner. Of the clinical isolates tested, 88/89 had DNA sequences that were highly homologous to those of reference strains accessioned in GenBank, and 87/89 gave a sequence-based identification result that correlated with the traditional identification. In contrast to relatively slow conventional methods of identification, a sequence-based identification from a pure culture can be obtained within 24 h of a DNA extraction carried out after a minimal period of culture growth. We conclude that this approach is rapid, and may be a more accurate cost-effective alternative than most phenotypic methods for identification of many medically important fungi frequently encountered in a routine diagnostic microbiology laboratory.     

An evaluation of statistical procedures for comparing an individual's performance with that of a group of controls

The single case methodology that is widely used in cognitive neuropsychology often requires a comparison of data from a single individual (the patient) with that from a group of controls, in order to ascertain whether the patient's mean score can be viewed as significantly different from that of controls. This article reviews methods that have been used to deal with such data. Although Analysis of Variance (ANOVA) provides one possible solution of comparing group means, unequal group sizes and differences in variability between patient and controls can violate the assumptions of the test. Using Monte Carlo simulations, it was found that differences in group size and a group of N = 1 did not significantly affect the reliability of the analysis. In contrast, unacceptably high Type I errors were obtained when, in addition to unequal group sizes, there were relatively modest differences between the variance of the patient and that of the controls. We suggest that ANOVA can be used for the comparison of the mean score of an individual with that of a group of controls, but that when there is a difference in variability between the two groups, revised F criteria should be used in order to make the analysis reliable. A table of modified F values is given, which can be used for various departures from homogeneity of variance.     

An evaluation of statistical procedures for comparing an individual's performance with that of a group of controls

The single case methodology that is widely used in cognitive neuropsychology often requires a comparison of data from a single individual (the patient) with that from a group of controls, in order to ascertain whether the patient's mean score can be viewed as significantly different from that of controls. This article reviews methods that have been used to deal with such data. Although Analysis of Variance (ANOVA) provides one possible solution of comparing group means, unequal group sizes and differences in variability between patient and controls can violate the assumptions of the test. Using Monte Carlo simulations, it was found that differences in group size and a group of N = 1 did not significantly affect the reliability of the analysis. In contrast, unacceptably high Type I errors were obtained when, in addition to unequal group sizes, there were relatively modest differences between the variance of the patient and that of the controls. We suggest that ANOVA can be used for the comparison of the mean score of an individual with that of a group of controls, but that when there is a difference in variability between the two groups, revised F criteria should be used in order to make the analysis reliable. A table of modified F values is given, which can be used for various departures from homogeneity of variance.     

Update of a selected technique for uric acid determination in plasma and serum. Experimental study and value of derivative spectrophotometry

A critical study of the candidate reference method for evaluation of uric acid in plasma proposed by the American Association of Clinical Chemistry is followed by testing in six laboratories. The dispersion of results is wide (CV greater than 5%). The importance of turbidity remaining after the deproteinization by trichloracetic acid is clearly demonstrated. This turbidity is really not reproducible from an operation to another one on the same serum. It is very likely responsible of the great dispersion of the results. After that, other deproteinization methods are tried. Ultrafiltration and ultracentrifugation give both defect errors because uric acid is in part bound to proteins. So it is necessary to find another technic which cancels the effects of turbidity on the absorbance readings in the ultra-violet domain. Experimental studies showed that uric acid may be evaluated with a good accuracy by derivative spectrophotometry, in lipid solutions as well as in cloudy ones. Turbidity was created by intralipid suspension additions. Various parameters hitting the method were examined (linearity, smoothing window...), taking as criterion the measure of overloaded serums at different levels. At last, the method is successfully transferred in several sites. In the case of blood serum, the respective influences of derivation and of repetition of final centrifugations are studied in order to estimate the effect of remaining turbidity. By the use of derivative spectrophotometry the improvement of the method of evaluation of uric acid proposed by A.A.C.C. is very noticeable; it reduces the variation coefficient between sites to less than 2%.     

A multisite performance study comparing the reading of immunohistochemical slides on a computer monitor with conventional manual microscopy for estrogen and progesterone receptor analysis

A multisite study was conducted to assess the performance of the Aperio digital pathology system (Aperio Technologies, Vista, CA) for reading estrogen receptor (ER) and progesterone receptor (PR) slides on a computer monitor. A total of 520 formalin-fixed breast tissue specimens were assayed at 3 clinical sites for ER and PR (260 each). Percentage and average staining intensity of positive nuclei were assessed. At each site, 3 pathologists performed a blinded reading of the glass slides using their microscopes initially and later using digital images on a computer monitor. Comparable percentages of agreements were obtained for manual microscopy (MM) and manual digital slide reading (MDR) (ER, percentage of positive nuclei with cutoffs: MM, 91.3%-99.0%/MDR, 91.3%-100.0%; PR, percentage of positive nuclei with cutoffs: MM, 83.8%-99.0%/MDR, 76.3%-100.0%). Reading ER and PR slides on a computer monitor using the Aperio digital pathology system is equivalent to reading the slides with a conventional light microscope.     

Iterative interim techniques for reduced costs and a better mammography workstation: An opinion

The promise of digital mammography has focused considerable resources on the challenges of mammography workstations, but the risk of wasted time and resources in such efforts is very high. Although final testing of the workstation's image quality and ergonomics is common, a number of interim evaluation and refinement techniques can be applied throughout the design and development process. The use of such techniques holds potential not only to save time and money but also to produce a superior workstation.     

A validation study of a dedicated software for an automated in vivo dosimetry control in radiotherapy

In vivo dosimetry (IVD) is the last step of a radiotherapy quality control program aimed to ensure that the dose delivered is in agreement with that prescribed. IVD procedures based on single detectors are time-consuming and impossible to use for the modern radiotherapy techniques, based on static or kinetic beams (modulated in intensity fluence); this means that more efficient and practical methods are highly recommended. The practical method SOFTDISO, based on the use of electronic portal image device (EPID), provides two tests (i) the R ratio between the reconstructed and the planned isocenter doses to verify an agreement within 5% and (ii) the γ-analysis of the EPID images, to verify γ%?≥?90% and γ_mean ≤?0.4. This paper reports the results of 11,357 IVD tests carried out for 823 patients treated by three-dimensional conformal radiation therapy and volumetric modulated arc therapy techniques. In particular, the dose disagreements are reported distinguishing two kinds of causes, those of (i) class 1 that includes the errors due to inadequate quality controls and (ii) the class 2, due to patient morphological changes. About the tests out of tolerance, 6% were by VMAT and 21% by 3DCRT, but taking into account the only class 1 of errors, i.e., removing the causes of class 2, only 7% of patients examined presented at least one of the three mean indexes out of tolerance. The workload for IVD on 9 patients/day per linac is about 52 min/day but recently, a new automated SOFTDISO version has been implemented to reduce the time to about 34 min/day.