Biomedical subjects

C E Floyd

Publications and source records attributed to C E Floyd.

At least 19 recordsLinked to original sources

PubMed1997

Fractal texture analysis in computer-aided diagnosis of solitary pulmonary nodules.

RATIONALE AND OBJECTIVES: The authors investigated the use of fractal texture characterization to improve the accuracy of solitary pulmonary nodule computer-aided diagnosis (CAD) systems. METHODS: Thirty chest radiographs were acquired from patients who had no pulmonary nodules. Thirty regions were selected that were considered remotely suspicious-looking for nodules. Artificial nodules of multiple shapes, sizes, and orientations were added at subtle levels of contrast to 30 non-suspicious-looking regions of the radiographs. Fractal dimensions of the 60 "nodule candidates" were calculated to quantify the texture of each region. Four radiologists also interpreted the images. RESULTS: The fractal dimension of each possible nodule provided statistically significant (P < .05) differentiation between regions that contained an artificial nodule and those that did not. The area under the receiver operating characteristic curve for the fractal analysis was significantly better (P < .05) than that for the radiologists. CONCLUSION: Fractal texture characterization provides useful information for the classification of potential solitary pulmonary nodules with CAD algorithms.

Diagnosis, Computer-Assisted↗

PubMed1997

Improved Bayesian image estimation for digital chest radiography.

PURPOSE: Previously, we have shown that Spatially Varying Bayesian Image Estimation (SVBIE) can be used to reduce scatter and improve contrast-to-noise ratios (CNR) in digital chest radiographs with no degradation of image resolution. This previous algorithm used a model for scatter compensation that was derived for emission tomography. Here, we develop and evaluate a new iterative SVBIE technique that incorporates a scatter model derived for projection radiography. MATERIALS AND METHODS: Portable digital radiographs of an anthropomorphic chest phantom were obtained along with quantitative scatter measurements using a calibrated photostimulable phosphor system. The new iterative SVBIE technique was applied to the phantom image to reduce scatter. Scatter fraction reduction, CNR improvement, and resolution degradation were evaluated. RESULTS: Residual scatter fractions were reduced to less than 2% in the lungs and 30% in the mediastinum at 14 iterations. CNR was improved by approximately 50% in the lung region and 187% in the mediastinum. Resolution was not degraded. CONCLUSIONS: The new SVBIE technique can reduce scatter to levels far below those provided by an antiscatter grid and can increase CNR without loss of resolution. The new technique outperforms the previous Bayesian techniques.

Bayes Theorem↗

PubMed1997

Quality control phantom for digital chest radiography.

PURPOSE: To develop and test a chest phantom for routine quality control testing of digital radiography systems. MATERIALS AND METHODS: The phantom was constructed from sheets of copper, aluminum, and acrylic, which were cut and arranged to yield a radiographic projection resembling that of a human thorax. Regional test objects allowed quantitative assessment of optical density, contrast detail, and spatial resolution. Validation tests were performed to assess image stability in a stable imaging environment and sensitivity to changes in image quality when they occur. RESULTS: The phantom yielded consistent pseudoclinical images when used in a routine quality control program and facilitated detection of simulated problems that were induced in imaging system performance. CONCLUSION: The chest phantom enables quantitative, full-system testing of digital radiography system as they are used clinically for chest radiography.

Humans↗

PubMed1997

Predicting breast cancer invasion with artificial neural networks on the basis of mammographic features.

PURPOSE: To evaluate whether an artificial neural network (ANN) can predict breast cancer invasion on the basis of readily available medical findings (ie, mammographic findings classified according to the American College of Radiology Breast Imaging Reporting and Data System and patient age). MATERIALS AND METHODS: In 254 adult patients, 266 lesions that had been sampled at biopsy were randomly selected for the study. There were 96 malignant and 170 benign lesions. On the basis of nine mammographic findings and patient age, a three-layer backpropagation network was developed to predict whether the malignant lesions were in situ or invasive. RESULTS: The ANN predicted invasion among malignant lesions with an area under the receiver operating characteristic curve (Az) of .91 +/- .03. It correctly identified all 28 in situ cancers (specificity, 100%) and 48 of 68 invasive cancers (sensitivity, 71%). CONCLUSION: The ANN used mammographic features and patient age to accurately classify invasion among breast cancers, information that was previously available only by means of biopsy. This knowledge may assist in surgical planning and may help reduce the cost and morbidity of unnecessary biopsy.

Biopsy↗

PubMed1997

Memory artifact related to selenium-based digital radiography systems.

Digital images acquired on radiography systems with amorphous selenium detectors are susceptible to "memory artifacts" from prior x-ray exposures. In routine clinical use and in a laboratory experiment, artifacts appeared in chest radiographs until the selenium recovered from initial exposure. Memory artifacts were eliminated when 3 minutes or more elapsed between acquisition of a lateral chest radiograph and acquisition of the next radiograph.

Aluminum↗

PubMed1996

Spatially varying Bayesian image estimation.

RATIONALE AND OBJECTIVES: Second-order neighborhoods and a spatially varying prior were incorporated into Bayesian image estimation (BIE) to improve image contrast-to-noise ratios (CNRs) while preserving image resolution. METHODS: Second-order neighborhoods were incorporated into the BIE algorithm. A spatially varying BIE (SVBIE) algorithm was developed by incorporating a spatially varying prior. The two algorithms were used to process an anthropomorphic chest phantom image. CNRs, resolution, and image appearance were evaluated. RESULTS: The use of second-order neighborhoods alone improved the CNR in the mediastinum and degraded the resolution. SVBIE demonstrated no degradation of resolution. In the lung region, SVBIE enhanced the CNR but did not perform as well as BIE. In the mediastinum, the SVBIE technique outperformed the older technique and provided a dramatic increase in the CNR over the original image. CONCLUSION: The SVBIE technique provides improved image CNR with no loss of resolution.

Algorithms↗

PubMed1996

Improved noninvasive diagnosis of acute pulmonary embolism with optimally selected clinical and chest radiographic findings.

RATIONALE AND OBJECTIVES: The authors improved the noninvasive diagnosis of acute pulmonary embolism (PE) by studying the clinical and chest radiographic findings of patients suspected of having PE and correlating those findings with the physicians' clinical impression. METHODS: A stepwise linear discriminant algorithm was developed on the basis of 1,064 patients from the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) study to select clinical and chest radiographic findings with the highest diagnostic power in patients suspected of having PE. Subsequently, a linear classifier and a nonlinear artificial neural network were developed to help diagnose PE on the basis of the reduced number of findings. RESULTS: Both classifiers produced a statistically significant improvement (Az = 0.77 +/- 0.02) in the clinical performance of the PIOPED physicians (Az = 0.72 +/- 0.02). Results are also presented separately for groups of patients classified on the basis of the difficulty level of their ventilation-perfusion lung scans. CONCLUSION: Two computer-aided diagnostic tools were developed to assist physicians in the assessment of the pretest likelihood of PE by using an optimally reduced number of findings.

Acute Disease↗

PubMed1996

Artificial neural network: improving the quality of breast biopsy recommendations.

PURPOSE: To evaluate the performance and inter- and intraobserver variability of an artificial neural network (ANN) for predicting breast biopsy outcome. MATERIALS AND METHODS: Five radiologists described 60 mammographically detected lesions with the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) nomenclature. A previously programmed ANN used the BI-RADS descriptors and patient histories to predict biopsy results. ANN predictive performance was compared with the clinical decision to perform biopsy. Inter- and intraobserver variability of radiologists' interpretations and ANN predictions were evaluated with Cohen kappa analysis. RESULTS: The ANN maintained 100% sensitivity (23 of 23 cancers) while improving the positive predictive value of biopsy results from 38% (23 of 60 lesions) to between 58% (23 of 40 lesions) and 66% (23 of 35 lesions; P < .001). Interobserver variability for interpretation of the lesions was significantly reduced by the ANN (P < .001); there was no statistically significant effect on nearly perfect intraobserver reproducibility. CONCLUSION: Use of an ANN with radiologists' descriptions of abnormal findings may improve interpretation of mammographic abnormalities.

Biopsy↗

PubMed1996

Breast imaging reporting and data system standardized mammography lexicon: observer variability in lesion description.

OBJECTIVE: The American College of Radiology has recommended the Breast Imaging Reporting and Data System (BI-RADS) as a standardized scheme for describing mammographic lesions. The objective of this study was to measure inter- and intraobserver variabilities of radiologists' descriptions of mammographic lesions with the BI-RADS standardized lexicon. MATERIALS AND METHODS: Sixty mammographic studies with abnormal findings were independently evaluated by five radiologists. Readers described each lesion by selecting a single term from the BI-RADS lexicon for each of eight morphologic categories: calcification distribution, number, and description; mass margin, shape, and density; associated findings; and special cases. Additionally, each reader assessed the significance of each lesion on a five-point scale. One observer read each case twice. Inter- and intraobserver variabilities for each description and interpretation category of the BI-RADS lexicon were determined with Cohen's kappa statistic. Radiologists' specific use of calcification descriptors was evaluated in detail. RESULTS: Substantial agreement was observed between readers for choosing terms to describe masses and calcifications (kappa value range, 0.50 +/- 0.02-0.77 +/- 0.03). Intraobserver agreement for these categories was similar (kappa value range, 0.57 +/- 0.07-0.84 +/- 0.09). Considerable inter- and intraobserver variabilities were noted for the "associated findings" and "special cases" categories (kappa value range, -0.02 +/- 0.14-0.38 +/- 0.12), a result that in part reflected the small number of cases to which these categories were assigned. Moderate interobserver variability and little intraobserver variability in the interpretation of lesion significance were noted when an assessment classification similar to that of BI-RADS was used. Use of terms to describe calcifications did not always conform to BI-RADS-defined levels of suspicion. CONCLUSION: BI-RADS is moderately successful in providing a standardized language for physicians to describe lesion morphology. Efforts to reevaluate specific terms and the diagnostic significance assigned to calcification descriptors may prove useful in maintaining the promise of improved quality with the BI-RADS standardized mammography lexicon.

Adult↗

PubMed1996

Diffuse nodular lung disease on chest radiographs: a pilot study of characterization by fractal dimension.

OBJECTIVE: We present a computer-aided diagnostic technique for identifying nodular interstitial lung disease on chest radiographs. The fractal dimension was used as a numerical measure of image texture on digital chest radiographs to distinguish patients with normal lung from those with a diffuse nodular interstitial abnormality. MATERIALS AND METHODS: Twenty digitized chest radiographs were classified as normal (n = 10) or as containing diffuse nodular abnormality (n = 10) on the basis of readings assigned according to the classification of the International Labour Organization. Regions of interest (ROIs) measuring 1.28 cm2 were selected from the intercostal spaces of these radiographs. The fractal dimension of these ROIs was estimated by power spectrum analysis. The cases were not subtle. RESULTS: The fractal dimension provided statistically significant discrimination between normal parenchyma and nodular interstitial lung disease. The area under the receiver operating characteristic curve was 0.90 (+/- 0.02). One operating point provides sensitivity of 88% with a specificity of 80%. CONCLUSION: The fractal dimension can provide a measure of lung parenchymal texture and shows promise as an element of computer-aided diagnosis, characterization, and follow-up of interstitial lung disease.

Fractals↗

PubMed1995

Lesion size quantification in SPECT using an artificial neural network classification approach.

An artificial neural network (ANN) has been developed to determine the size of lesions detected in single photon emission computed tomographic images. The network is the Learning Vector Quantizer and is trained to perform size quantification based on image neighborhoods extracted around the lesions. The ANN is compared to the optimal, Bayesian algorithm developed to perform the same task using the unreconstructed, projection data. The performance of the neural network is evaluated at two different noise levels. The Bayesian algorithm provides the upper bound for size quantification performance against which the ANN is compared. In the ideal case where the Bayesian algorithm has explicit knowledge of the underlying distributions, its performance is superior to that of the neural network. However, in the more realistic case where the distributions need to be estimated from the same learning sample the ANN was trained on, the two algorithms have comparable performances.

Algorithms↗

PubMed1995

A Monte Carlo investigation of dual-energy-window scatter correction for volume-of-interest quantification in 99Tcm SPECT.

Using Monte Carlo simulation of 99Tcm single-photon-emission computed tomography (SPECT), we investigate the effects of tissue-background activity, tumour location, patient size, uncertainty of energy windows, and definition of tumour region on the accuracy of quantification. The dual-energy-window method of correction for Compton scattering is employed and the multiplier which yields correct activity for the VI as a whole calculated. The model is usually a sphere containing radioactive water located within a cylinder filled with a more dilute solution of radioactivity. Two simulation codes are employed. Reconstruction is by ML-EM algorithm with attenuation compensation. The scatter multiplier depends only slightly on the sphere location or the cylinder diameter. It also depends little on whether correction is before or after reconstruction. At low background level, it changes with VOI size, but not at higher background. For a geometrical VOI, it is 1.25 at zero background, decreases sharply to 0.56 for equal concentrations, and is 0.44 when the background concentration is very large. Quantification is accurate (less than 9% error) if the test background is reasonably close to that used in setting the universal scatter-multiplier value, or if the test backgrounds are always large and so is the universal-value background, but not if the test backgrounds cover a large range of values including zero. Results largely agree with those from experiment after the experimental data with background is re-evaluated with prejudice.

Humans↗

PubMed1995

Bayesian image estimation of digital chest radiography: interdependence of noise, resolution, and scatter fraction.

Previously, it has been shown that Bayesian image estimation (BIE) can reduce the effects of scattered radiation and improve contrast-to-noise ratios (CNR) in digital radiographs of anthropomorphic chest phantoms by improving contrast while constraining noise. Here, the use of BIE as a noise reduction technique is reported. An anthropomorphic phantom was imaged with a previously calibrated photostimulable phosphor system using standard bedside chest radiography protocols. The Bayesian technique was then used to process this image. BIE incorporates a radial exponential convolution scatter model with two adjustable parameters. In previous reports, these parameters were optimized to reduce the residual fraction of scattered radiation in the processed image. Here, the parameters were adjusted to evaluate the potential of BIE to reduce image noise. While the full width at half maximum of the scatter model was held constant, the magnitude was varied. Evaluation was based on residual scatter fractions and CNR. The magnitude of the kernel in the scatter model was varied from 0.0 to 2.5 in steps of 0.5. Previously, it was found that an "ideal" scatter kernel magnitude of 2.33 provided a minimum residual scatter fraction. This magnitude corresponds to the average scatter-to-primary ratio in the chest radiograph. As the magnitude was increased, the residual scatter fraction decreased and the CNR increased in both the lungs and the mediastinum. However, as the magnitude was decreased, the percent noise also decreased; therefore, a lower magnitude kernel reduces noise. By varying the magnitude of the kernel used, differing amounts of noise reduction and contrast enhancement can be obtained.(ABSTRACT TRUNCATED AT 250 WORDS)

Bayes Theorem↗

PubMed1995

Artificial neural network for diagnosis of acute pulmonary embolism: effect of case and observer selection.

PURPOSE: To compare the diagnostic performance of an artificial neural network (ANN) with that of physicians in patients with suspected pulmonary embolism (PE). MATERIALS AND METHODS: An ANN was developed to predict PE by using findings from ventilation-perfusion lung scans and chest radiographs. First, the network was evaluated on 1,064 cases from the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) study that had a definitive angiographic outcome. An upper and lower bound of its diagnostic performance was provided depending on case difficulty. Then, the network was tested on 104 patients with suspected PE in whom pulmonary angiography was essential for diagnosis. The diagnostic performance of the ANN was compared with that of (a) two nuclear medicine physicians who read the scans for the needs of this study and (b) the nuclear medicine physicians who originally read the scans. The effects of case and observer selection on performance were addressed. RESULTS: The ANN outperformed the physicians when they used the PIOPED criteria for categoric assessment, and it performed as well as the two study physicians on the basis of their probability assessments. CONCLUSION: The ANN can detect or exclude PE in a highly selected group of difficult cases with a consistency equivalent to that of very experienced physicians.

Algorithms↗

PubMed1995

Technical evaluation of a digital chest radiography system that uses a selenium detector.

PURPOSE: To evaluate a digital chest radiography system that uses a selenium detector. MATERIALS AND METHODS: The relative amounts of scattered radiation in the images (scatter fractions), the effect of x-ray exposure levels on image appearance, the potential "throughput" in a clinical environment, and the effects of image processing options were evaluated. RESULTS: Scatter fractions in digital images acquired with an antiscatter grid were lower in the lung region and higher in the retrocardiac and central mediastinal regions than in conventional images. Digital images acquired without an antiscatter grid had higher scatter fractions in all areas. Increases in exposure intensity reduced the appearance of noise. A new image could be acquired every 37 seconds, and a "preview image" appeared on the monitor after approximately 23 seconds. Laser-printed images were available after at least 5 minutes; the time required increased when many images were acquired in a short time. CONCLUSION: The selenium-based chest radiography system allows for rapid chest examination and excellent image quality when used with an antiscatter grid.

Evaluation Studies as Topic↗

PubMed1995

Breast cancer: prediction with artificial neural network based on BI-RADS standardized lexicon.

PURPOSE: To determine if an artificial neural network (ANN) to categorize benign and malignant breast lesions can be standardized for use by all radiologists. MATERIALS AND METHODS: An ANN was constructed based on the standardized lexicon of the Breast Imaging Recording and Data System (BI-RADS) of the American College of Radiology. Eighteen inputs to the network included 10 BI-RADS lesion descriptors and eight input values from the patient's medical history. The network was trained and tested on 206 cases (133 benign, 73 malignant cases). Receiver operating characteristic curves for the network and radiologists were compared. RESULTS: At a specified output threshold, the ANN would have improved the positive predictive value (PPV) of biopsy from 35% to 61% with a relative sensitivity of 100%. At a fixed sensitivity of 95%, the specificity of the ANN (62%) was significantly greater than the specificity of radiologists (30%) (P < .01). CONCLUSION: The BI-RADS lexicon provides a standardized language between mammographers and an ANN that can improve the PPV of breast biopsy.

Adult↗

PubMed1995

Selenium-based digital radiography of the chest: radiologists' preference compared with film-screen radiographs.

OBJECTIVE: A new digital thoracic radiography system (Thoravision; Philips Medical Systems, Hamburg, Germany), which uses selenium as a detector material, was evaluated for observer preference. The system has been shown to have higher detection efficiency than conventional film-screen systems and thus could provide an image with reduced noise. The hypothesis tested in this study was that the selenium-based digital system would provide an image appearance for conventional thoracic imaging that would be equal or superior to that provided by a conventional film-screen system. MATERIALS AND METHODS: Fifty-three patient volunteers were imaged at 120 kV with both the selenium-based system and a thoracic film-screen combination system (InSight HC; Kodak, Rochester, NY). Posteroanterior and lateral images were acquired with both systems, for a total of 212 images. Both imaging systems included a stationary 12:1 antiscatter grid. Exposures were the same for both imaging systems, and the digital images were printed to film. Images for the same patient were compared by six observers--three specialized chest radiologists and three general radiologists. Images included both normal chest radiographs and radiographs with abnormal findings. Each pair of images was ranked on a scale from 1 to 5 for preference of technique, with a score of 3 indicating no preference. Eleven anatomic features were evaluated in the posteroanterior views, and six features were evaluated in the lateral views. Statistical significance of preference was evaluated with Student's t test. RESULTS: The chest radiologists had a statistically significant preference for the selenium-based system for all 17 features (p < .001). The general radiologists had a statistically significant preference for the selenium-based system for visualization of 10 of the 17 features (p < .05). Neither group had a statistically significant preference for the conventional images in any category. CONCLUSION: The selenium-based system provided an image appearance that was significantly preferred by all radiologists, more strongly by those specializing in chest radiography. This study demonstrates that a digital thoracic imaging system can routinely produce images that are perceived as equal or superior to conventional images.

Artifacts↗

PubMed1994

Prediction of breast cancer malignancy using an artificial neural network.

BACKGROUND: An artificial neural network (ANN) was developed to predict breast cancer from mammographic findings. This network was evaluated in a retrospective study. METHODS: For a set of patients who were scheduled for biopsy, radiologists interpreted the mammograms and provided data on eight mammographic findings as part of the standard mammographic workup. These findings were encoded as features for an ANN. Results of biopsies were taken as truth in the diagnosis of malignancy. The ANN was trained and evaluated using a jackknife sampling on a set of 260 patient records. Performance of the network was evaluated in terms of sensitivity and specificity over a range of decision thresholds and was expressed as a receiver operating characteristic curve. RESULTS: The ANN performed more accurately than the radiologists (P < 0.08) with a relative sensitivity of 1.0 and specificity of 0.59. CONCLUSIONS: An ANN can be trained to predict malignancy from mammographic findings with a high degree of accuracy.

Algorithms↗