Your browser doesn't support javascript.
Mostrar: 20 | 50 | 100
Resultados 1 - 20 de 46
Filtrar
Mais filtros










Base de dados
Intervalo de ano de publicação
1.
Radiology ; 294(1): 199-209, 2020 Jan.
Artigo em Inglês | MEDLINE | ID: mdl-31714194

RESUMO

Background Multicenter studies are required to validate the added benefit of using deep convolutional neural network (DCNN) software for detecting malignant pulmonary nodules on chest radiographs. Purpose To compare the performance of radiologists in detecting malignant pulmonary nodules on chest radiographs when assisted by deep learning-based DCNN software with that of radiologists or DCNN software alone in a multicenter setting. Materials and Methods Investigators at four medical centers retrospectively identified 600 lung cancer-containing chest radiographs and 200 normal chest radiographs. Each radiograph with a lung cancer had at least one malignant nodule confirmed by CT and pathologic examination. Twelve radiologists from the four centers independently analyzed the chest radiographs and marked regions of interest. Commercially available deep learning-based computer-aided detection software separately trained, tested, and validated with 19 330 radiographs was used to find suspicious nodules. The radiologists then reviewed the images with the assistance of DCNN software. The sensitivity and number of false-positive findings per image of DCNN software, radiologists alone, and radiologists with the use of DCNN software were analyzed by using logistic regression and Poisson regression. Results The average sensitivity of radiologists improved (from 65.1% [1375 of 2112; 95% confidence interval {CI}: 62.0%, 68.1%] to 70.3% [1484 of 2112; 95% CI: 67.2%, 73.1%], P < .001) and the number of false-positive findings per radiograph declined (from 0.2 [488 of 2400; 95% CI: 0.18, 0.22] to 0.18 [422 of 2400; 95% CI: 0.16, 0.2], P < .001) when the radiologists re-reviewed radiographs with the DCNN software. For the 12 radiologists in this study, 104 of 2400 radiographs were positively changed (from false-negative to true-positive or from false-positive to true-negative) using the DCNN, while 56 of 2400 radiographs were changed negatively. Conclusion Radiologists had better performance with deep convolutional network software for the detection of malignant pulmonary nodules on chest radiographs than without. © RSNA, 2019 Online supplemental material is available for this article. See also the editorial by Jacobson in this issue.

2.
J Am Coll Radiol ; 16(11): 1516-1521, 2019 Nov.
Artigo em Inglês | MEDLINE | ID: mdl-31585696

RESUMO

This is a condensed summary of an international multisociety statement on ethics of artificial intelligence (AI) in radiology produced by the ACR, European Society of Radiology, RSNA, Society for Imaging Informatics in Medicine, European Society of Medical Imaging Informatics, Canadian Association of Radiologists, and American Association of Physicists in Medicine. AI has great potential to increase efficiency and accuracy throughout radiology, but it also carries inherent pitfalls and biases. Widespread use of AI-based intelligent and autonomous systems in radiology can increase the risk of systemic errors with high consequence and highlights complex ethical and societal issues. Currently, there is little experience using AI for patient care in diverse clinical settings. Extensive research is needed to understand how to best deploy AI in clinical practice. This statement highlights our consensus that ethical use of AI in radiology should promote well-being, minimize harm, and ensure that the benefits and harms are distributed among stakeholders in a just manner. We believe AI should respect human rights and freedoms, including dignity and privacy. It should be designed for maximum transparency and dependability. Ultimate responsibility and accountability for AI remains with its human designers and operators for the foreseeable future. The radiology community should start now to develop codes of ethics and practice for AI that promote any use that helps patients and the common good and should block use of radiology data and algorithms for financial gain without those two attributes.

3.
Can Assoc Radiol J ; 70(4): 329-334, 2019 Nov.
Artigo em Inglês | MEDLINE | ID: mdl-31585825

RESUMO

This is a condensed summary of an international multisociety statement on ethics of artificial intelligence (AI) in radiology produced by the ACR, European Society of Radiology, RSNA, Society for Imaging Informatics in Medicine, European Society of Medical Imaging Informatics, Canadian Association of Radiologists, and American Association of Physicists in Medicine. AI has great potential to increase efficiency and accuracy throughout radiology, but it also carries inherent pitfalls and biases. Widespread use of AI-based intelligent and autonomous systems in radiology can increase the risk of systemic errors with high consequence and highlights complex ethical and societal issues. Currently, there is little experience using AI for patient care in diverse clinical settings. Extensive research is needed to understand how to best deploy AI in clinical practice. This statement highlights our consensus that ethical use of AI in radiology should promote well-being, minimize harm, and ensure that the benefits and harms are distributed among stakeholders in a just manner. We believe AI should respect human rights and freedoms, including dignity and privacy. It should be designed for maximum transparency and dependability. Ultimate responsibility and accountability for AI remains with its human designers and operators for the foreseeable future. The radiology community should start now to develop codes of ethics and practice for AI that promote any use that helps patients and the common good and should block use of radiology data and algorithms for financial gain without those two attributes.


Assuntos
Inteligência Artificial/ética , Radiologia/ética , Canadá , Consenso , Europa (Continente) , Humanos , Radiologistas/ética , Sociedades Médicas , Estados Unidos
4.
Insights Imaging ; 10(1): 101, 2019 Oct 01.
Artigo em Inglês | MEDLINE | ID: mdl-31571015

RESUMO

This is a condensed summary of an international multisociety statement on ethics of artificial intelligence (AI) in radiology produced by the ACR, European Society of Radiology, RSNA, Society for Imaging Informatics in Medicine, European Society of Medical Imaging Informatics, Canadian Association of Radiologists, and American Association of Physicists in Medicine.AI has great potential to increase efficiency and accuracy throughout radiology, but also carries inherent pitfalls and biases. Widespread use of AI-based intelligent and autonomous systems in radiology can increase the risk of systemic errors with high consequence, and highlights complex ethical and societal issues. Currently, there is little experience using AI for patient care in diverse clinical settings. Extensive research is needed to understand how to best deploy AI in clinical practice.This statement highlights our consensus that ethical use of AI in radiology should promote well-being, minimize harm, and ensure that the benefits and harms are distributed among stakeholders in a just manner. We believe AI should respect human rights and freedoms, including dignity and privacy. It should be designed for maximum transparency and dependability. Ultimate responsibility and accountability for AI remains with its human designers and operators for the foreseeable future.The radiology community should start now to develop codes of ethics and practice for AI which promote any use that helps patients and the common good and should block use of radiology data and algorithms for financial gain without those two attributes.

5.
Radiology ; 293(2): 436-440, 2019 Nov.
Artigo em Inglês | MEDLINE | ID: mdl-31573399

RESUMO

This is a condensed summary of an international multisociety statement on ethics of artificial intelligence (AI) in radiology produced by the ACR, European Society of Radiology, RSNA, Society for Imaging Informatics in Medicine, European Society of Medical Imaging Informatics, Canadian Association of Radiologists, and American Association of Physicists in Medicine. AI has great potential to increase efficiency and accuracy throughout radiology, but it also carries inherent pitfalls and biases. Widespread use of AI-based intelligent and autonomous systems in radiology can increase the risk of systemic errors with high consequence and highlights complex ethical and societal issues. Currently, there is little experience using AI for patient care in diverse clinical settings. Extensive research is needed to understand how to best deploy AI in clinical practice. This statement highlights our consensus that ethical use of AI in radiology should promote well-being, minimize harm, and ensure that the benefits and harms are distributed among stakeholders in a just manner. We believe AI should respect human rights and freedoms, including dignity and privacy. It should be designed for maximum transparency and dependability. Ultimate responsibility and accountability for AI remains with its human designers and operators for the foreseeable future. The radiology community should start now to develop codes of ethics and practice for AI that promote any use that helps patients and the common good and should block use of radiology data and algorithms for financial gain without those two attributes. This article is a simultaneous joint publication in Radiology, Journal of the American College of Radiology, Canadian Association of Radiologists Journal, and Insights into Imaging. Published under a CC BY-NC-ND 4.0 license. Online supplemental material is available for this article.

6.
PLoS One ; 14(3): e0213339, 2019.
Artigo em Inglês | MEDLINE | ID: mdl-30835766

RESUMO

PURPOSE: To compare the diagnostic performance and raters´confidence of radiography, radiography equivalent dose multi-detector computed tomography (RED-MDCT) and radiography equivalent dose cone beam computed tomography (RED-CBCT) for finger fractures. METHODS: Fractures were inflicted artificially and randomly to 10 cadaveric hands of body donors. Radiography as well as RED-MDCT and RED-CBCT imaging were performed at dose settings equivalent to radiography. Images were de-identified and analyzed by three radiologists regarding finger fractures, joint involvement and confidence with their findings. Reference standard was consensus reading by two radiologists of the fracturing protocol and high-dose multi-detector computed tomography (MDCT) images. Sensitivity and specificity were calculated and compared with Cochrane´s Q and post hoc analysis. Rater´s confidence was calculated with Friedman Test and post hoc Nemenyi Test. RESULTS: Rater´s confidence, inter-rater correlation, specificity for fractures and joint involvement were higher in RED-MDCT and RED-CBCT compared to radiography. No differences between the modalities were found regarding sensitivity. CONCLUSION: In this phantom study, radiography equivalent dose computed tomography (RED-CT) demonstrates a partly higher diagnostic accuracy than radiography. Implementing RED-CT in the diagnostic work-up of finger fractures could improve diagnostics, support correct classification and adequate treatment. Clinical studies should be performed to confirm these preliminary results.


Assuntos
Tomografia Computadorizada de Feixe Cônico/métodos , Traumatismos dos Dedos/diagnóstico , Fraturas Ósseas/diagnóstico , Tomografia Computadorizada Multidetectores/métodos , Imagens de Fantasmas , Radiografia/métodos , Traumatismos dos Dedos/diagnóstico por imagem , Fraturas Ósseas/diagnóstico por imagem , Humanos , Doses de Radiação
7.
Transpl Infect Dis ; 20(6): e12993, 2018 Dec.
Artigo em Inglês | MEDLINE | ID: mdl-30187615

RESUMO

Long-term success of lung transplantation is limited by allograft dysfunction and frequent infections. Varicella zoster virus infection (VZV) is one of the most common opportunistic infections among solid organ transplantation recipients. However the occurrence of visceral involvement or disseminated disease, as seen after bone marrow transplantation, is rare. We report a case of a 59-year-old woman who underwent double-lung transplantation with a fatal visceral and disseminated varicella zoster virus infection.


Assuntos
Herpesvirus Humano 3/isolamento & purificação , Imunossupressão/efeitos adversos , Transplante de Pulmão/efeitos adversos , Fibrose Pulmonar/cirurgia , Choque Séptico/imunologia , Infecção pelo Vírus da Varicela-Zoster/imunologia , Dor Abdominal/imunologia , Dor Abdominal/virologia , Exantema/imunologia , Exantema/microbiologia , Evolução Fatal , Feminino , Humanos , Imunossupressão/métodos , Pessoa de Meia-Idade , Choque Séptico/virologia , Inconsciência/imunologia , Inconsciência/virologia , Infecção pelo Vírus da Varicela-Zoster/complicações , Infecção pelo Vírus da Varicela-Zoster/virologia
8.
Ann N Y Acad Sci ; 1434(1): 274-281, 2018 12.
Artigo em Inglês | MEDLINE | ID: mdl-29766512

RESUMO

Significant technical advances have been made in radiology since the first discovery of X-rays. Diagnostic techniques have become more and more complex, workflows have been digitized, and data production has increased exponentially. However, the radiology report as the main method for communicating examination results has largely remained unchanged. Growing evidence supports that more structured radiology reports offer various benefits over conventional narrative reports. Various efforts have been made to further develop and promote structured reporting. However, regardless of the potential benefits, structured reporting has still not seen widespread implementation into the clinical routine. With recent technical advances, especially new research topics such as big data and machine learning, structured reporting could prove essential for the future of radiology. New interoperable solutions are needed to facilitate the implementation of template-based structured reporting into the clinical routine.


Assuntos
Processamento de Imagem Assistida por Computador/instrumentação , Processamento de Imagem Assistida por Computador/métodos , Processamento de Imagem Assistida por Computador/tendências , Aprendizado de Máquina , Humanos
9.
J Arthroplasty ; 33(8): 2652-2659.e3, 2018 08.
Artigo em Inglês | MEDLINE | ID: mdl-29615377

RESUMO

BACKGROUND: Correct positioning of the cup is an important factor in total hip arthroplasty. Assessing its position from a plain anteroposterior pelvic radiograph is known to be hampered by systemic errors. This study focuses on developing a correction method to adjust for these potential sources of error and to eliminate them based on a 3D geometric analysis. METHODS: Computed tomography scans of 113 (66 male, 47 female) pelvices were reconstructed and virtually projected onto a plain radiograph with varying rotational and translational positions. Thus cup inclination and anteversion as measured on a 2D-radiograph and in the 3D environment were correlated. Projected offset of the symphysis from the mid-sacrum served as a mean to measure pelvic right/left-rotation. Pelvic tilt was determined from the projected height of the contour of the small pelvis. Correction formulas were verified by projecting a gimbal-mounted artificial pelvis with a cup implanted in a known position. RESULTS: We found gender-specific formulas that correct for malrotated and off-centered radiographs. Applying these formulas cup inclination was assessed as close as 1.3° (±1.90°) to the true 3D value and cup anteversion as close as 1° (±1.91°) although deviations between directly measured plain values and corrected values rose up to 18°. CONCLUSION: Inherent effects of central projection and malrotations due to pelvic tilt, pelvic rotation, and noncentered radiographs are corrected. Evaluation of radiographic inclination and anteversion of acetabular cups from plain 2D-radiographs show improved precision. Real values are approached better than 1.3° when applying our correction formulas.


Assuntos
Artroplastia de Quadril/métodos , Prótese de Quadril , Tomografia Computadorizada por Raios X/métodos , Acetábulo/diagnóstico por imagem , Acetábulo/cirurgia , Adulto , Idoso , Idoso de 80 Anos ou mais , Feminino , Humanos , Processamento de Imagem Assistida por Computador , Imagem Tridimensional , Masculino , Pessoa de Meia-Idade , Posicionamento do Paciente , Pelve/diagnóstico por imagem , Pelve/cirurgia , Radiografia , Rotação , Fatores Sexuais
10.
Sci Rep ; 8(1): 3906, 2018 03 02.
Artigo em Inglês | MEDLINE | ID: mdl-29500380

RESUMO

The aim of this study was to evaluate and compare the diagnostic accuracy, the inter-rater agreement and raters' certainty of cone beam computed tomography (CBCT) and radiography for the detection of scaphoid fractures. Our hypothesis is that the CBCT has a higher diagnostic accuracy for scaphoid fractures than radiography. We retrospectively analysed patients who underwent both radiography and CBCT examinations within 4 days to rule out a scaphoid fracture over a 2-year period in our institution. 4 blinded radiologists and orthopaedic surgeons independently rated the images regarding the presence of a scaphoid fracture. The reference standard was evaluated by two radiologists in a consensus reading. Inter-rater correlation was evaluated, pooled sensitivity, specificity, positive and negative predictive values were calculated and compared. 102 patients met the inclusion criteria. 52% of them had a scaphoid fracture. The inter-rater correlation was higher in the CBCT compared to radiography (P < 0.001). Sensitivity, specificity, positive and negative predictive values were higher for CBCT than for radiography (P < 0.019). Observers' fracture classifications showed a higher correlation with the reference standard in the CBCT. Observers' certainty for fracture detection and classification were higher in the CBCT. CBCT shows a higher diagnostic accuracy for scaphoid fractures than radiography.


Assuntos
Tomografia Computadorizada de Feixe Cônico/métodos , Radiografia/métodos , Fraturas do Rádio/diagnóstico , Osso Escafoide/patologia , Adulto , Feminino , Humanos , Masculino , Prognóstico , Fraturas do Rádio/diagnóstico por imagem , Estudos Retrospectivos , Osso Escafoide/diagnóstico por imagem , Osso Escafoide/lesões
11.
Int J Comput Assist Radiol Surg ; 12(3): 485-491, 2017 Mar.
Artigo em Inglês | MEDLINE | ID: mdl-27722873

RESUMO

PURPOSE: In the diagnostic process of primary bone tumors, patient age, tumor localization and to a lesser extent sex affect the differential diagnosis. We therefore aim to develop a pretest probability calculator for primary malignant bone tumors based on population data taking these variables into account. METHODS: We access the SEER (Surveillance, Epidemiology and End Results Program of the National Cancer Institute, 2015 release) database and analyze data of all primary malignant bone tumors diagnosed between 1973 and 2012. We record age at diagnosis, tumor localization according to the International Classification of Diseases (ICD-O-3) and sex. We take relative probability of the single tumor entity as a surrogate parameter for unadjusted pretest probability. We build a probabilistic (naïve Bayes) classifier to calculate pretest probabilities adjusted for age, tumor localization and sex. RESULTS: We analyze data from 12,931 patients (647 chondroblastic osteosarcomas, 3659 chondrosarcomas, 1080 chordomas, 185 dedifferentiated chondrosarcomas, 2006 Ewing's sarcomas, 281 fibroblastic osteosarcomas, 129 fibrosarcomas, 291 fibrous malignant histiocytomas, 289 malignant giant cell tumors, 238 myxoid chondrosarcomas, 3730 osteosarcomas, 252 parosteal osteosarcomas, 144 telangiectatic osteosarcomas). We make our probability calculator accessible at http://ebm-radiology.com/bayesbone/index.html . We provide exhaustive tables for age and localization data. Results from tenfold cross-validation show that in 79.8 % of cases the pretest probability is correctly raised. CONCLUSIONS: Our approach employs population data to calculate relative pretest probabilities for primary malignant bone tumors. The calculator is not diagnostic in nature. However, resulting probabilities might serve as an initial evaluation of probabilities of tumors on the differential diagnosis list.


Assuntos
Neoplasias Ósseas/diagnóstico por imagem , Cordoma/diagnóstico por imagem , Tumor de Células Gigantes do Osso/diagnóstico por imagem , Modelos Estatísticos , Sarcoma/diagnóstico por imagem , Teorema de Bayes , Neoplasias Ósseas/diagnóstico , Condrossarcoma/diagnóstico , Condrossarcoma/diagnóstico por imagem , Cordoma/diagnóstico , Bases de Dados Factuais , Diagnóstico Diferencial , Feminino , Fibrossarcoma/diagnóstico , Fibrossarcoma/diagnóstico por imagem , Tumor de Células Gigantes do Osso/diagnóstico , Histiocitoma Fibroso Maligno/diagnóstico , Histiocitoma Fibroso Maligno/diagnóstico por imagem , Humanos , Masculino , Osteossarcoma/diagnóstico , Osteossarcoma/diagnóstico por imagem , Probabilidade , Radiografia , Programa de SEER , Sarcoma/diagnóstico , Sarcoma de Ewing/diagnóstico , Sarcoma de Ewing/diagnóstico por imagem , Fatores Sexuais
12.
Medicine (Baltimore) ; 95(44): e5309, 2016 Nov.
Artigo em Inglês | MEDLINE | ID: mdl-27858910

RESUMO

The mixed convolution kernel alters his properties geographically according to the depicted organ structure, especially for the lung. Therefore, we compared the image quality of the mixed convolution kernel to standard soft and hard kernel reconstructions for different organ structures in thoracic computed tomography (CT) images.Our Ethics Committee approved this prospective study. In total, 31 patients who underwent contrast-enhanced thoracic CT studies were included after informed consent. Axial reconstructions were performed with hard, soft, and mixed convolution kernel. Three independent and blinded observers rated the image quality according to the European Guidelines for Quality Criteria of Thoracic CT for 13 organ structures. The observers rated the depiction of the structures in all reconstructions on a 5-point Likert scale. Statistical analysis was performed with the Friedman Test and post hoc analysis with the Wilcoxon rank-sum test.Compared to the soft convolution kernel, the mixed convolution kernel was rated with a higher image quality for lung parenchyma, segmental bronchi, and the border between the pleura and the thoracic wall (P < 0.03). Compared to the hard convolution kernel, the mixed convolution kernel was rated with a higher image quality for aorta, anterior mediastinal structures, paratracheal soft tissue, hilar lymph nodes, esophagus, pleuromediastinal border, large and medium sized pulmonary vessels and abdomen (P < 0.004) but a lower image quality for trachea, segmental bronchi, lung parenchyma, and skeleton (P < 0.001).The mixed convolution kernel cannot fully substitute the standard CT reconstructions. Hard and soft convolution kernel reconstructions still seem to be mandatory for thoracic CT.


Assuntos
Algoritmos , Radiografia Torácica , Tórax/diagnóstico por imagem , Tomografia Computadorizada por Raios X , Feminino , Humanos , Masculino , Pessoa de Meia-Idade , Estudos Prospectivos , Radiografia Torácica/métodos , Análise Espacial , Tomografia Computadorizada por Raios X/métodos
13.
PLoS One ; 11(10): e0164859, 2016.
Artigo em Inglês | MEDLINE | ID: mdl-27788215

RESUMO

PURPOSE: To compare the diagnostic accuracy of radiography, to radiography equivalent dose multidetector computed tomography (RED-MDCT) and to radiography equivalent dose cone beam computed tomography (RED-CBCT) for wrist fractures. METHODS: As study subjects we obtained 10 cadaveric human hands from body donors. Distal radius, distal ulna and carpal bones (n = 100) were artificially fractured in random order in a controlled experimental setting. We performed radiation dose equivalent radiography (settings as in standard clinical care), RED-MDCT in a 320 row MDCT with single shot mode and RED-CBCT in a device dedicated to musculoskeletal imaging. Three raters independently evaluated the resulting images for fractures and the level of confidence for each finding. Gold standard was evaluated by consensus reading of a high-dose MDCT. RESULTS: Pooled sensitivity was higher in RED-MDCT with 0.89 and RED-MDCT with 0.81 compared to radiography with 0.54 (P = < .004). No significant differences were detected concerning the modalities' specificities (with values between P = .98). Raters' confidence was higher in RED-MDCT and RED-CBCT compared to radiography (P < .001). CONCLUSION: The diagnostic accuracy of RED-MDCT and RED-CBCT for wrist fractures proved to be similar and in some parts even higher compared to radiography. Readers are more confident in their reporting with the cross sectional modalities. Dose equivalent cross sectional computed tomography of the wrist could replace plain radiography for fracture diagnosis in the long run.


Assuntos
Tomografia Computadorizada de Feixe Cônico , Tomografia Computadorizada Multidetectores , Radiografia , Traumatismos do Punho/diagnóstico por imagem , Adulto , Ossos do Carpo/diagnóstico por imagem , Tomografia Computadorizada de Feixe Cônico/métodos , Tomografia Computadorizada de Feixe Cônico/normas , Fraturas Mal-Unidas/diagnóstico por imagem , Humanos , Tomografia Computadorizada Multidetectores/métodos , Tomografia Computadorizada Multidetectores/normas , Doses de Radiação , Radiografia/métodos , Radiografia/normas , Fraturas do Rádio/diagnóstico por imagem , Reprodutibilidade dos Testes , Sensibilidade e Especificidade , Fraturas da Ulna/diagnóstico por imagem , Punho/diagnóstico por imagem
14.
Medicine (Baltimore) ; 94(31): e1231, 2015 Aug.
Artigo em Inglês | MEDLINE | ID: mdl-26252281

RESUMO

To compare the visualization of cortical fractures, cortical defects, and orthopedic screws in a dedicated extremity flat-panel computed tomography (FPCT) scanner and a multidetector computed tomography (MDCT) scanner.We used feet of European roe deer as phantoms for cortical fractures, cortical defects, and implanted orthopedic screws. FPCT and MDCT scans were performed with equivalent dose settings. Six observers rated the scans according to number of fragments, size of defects, size of defects opposite orthopedic screws, and the length of different screws. The image quality regarding depiction of the cortical bone was assessed. The gold standard (real number of fragments) was evaluated by autopsy.The correlation of reader assessment of fragments, cortical defects, and screws with the gold standard was similar for FPCT and MDCT. Three readers rated the subjective image quality of the MDCT to be higher, whereas the others showed no preferences.Although the image quality was rated higher in the MDCT than in the FPCT by 3 out of 6 observers, both modalities proved to be comparable regarding the visualization of cortical fractures, cortical defects, and orthopedic screws and of use to musculoskeletal radiology regarding fracture detection and postsurgical evaluation in our experimental setting.


Assuntos
Parafusos Ósseos , Fraturas Ósseas/diagnóstico por imagem , Ossos do Metatarso/lesões , Tomografia Computadorizada por Raios X/métodos , Animais , Cervos , Ossos do Metatarso/diagnóstico por imagem , Tomografia Computadorizada Multidetectores , Variações Dependentes do Observador , Distribuição Aleatória
15.
Med Phys ; 42(8): 4987-96, 2015 Aug.
Artigo em Inglês | MEDLINE | ID: mdl-26233224

RESUMO

PURPOSE: Lesions detected at mammography are described with a highly standardized terminology: the breast imaging-reporting and data system (BI-RADS) lexicon. Up to now, no validated semantic computer assisted classification algorithm exists to interactively link combinations of morphological descriptors from the lexicon to a probabilistic risk estimate of malignancy. The authors therefore aim at the external validation of the mammographic mass diagnosis (MMassDx) algorithm. A classification algorithm like MMassDx must perform well in a variety of clinical circumstances and in datasets that were not used to generate the algorithm in order to ultimately become accepted in clinical routine. METHODS: The MMassDx algorithm uses a naïve Bayes network and calculates post-test probabilities of malignancy based on two distinct sets of variables, (a) BI-RADS descriptors and age ("descriptor model") and (b) BI-RADS descriptors, age, and BI-RADS assessment categories ("inclusive model"). The authors evaluate both the MMassDx (descriptor) and MMassDx (inclusive) models using two large publicly available datasets of mammographic mass lesions: the digital database for screening mammography (DDSM) dataset, which contains two subsets from the same examinations-a medio-lateral oblique (MLO) view and cranio-caudal (CC) view dataset-and the mammographic mass (MM) dataset. The DDSM contains 1220 mass lesions and the MM dataset contains 961 mass lesions. The authors evaluate discriminative performance using area under the receiver-operating-characteristic curve (AUC) and compare this to the BI-RADS assessment categories alone (i.e., the clinical performance) using the DeLong method. The authors also evaluate whether assigned probabilistic risk estimates reflect the lesions' true risk of malignancy using calibration curves. RESULTS: The authors demonstrate that the MMassDx algorithms show good discriminatory performance. AUC for the MMassDx (descriptor) model in the DDSM data is 0.876/0.895 (MLO/CC view) and AUC for the MMassDx (inclusive) model in the DDSM data is 0.891/0.900 (MLO/CC view). AUC for the MMassDx (descriptor) model in the MM data is 0.862 and AUC for the MMassDx (inclusive) model in the MM data is 0.900. In all scenarios, MMassDx performs significantly better than clinical performance, P < 0.05 each. The authors furthermore demonstrate that the MMassDx algorithm systematically underestimates the risk of malignancy in the DDSM and MM datasets, especially when low probabilities of malignancy are assigned. CONCLUSIONS: The authors' results reveal that the MMassDx algorithms have good discriminatory performance but less accurate calibration when tested on two independent validation datasets. Improvement in calibration and testing in a prospective clinical population will be important steps in the pursuit of translation of these algorithms to the clinic.


Assuntos
Algoritmos , Neoplasias da Mama/diagnóstico por imagem , Neoplasias da Mama/diagnóstico , Mamografia/métodos , Interpretação de Imagem Radiográfica Assistida por Computador/métodos , Acesso à Informação , Área Sob a Curva , Teorema de Bayes , Calibragem , Bases de Dados Factuais , Diagnóstico Diferencial , Humanos , Curva ROC
16.
Eur Radiol ; 25(6): 1768-75, 2015 Jun.
Artigo em Inglês | MEDLINE | ID: mdl-25576230

RESUMO

PURPOSE: To develop and validate a decision support tool for mammographic mass lesions based on a standardized descriptor terminology (BI-RADS lexicon) to reduce variability of practice. MATERIALS AND METHODS: We used separate training data (1,276 lesions, 138 malignant) and validation data (1,177 lesions, 175 malignant). We created naïve Bayes (NB) classifiers from the training data with tenfold cross-validation. Our "inclusive model" comprised BI-RADS categories, BI-RADS descriptors, and age as predictive variables; our "descriptor model" comprised BI-RADS descriptors and age. The resulting NB classifiers were applied to the validation data. We evaluated and compared classifier performance with ROC-analysis. RESULTS: In the training data, the inclusive model yields an AUC of 0.959; the descriptor model yields an AUC of 0.910 (P < 0.001). The inclusive model is superior to the clinical performance (BI-RADS categories alone, P < 0.001); the descriptor model performs similarly. When applied to the validation data, the inclusive model yields an AUC of 0.935; the descriptor model yields an AUC of 0.876 (P < 0.001). Again, the inclusive model is superior to the clinical performance (P < 0.001); the descriptor model performs similarly. CONCLUSION: We consider our classifier a step towards a more uniform interpretation of combinations of BI-RADS descriptors. We provide our classifier at www.ebm-radiology.com/nbmm/index.html . KEY POINTS: • We provide a decision support tool for mammographic masses at www.ebm-radiology.com/nbmm/index.html . • Our tool may reduce variability of practice in BI-RADS category assignment. • A formal analysis of BI-RADS descriptors may enhance radiologists' diagnostic performance.


Assuntos
Neoplasias da Mama/diagnóstico por imagem , Sistemas de Apoio a Decisões Clínicas/normas , Mamografia/métodos , Adulto , Idoso , Teorema de Bayes , Feminino , Humanos , Internet , Mamografia/normas , Pessoa de Meia-Idade , Curva ROC , Radiologia/educação , Terminologia como Assunto , Estados Unidos
17.
Med Phys ; 41(5): 051902, 2014 May.
Artigo em Inglês | MEDLINE | ID: mdl-24784381

RESUMO

PURPOSE: The Digital Database for Screening Mammography (DDSM) is the largest publicly available resource for mammographic image analysis research and has been used extensively in the past for computer assisted diagnosis (CADx) studies. However, the database has not been searchable for a specific kind of lesion, which rendered the case selection process in past studies often times arbitrary. Therefore, the authors want to provide the complete metadata of the DDSM in an accessible format. METHODS: The authors semiautomatically transformed the data available athttp://marathon.csee.usf.edu/Mammography/Database.html into table format. The 1769 cases (914 from cancer volumes, 855 from benign volumes) comprise 1220 mass lesions (578 benign, 642 malignant) and 859 calcifications (433 benign, 426 malignant). Additionally, 694 normal cases were processed to allow for matching according to age and breast density. RESULTS: The authors provide the entire DDSM metadata (for benign, malignant, and normal cases) as tab-delimited text files[see supplementary material at http://dx.doi.org/10.1118/1.4870379E-MPHYA6-41-006405 for DDSM metadata]. CONCLUSIONS: The data provided make the case selection for future studies using the DDSM reproducible. Furthermore, it may serve as a validation dataset for CADx approaches using the BI-RADS lexicon.


Assuntos
Bases de Dados Factuais , Mamografia , Doenças Mamárias/diagnóstico por imagem , Neoplasias da Mama/diagnóstico por imagem , Calcinose/diagnóstico por imagem , Feminino , Humanos , Processamento de Imagem Assistida por Computador , Mamografia/normas
18.
Invest Radiol ; 49(7): 491-7, 2014 Jul.
Artigo em Inglês | MEDLINE | ID: mdl-24637586

RESUMO

OBJECTIVES: The aim of this study was to compare the image quality of a compact mobile flat-panel computed tomography (FPCT) capable of extremity imaging and a multidetector computed tomography (MDCT) in examinations with the same radiation dose. MATERIAL AND METHODS: Imaging with the FPCT was performed with default settings. Monte Carlo simulations were used to calculate equivalent dose settings for the 320-row MDCT. Simulations were based on and validated by dose measurements. Homogeneity, geometric distortion, artifacts, accuracy of Hounsfield values, contrast, and spatial resolution were evaluated in different imaging phantoms. Whitney-Mann U Test and Spearman ρ were used for statistical analysis. RESULTS: Homogeneity reached 2.5% for the FPCT and 0.5% for the MDCT. Hounsfield values were more accurate and contrast to noise ratios were higher for the MDCT than the FPCT (P ≤ 0.001). The MDCT depicted more rod inserts than the FPCT did. No significant geometric distortion was detected in either modality. The FPCT was more prone to artifacts around Krischner wires with a diameter of 2 mm (P = 0.05-0.001), whereas the MDCT showed a higher amount of artifacts around wires with a diameter of 0.8 mm (P ≤ 0.001). Spatial resolution was 1 lp/mm (xy), 1.7 lp/mm (z) for the FPCT and 1 lp/mm (xy), less than 1 lp/mm (z) for the MDCT. CONCLUSIONS: We compared a mobile FPCT and a 320-row MDCT by using the same radiation dose for scans. We found the spatial resolution to be higher in the FPCT. Hounsfield units were more accurate and homogeneity and contrast resolution were better in MDCT. The MDCT was also less prone to artifacts from thick Kirschner wires but showed comparably more artifacts around thin wires.


Assuntos
Tomografia Computadorizada Multidetectores/instrumentação , Interpretação de Imagem Radiográfica Assistida por Computador/instrumentação , Interpretação de Imagem Radiográfica Assistida por Computador/métodos , Ecrans Intensificadores para Raios X , Simulação por Computador , Projeto Auxiliado por Computador , Desenho de Equipamento , Análise de Falha de Equipamento , Miniaturização , Modelos Estatísticos , Imagens de Fantasmas , Reprodutibilidade dos Testes , Sensibilidade e Especificidade
19.
J Neuroradiol ; 41(4): 259-68, 2014 Oct.
Artigo em Inglês | MEDLINE | ID: mdl-24411522

RESUMO

BACKGROUND AND PURPOSE: CT angiography (CTA) is an increasingly used method for evaluation of stented vessel segments. Our aim was to compare the appearance of different carotid artery stents in vitro on CTA using different CT scanners. Of particular interest was the measurement of artificial lumen narrowing (ALN) caused by the stent material within the stented vessel segment to determine whether CTA can be used to detect in-stent restenosis. MATERIAL AND METHODS: CTA appearances of 16 carotid artery stents of different designs and sizes (4.0 to 11.0 mm) were investigated in vitro. CTA was performed using 16-, 64- and 320-row CT scanners. For each stent, artificial lumen narrowing (ALN) was calculated. RESULTS: ALN ranged from 18.77% to 59.86%. ALN in different stents differed significantly. In most stents, ALN decreased with increasing stent diameter. In all but one stents, ALN using sharp image kernels was significantly lower than ALN using medium image kernels. Considering all stents, ALN did not significantly differ using different CT scanners or imaging protocols. CONCLUSION: CTA evaluation of vessel patency after stent placement is possible, but is considerably impaired by ALN. Investigators should be informed about the method of choice for every stent and stent manufacturers should be aware of potential artifacts caused by their stents during noninvasive diagnostic methods such as CTA.


Assuntos
Angiografia/instrumentação , Prótese Vascular/efeitos adversos , Artérias Carótidas/diagnóstico por imagem , Doenças das Artérias Carótidas/diagnóstico por imagem , Oclusão de Enxerto Vascular/diagnóstico por imagem , Stents/efeitos adversos , Tomografia Computadorizada por Raios X/instrumentação , Angiografia/métodos , Artérias Carótidas/cirurgia , Doenças das Artérias Carótidas/etiologia , Desenho de Equipamento , Análise de Falha de Equipamento , Oclusão de Enxerto Vascular/etiologia , Técnicas In Vitro , Reprodutibilidade dos Testes , Sensibilidade e Especificidade , Tomografia Computadorizada por Raios X/métodos
20.
J Magn Reson Imaging ; 37(6): 1486-92, 2013 Jun.
Artigo em Inglês | MEDLINE | ID: mdl-23165951

RESUMO

PURPOSE: To show the feasibility of a stress magnetic resonance imaging (MRI) as a new method for simultaneous evaluation of the morphology and the functional integrity of the acromioclavicular joint (ACJ) ligamentous stabilizers. MATERIALS AND METHODS: MRI of four volunteers, 10 patients with acute, and six with chronic ACJ injuries was performed using a 0.25 T open MRI scanner. A 2D-proton-density and a 3D-gradient-echo sequence at rest and under 6.5 kg shoulder traction were performed. Comparative measurements of the coracoclavicular and the acromioclavicular distance were performed. Additionally, the conoid and trapezoid ligament lengths were measured with multiplanar reconstructions. RESULTS: MRI at rest correctly identified tears of the coracoclavicular and the acromioclavicular ligaments in eight patients suffering acute ACJ injuries. Stress application helped to distinguish between partial and complete coracoclavicular ligament tears in two cases. Insufficiency of the ACJ ligaments was present in all acute and chronic ACJ injuries. Stress application in chronic ACJ ligaments revealed isolated insufficiency of the conoid ligament in three cases and of the trapezoid ligament in one case. Combined insufficiency was present in two cases. CONCLUSION: Stress MRI facilitates simultaneous acquisition of morphologic and functional information of the ACJ stabilizers. In acute ACJ injuries it helps to distinguish between partial and complete ligament tears. In chronic ACJ injuries it provides functional information of the ligament regrinds.


Assuntos
Articulação Acromioclavicular/lesões , Articulação Acromioclavicular/patologia , Instabilidade Articular/patologia , Ligamentos/lesões , Ligamentos/patologia , Imagem por Ressonância Magnética/métodos , Tração/métodos , Doença Aguda , Adulto , Doença Crônica , Estudos de Viabilidade , Feminino , Humanos , Masculino , Pessoa de Meia-Idade , Reprodutibilidade dos Testes , Sensibilidade e Especificidade , Ombro
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA