Search

Breast cancer has been known as the most prevalent and common cause of death among Malaysian woman especially over the age of 40. Breast cancer can usually be identified as either benign or malignant with invasive biopsy procedure. The treatment protocol is allocated based on the whether the mass is benign or malignant. Fortunately, breast cancer like many other cancer types are curable and patient survival can be improved, subject to early diagnosis. Radiograph images lies numbers of features that useful for computer aided diagnosis. In this thesis, the work is divided into two main phases; 1) evaluating the reproducibility of radiomics features derived from manual delineation and semiautomatic segmentation after two different contrast enhancement techniques on masses in two-dimensional (2D) mammography images and 2) to implement the Automated Machine Learning (AutoML) in classifying types of mass in mammogram images. With introduction of ML techniques, breast cancer can be diagnosed in early stage without any invasive and risky procedure. The methodology presented in this research consist of several stages including, image acquisition, image segmentation, feature extraction/selection and, classification using AutoML. The first phase determines the reproducibility between Contrast Limited Adaptive Histogram Equalization (CLAHE) and Adaptive Histogram Equalization (AHE) techniques. The semiautomatic segmentation techniques used in the first phase is Active Contour Method (ACM) with 100 iterations. Three types of radiomics features were extracted including first order, second order and shape features. 37 features were extracted from each tumor in three different techniques mentioned: 9 of these were shape-based features, while 28 were texture-based features. Notably the CLAHE group (ICC = 0.890 ± 0.554, p < 0.05) had the highest reproducibility compared to the features extracted from the AHE group (ICC = 0.850 ± 0.933, p < 0.05) and manual delineation (ICC = 0.673 ±

Breast cancer has been known as the most prevalent and common cause of death among Malaysian woman especially over the age of 40. Breast cancer can usually be identified as either benign or malignant with invasive biopsy procedure. The treatment protocol is allocated based on the whether the mass is benign or malignant. Fortunately, breast cancer like many other cancer types are curable and patient survival can be improved, subject to early diagnosis. Radiograph images lies numbers of features that useful for computer aided diagnosis. In this thesis, the work is divided into two main phases; 1) evaluating the reproducibility of radiomics features derived from manual delineation and semiautomatic segmentation after two different contrast enhancement techniques on masses in two-dimensional (2D) mammography images and 2) to implement the Automated Machine Learning (AutoML) in classifying types of mass in mammogram images. With introduction of ML techniques, breast cancer can be diagnosed in early stage without any invasive and risky procedure. The methodology presented in this research consist of several stages including, image acquisition, image segmentation, feature extraction/selection and, classification using AutoML. The first phase determines the reproducibility between Contrast Limited Adaptive Histogram Equalization (CLAHE) and Adaptive Histogram Equalization (AHE) techniques. The semiautomatic segmentation techniques used in the first phase is Active Contour Method (ACM) with 100 iterations. Three types of radiomics features were extracted including first order, second order and shape features. 37 features were extracted from each tumor in three different techniques mentioned: 9 of these were shape-based features, while 28 were texture-based features. Notably the CLAHE group (ICC = 0.890 ± 0.554, p < 0.05) had the highest reproducibility compared to the features extracted from the AHE group (ICC = 0.850 ± 0.933, p < 0.05) and manual delineation (ICC = 0.673 ±

Feature learning refers to techniques that learn to transform raw data input into an effective representation for further higher-level processing in many computer vision tasks. This chapter presents the investigations and the results of feature learning using convolutional neural networks to automatically assess knee osteoarthritis (OA) severity and the associated clinical and diagnostic features of knee OA from radiographs (X-ray images). Also, this chapter demonstrates that feature learning in a supervised manner is more effective than using conventional handcrafted features for automatic detection of knee joints and fine-grained knee OA image classification. In the general machine learning approach to automatically assess knee OA severity, the first step is to localize the region of interest that is to detect and extract the knee joint regions from the radiographs, and the next step is to classify the localized knee joints based on a radiographic classification scheme such as Kellgren and Lawrence grades. First, the existing approaches for detecting (or localizing) the knee joint regions based on handcrafted features are reviewed and outlined in this chapter. Next, three new approaches are introduced: 1) to automatically detect the knee joint region using a fully convolutional network, 2) to automatically assess the radiographic knee OA using CNNs trained from scratch for classification and regression of knee joint images to predict KL grades in ordinal and continuous scales, and 3) to quantify the knee OA severity optimizing a weighted ratio of two loss functions: categorical cross entropy and mean-squared error using multi-objective convolutional learning. The results from these methods show progressive improvement in the overall quantification of the knee OA severity. Two public datasets: the OAI and the MOST are used to evaluate the approaches with promising results that outperform existing approaches. In summary, this work primarily contributes to the field

Feature learning refers to techniques that learn to transform raw data input into an effective representation for further higher-level processing in many computer vision tasks. This chapter presents the investigations and the results of feature learning using convolutional neural networks to automatically assess knee osteoarthritis (OA) severity and the associated clinical and diagnostic features of knee OA from radiographs (X-ray images). Also, this chapter demonstrates that feature learning in a supervised manner is more effective than using conventional handcrafted features for automatic detection of knee joints and fine-grained knee OA image classification. In the general machine learning approach to automatically assess knee OA severity, the first step is to localize the region of interest that is to detect and extract the knee joint regions from the radiographs, and the next step is to classify the localized knee joints based on a radiographic classification scheme such as Kellgren and Lawrence grades. First, the existing approaches for detecting (or localizing) the knee joint regions based on handcrafted features are reviewed and outlined in this chapter. Next, three new approaches are introduced: 1) to automatically detect the knee joint region using a fully convolutional network, 2) to automatically assess the radiographic knee OA using CNNs trained from scratch for classification and regression of knee joint images to predict KL grades in ordinal and continuous scales, and 3) to quantify the knee OA severity optimizing a weighted ratio of two loss functions: categorical cross entropy and mean-squared error using multi-objective convolutional learning. The results from these methods show progressive improvement in the overall quantification of the knee OA severity. Two public datasets: the OAI and the MOST are used to evaluate the approaches with promising results that outperform existing approaches. In summary, this work primarily contributes to the field

Feature learning refers to techniques that learn to transform raw data input into an effective representation for further higher-level processing in many computer vision tasks. This chapter presents the investigations and the results of feature learning using convolutional neural networks to automatically assess knee osteoarthritis (OA) severity and the associated clinical and diagnostic features of knee OA from radiographs (X-ray images). Also, this chapter demonstrates that feature learning in a supervised manner is more effective than using conventional handcrafted features for automatic detection of knee joints and fine-grained knee OA image classification. In the general machine learning approach to automatically assess knee OA severity, the first step is to localize the region of interest that is to detect and extract the knee joint regions from the radiographs, and the next step is to classify the localized knee joints based on a radiographic classification scheme such as Kellgren and Lawrence grades. First, the existing approaches for detecting (or localizing) the knee joint regions based on handcrafted features are reviewed and outlined in this chapter. Next, three new approaches are introduced: 1) to automatically detect the knee joint region using a fully convolutional network, 2) to automatically assess the radiographic knee OA using CNNs trained from scratch for classification and regression of knee joint images to predict KL grades in ordinal and continuous scales, and 3) to quantify the knee OA severity optimizing a weighted ratio of two loss functions: categorical cross entropy and mean-squared error using multi-objective convolutional learning. The results from these methods show progressive improvement in the overall quantification of the knee OA severity. Two public datasets: the OAI and the MOST are used to evaluate the approaches with promising results that outperform existing approaches. In summary, this work primarily contributes to the field

Feature learning refers to techniques that learn to transform raw data input into an effective representation for further higher-level processing in many computer vision tasks. This chapter presents the investigations and the results of feature learning using convolutional neural networks to automatically assess knee osteoarthritis (OA) severity and the associated clinical and diagnostic features of knee OA from radiographs (X-ray images). Also, this chapter demonstrates that feature learning in a supervised manner is more effective than using conventional handcrafted features for automatic detection of knee joints and fine-grained knee OA image classification. In the general machine learning approach to automatically assess knee OA severity, the first step is to localize the region of interest that is to detect and extract the knee joint regions from the radiographs, and the next step is to classify the localized knee joints based on a radiographic classification scheme such as Kellgren and Lawrence grades. First, the existing approaches for detecting (or localizing) the knee joint regions based on handcrafted features are reviewed and outlined in this chapter. Next, three new approaches are introduced: 1) to automatically detect the knee joint region using a fully convolutional network, 2) to automatically assess the radiographic knee OA using CNNs trained from scratch for classification and regression of knee joint images to predict KL grades in ordinal and continuous scales, and 3) to quantify the knee OA severity optimizing a weighted ratio of two loss functions: categorical cross entropy and mean-squared error using multi-objective convolutional learning. The results from these methods show progressive improvement in the overall quantification of the knee OA severity. Two public datasets: the OAI and the MOST are used to evaluate the approaches with promising results that outperform existing approaches. In summary, this work primarily contributes to the field

Feature learning refers to techniques that learn to transform raw data input into an effective representation for further higher-level processing in many computer vision tasks. This chapter presents the investigations and the results of feature learning using convolutional neural networks to automatically assess knee osteoarthritis (OA) severity and the associated clinical and diagnostic features of knee OA from radiographs (X-ray images). Also, this chapter demonstrates that feature learning in a supervised manner is more effective than using conventional handcrafted features for automatic detection of knee joints and fine-grained knee OA image classification. In the general machine learning approach to automatically assess knee OA severity, the first step is to localize the region of interest that is to detect and extract the knee joint regions from the radiographs, and the next step is to classify the localized knee joints based on a radiographic classification scheme such as Kellgren and Lawrence grades. First, the existing approaches for detecting (or localizing) the knee joint regions based on handcrafted features are reviewed and outlined in this chapter. Next, three new approaches are introduced: 1) to automatically detect the knee joint region using a fully convolutional network, 2) to automatically assess the radiographic knee OA using CNNs trained from scratch for classification and regression of knee joint images to predict KL grades in ordinal and continuous scales, and 3) to quantify the knee OA severity optimizing a weighted ratio of two loss functions: categorical cross entropy and mean-squared error using multi-objective convolutional learning. The results from these methods show progressive improvement in the overall quantification of the knee OA severity. Two public datasets: the OAI and the MOST are used to evaluate the approaches with promising results that outperform existing approaches. In summary, this work primarily contributes to the field

Feature learning refers to techniques that learn to transform raw data input into an effective representation for further higher-level processing in many computer vision tasks. This chapter presents the investigations and the results of feature learning using convolutional neural networks to automatically assess knee osteoarthritis (OA) severity and the associated clinical and diagnostic features of knee OA from radiographs (X-ray images). Also, this chapter demonstrates that feature learning in a supervised manner is more effective than using conventional handcrafted features for automatic detection of knee joints and fine-grained knee OA image classification. In the general machine learning approach to automatically assess knee OA severity, the first step is to localize the region of interest that is to detect and extract the knee joint regions from the radiographs, and the next step is to classify the localized knee joints based on a radiographic classification scheme such as Kellgren and Lawrence grades. First, the existing approaches for detecting (or localizing) the knee joint regions based on handcrafted features are reviewed and outlined in this chapter. Next, three new approaches are introduced: 1) to automatically detect the knee joint region using a fully convolutional network, 2) to automatically assess the radiographic knee OA using CNNs trained from scratch for classification and regression of knee joint images to predict KL grades in ordinal and continuous scales, and 3) to quantify the knee OA severity optimizing a weighted ratio of two loss functions: categorical cross entropy and mean-squared error using multi-objective convolutional learning. The results from these methods show progressive improvement in the overall quantification of the knee OA severity. Two public datasets: the OAI and the MOST are used to evaluate the approaches with promising results that outperform existing approaches. In summary, this work primarily contributes to the field

Feature learning refers to techniques that learn to transform raw data input into an effective representation for further higher-level processing in many computer vision tasks. This chapter presents the investigations and the results of feature learning using convolutional neural networks to automatically assess knee osteoarthritis (OA) severity and the associated clinical and diagnostic features of knee OA from radiographs (X-ray images). Also, this chapter demonstrates that feature learning in a supervised manner is more effective than using conventional handcrafted features for automatic detection of knee joints and fine-grained knee OA image classification. In the general machine learning approach to automatically assess knee OA severity, the first step is to localize the region of interest that is to detect and extract the knee joint regions from the radiographs, and the next step is to classify the localized knee joints based on a radiographic classification scheme such as Kellgren and Lawrence grades. First, the existing approaches for detecting (or localizing) the knee joint regions based on handcrafted features are reviewed and outlined in this chapter. Next, three new approaches are introduced: 1) to automatically detect the knee joint region using a fully convolutional network, 2) to automatically assess the radiographic knee OA using CNNs trained from scratch for classification and regression of knee joint images to predict KL grades in ordinal and continuous scales, and 3) to quantify the knee OA severity optimizing a weighted ratio of two loss functions: categorical cross entropy and mean-squared error using multi-objective convolutional learning. The results from these methods show progressive improvement in the overall quantification of the knee OA severity. Two public datasets: the OAI and the MOST are used to evaluate the approaches with promising results that outperform existing approaches. In summary, this work primarily contributes to the field

The numbers of students enrolled in the Radiography programme at the Central University of Technology (CUT) are increasing each year, and selectors are therefore faced with critical admission decisions when selecting prospective students for a programme with limited spaces. This research study was undertaken to assess current student selection criteria in Radiography education, by retrospectively analysing the degree to which these criteria predicted actual academic performance over a three-year period amongst entry-level students enrolled in a three-year National Diploma in Radiography. The rationale behind the interest in student selection in Radiography education was the fact that the South African Qualification Authority (SAQA) has approved and registered a professional Bachelor’s degree in Radiography (480 credits), with a new curriculum and altered exit-level outcomes, including research skills. The transition from a diploma-level to a degree-level qualification in Radiography in South Africa indicated that a review and an assessment of current diploma-level student selection criteria could provide a benchmark for selecting students for degree-level education. An extensive literature study and document analysis provided background information and perspectives related to student admission criteria and selection tests in the health sciences, both nationally and internationally. The research design and methodology included a retrospective quantitative study and document analysis. In the research process, data were collected for each student, including biographical data, prior tertiary education, matriculation or National Senior Certificate (NSC) subject marks and a total Admission Points Score (APS) as calculated from each matriculation certificate. The CUT selection test results for each student were also documented, and included a General Scholastic Aptitude Test (GSAT) score, an English Proficiency Test score and a Self-Directed Search Questionnaire score. By

In hospitals, clinics, nursing homes, and medical laboratories, diagnostic radiography is extensively used to view internal structures of a patient and to aid radiologists and doctors diagnose and treat illness and injury. The most common medical radiography is by passing x-rays through the body and captures the image of the patient by means of a sensitized film. This radiographic procedure must feature the system of dose optimization and limitation known as the practice of ALARA (as low as reasonably achievable) recommended by the International Commission on Radiological Protection. The measurements of the entrance surface dose (ESD) and the critical organ dose (COD) in common medical radiography are very important to ensure the dose limit received by the patient in a single radiographic exposure must fulfill the ALARA principle. The most useful way to evaluate ESD is either by indirect measurement on a phantom using an ionization chamber or thermo luminescent dosimeter (TLD) or using calculation based on mathematical model.

In hospitals, clinics, nursing homes, and medical laboratories, diagnostic radiography is extensively used to view internal structures of a patient and to aid radiologists and doctors diagnose and treat illness and injury. The most common medical radiography is by passing x-rays through the body and captures the image of the patient by means of a sensitized film. This radiographic procedure must feature the system of dose optimization and limitation known as the practice of ALARA (as low as reasonably achievable) recommended by the International Commission on Radiological Protection. The measurements of the entrance surface dose (ESD) and the critical organ dose (COD) in common medical radiography are very important to ensure the dose limit received by the patient in a single radiographic exposure must fulfill the ALARA principle. The most useful way to evaluate ESD is either by indirect measurement on a phantom using an ionization chamber or thermo luminescent dosimeter (TLD) or using calculation based on mathematical model.

Computed tomography colonography (CTC) is emerging as an alternative to conventional colonoscopy (CC). However CTC is not yet in widespread use due in part to the lack of suitably trained radiologists. We have developed a novel remote access system to train radiologists for colorectal cancer screening using CTC. To ensure that radiologists can gain the relevant experience without the need for any specialist equipment or software, we opted for designing a system that is accessible via the Internet using a standard browser. The interface lets the user locate and characterise polyps with the help of appropriate tools such as windowing, polyp measurement, zooming and a 3-D view. Each user has an account in order to allow monitoring of their training. They can also run an automatic evaluation of their work based on gold standard information previously gathered from specialists. This thesis also describes an initial implementation exclusively made up of Java Servlets. The evaluation of this system has been discussed in order to determine a better approach. The final system has been developed using a combination of Java Servlets and Applets. This approach offers fast response time to the user-interface. An iteration of lumen tracking using the system takes approximately 45 seconds. This research has yielded an operational system that meets the needs of remote access users.

Computed tomography colonography (CTC) is emerging as an alternative to conventional colonoscopy (CC). However CTC is not yet in widespread use due in part to the lack of suitably trained radiologists. We have developed a novel remote access system to train radiologists for colorectal cancer screening using CTC. To ensure that radiologists can gain the relevant experience without the need for any specialist equipment or software, we opted for designing a system that is accessible via the Internet using a standard browser. The interface lets the user locate and characterise polyps with the help of appropriate tools such as windowing, polyp measurement, zooming and a 3-D view. Each user has an account in order to allow monitoring of their training. They can also run an automatic evaluation of their work based on gold standard information previously gathered from specialists. This thesis also describes an initial implementation exclusively made up of Java Servlets. The evaluation of this system has been discussed in order to determine a better approach. The final system has been developed using a combination of Java Servlets and Applets. This approach offers fast response time to the user-interface. An iteration of lumen tracking using the system takes approximately 45 seconds. This research has yielded an operational system that meets the needs of remote access users.

Computed tomography colonography (CTC) is emerging as an alternative to conventional colonoscopy (CC). However CTC is not yet in widespread use due in part to the lack of suitably trained radiologists. We have developed a novel remote access system to train radiologists for colorectal cancer screening using CTC. To ensure that radiologists can gain the relevant experience without the need for any specialist equipment or software, we opted for designing a system that is accessible via the Internet using a standard browser. The interface lets the user locate and characterise polyps with the help of appropriate tools such as windowing, polyp measurement, zooming and a 3-D view. Each user has an account in order to allow monitoring of their training. They can also run an automatic evaluation of their work based on gold standard information previously gathered from specialists. This thesis also describes an initial implementation exclusively made up of Java Servlets. The evaluation of this system has been discussed in order to determine a better approach. The final system has been developed using a combination of Java Servlets and Applets. This approach offers fast response time to the user-interface. An iteration of lumen tracking using the system takes approximately 45 seconds. This research has yielded an operational system that meets the needs of remote access users.

Computed tomography colonography (CTC) is emerging as an alternative to conventional colonoscopy (CC). However CTC is not yet in widespread use due in part to the lack of suitably trained radiologists. We have developed a novel remote access system to train radiologists for colorectal cancer screening using CTC. To ensure that radiologists can gain the relevant experience without the need for any specialist equipment or software, we opted for designing a system that is accessible via the Internet using a standard browser. The interface lets the user locate and characterise polyps with the help of appropriate tools such as windowing, polyp measurement, zooming and a 3-D view. Each user has an account in order to allow monitoring of their training. They can also run an automatic evaluation of their work based on gold standard information previously gathered from specialists. This thesis also describes an initial implementation exclusively made up of Java Servlets. The evaluation of this system has been discussed in order to determine a better approach. The final system has been developed using a combination of Java Servlets and Applets. This approach offers fast response time to the user-interface. An iteration of lumen tracking using the system takes approximately 45 seconds. This research has yielded an operational system that meets the needs of remote access users.

Le but de notre travail a été d’évaluer l’effet de la réduction de la dose d’irradiation en TDM multibarrette quant à la performance diagnostique, la confiance de l’observateur dans le diagnostic proposé, la capacité à suggérer un diagnostic alternatif dans quelques pathologies courantes et/ou caractérisées par des situations de faibles contrastes entre les structures anatomiques normales ou pathologiques. Nous avons donc comparé ces paramètres entre des TDM à doses réduites et à doses standard telles que couramment rapportées dans la littérature dans les circonstances suivantes :•, Doctorat en sciences médicales, info:eu-repo/semantics/nonPublished

Le but de notre travail a été d’évaluer l’effet de la réduction de la dose d’irradiation en TDM multibarrette quant à la performance diagnostique, la confiance de l’observateur dans le diagnostic proposé, la capacité à suggérer un diagnostic alternatif dans quelques pathologies courantes et/ou caractérisées par des situations de faibles contrastes entre les structures anatomiques normales ou pathologiques. Nous avons donc comparé ces paramètres entre des TDM à doses réduites et à doses standard telles que couramment rapportées dans la littérature dans les circonstances suivantes :•, Doctorat en sciences médicales, info:eu-repo/semantics/nonPublished

Mammography is currently used to screen women in targeted risk classes for breast cancer. Computer assisted diagnosis of mammograms attempts to lower the workload on radiologists by either automating some of their tasks or acting as a second reader. The task of mammogram segmentation based on pixel intensity is addressed in this thesis. The mammographic process leads to images where intensity in the image is related to the composition of tissue in the breast; it is therefore possible to segment a mammogram into several regions using a combination of global thresholds, local thresholds and higher-level information based on the intensity histogram. A hierarchical view is taken of the segmentation process, with a series of steps that feed into each other. Methods are presented for segmentation of: 1. image background regions; 2. skin-air interface; 3. pectoral muscle; and 4. segmentation of the database by classification of mammograms into tissue types and determining a similarity measure between mammograms. All methods are automatic. After a detailed analysis of minimum cross-entropy thresholding, multi-level thresholding is used to segment the main breast tissue from the background. Scanning artefacts and high intensity noise are separated from the breast tissue using binary image operations, rectangular labels are identified from the binary image by their shape, the Radon transform is used to locate the edges of tape artefacts, and a filter is used to locate vertical running roller scratching. Orientation of the image is determined using the shape of the breast and properties of the breast tissue near the breast edge. Unlike most existing orientation algorithms, which only distinguish between left facing or right facing breasts, the algorithm developed determines orientation for images flipped upside down or rotated onto their side and works successfully on all images of the testing database. Orientation is an integral part of the segmentation process, as skin-air i, Thesis (Ph.D.)--University of Western Australia, 2004

Free chest X-rays made possible by Christmas seal sales draws hundreds of University of Minnesota students to the mobile unit for examinations. The queue forms on the mall at Northrop auditorium.

A positive test leads to the chest X-ray shown being given at the Litchfield hospital by technician Flynn.

Barbara Shepherd, left, General hospital X-ray technician, lines up the X-ray machine over Smith's broken wrist. Smith fell from a ladder as she took down several hornets' nests.

Erik Olsson, X-ray technician from Glen Lake sanatorium, prepares to make a chest X-ray photo of Emil Olson.

Attendants hold the subject while Olek (left), member of the public library Nature Study group, looks at a leg fracture of a dog through the fluoroscope. Modern equipment and scientific devices for treatment of animals were shown to the study group visitors at the Morgan Animal hospital. Other individuals in the photo are unidentified.

United States Army Private First Class Redman, from Buffalo Lake, Minnesota, receives a chest x-ray as part of the medical screening portion of his discharge process at Camp McCoy's Separation Center in Wisconsin.

Andrew Nelson, age 15, lies in a hospital bed as Nurse Stella Glaim prepares him for X-ray. He was injured in a traffic accident while on his motor scooter.

Dr. Ancel Keys, with the University of Minnesota, moves United States Army Private Stacy, stationed at Fort Snelling, closer to the X-ray machine to measure the amount of blood the soldier's heart is pumping after exercising on the laboratory treadmill, as well as any change in size.

L to R: School children, Dwight Davis and Judy Brehm, ride through Richfield with Fire Chief Meeker on a fire engine with a sign reminding people that free tuberculosis screening will be provided by the Hennepin County Tuberculosis Association's mobile X-Ray unit.

Dr. Ancel Keys, with the University of Minnesota, moves United States Army Private Stacy, stationed at Fort Snelling, closer to the X-ray machine to measure the amount of blood the soldier's heart is pumping after exercising on the laboratory treadmill, as well as any change in size.

Swisher, 70-year-old department store Santa, given chest x-ray by Technician Acker with the Christmas Seal mobile unit to screen for tuberculosis before contact with children.

Hennepin County Tuberculosis Association is giving free chest x-rays to elementary school teachers first. L to R: Staff member Mason directs toward X-ray Technician Carter a line of Penn School teachers; Segal, Carlson, Weber, Faulkner and Holbeck.

General Hospital X-Ray Technicians, Weekley (left) and Steen show the tiny plaster cast on the owl's broken leg with a magnifying glass.

Christensen, x-ray technician, positions Nobles for X-ray.

L to R: Rovang, Erickson, Simertz

L to R: Ramsey Junior High School 7th Grade student, Pat Anderson, stands behind a model "fluoroscope" made by Jimmy Beiswanger and his partner for their radiographic science project at the school's Science Fair. Beiswanger's father, Dr. Richard Beiswanger (not pictured), is a radiologist.