RESUMEN
A new digital processing method was developed to improve the specific diagnosis of small (up to 3 cm) solitary pulmonary nodules on lung tomograms. The proposed method gives advanced imaging of diagnostically important details and structures of the nodule. It helps the physician easily identify known diagnostic features and discover some additional ones that are useful for reliable verification of cancer, tuberculoma, and hamartoma. X-Ray morphological comparison confirmed that all details, displayed on the processed tomogram, corresponded to morphological structures on postsurgical histotopograms. The method can give additional information to data from conventional and computed tomography, especially in the case of small radiographically indeterminate pulmonary opacities.
Asunto(s)
Intensificación de Imagen Radiográfica , Nódulo Pulmonar Solitario/diagnóstico por imagen , Diagnóstico Diferencial , Técnicas de Diagnóstico Quirúrgico , Hamartoma/diagnóstico por imagen , Hamartoma/patología , Humanos , Procesamiento de Imagen Asistido por Computador , Pulmón/anomalías , Pulmón/diagnóstico por imagen , Pulmón/patología , Neoplasias Pulmonares/diagnóstico por imagen , Neoplasias Pulmonares/patología , Nódulo Pulmonar Solitario/patología , Tomografía por Rayos X , Tomografía Computarizada por Rayos X , Tuberculoma/diagnóstico por imagen , Tuberculoma/patología , Tuberculosis Pulmonar/diagnóstico por imagen , Tuberculosis Pulmonar/patologíaRESUMEN
The authors propose and implement a roentgenologist's IBM PC/AT-compatible computer-aided automatic working place designed for the digital processing and analysis of medical introscopic images. The two-levelled architecture of APM-BP software allows one to optimize the speed of computations and to apply as computer technologies the know-how and experience gained in solving specific problems in the analysis of informative features of medical images. A procedure for the digital processing of linear pulmonary tomograms by using the designed working place in order to solve the problems of preclinical diagnosis of peripheral lung cancer is described.
Asunto(s)
Procesamiento Automatizado de Datos , Procesamiento de Imagen Asistido por Computador , Tecnología Radiológica/métodos , Lugar de Trabajo , Anciano , Femenino , Humanos , Neoplasias Pulmonares/diagnóstico , MatemáticaRESUMEN
Additional information on an x-ray picture of peribronchial changes in tumorous and non-tumorous lesions of the lungs were obtained in 54 patients (40 with bronchogenic, mostly central cancer and 14 with chronic inflammatory diseases) in order to study roentgenological symptomatology before and after computerized processing of tomographic images using a method of linear filtration. Image processing with the measurement of densitometric indicators provides additional information for the estimation of bronchial affection by tumor infiltration. The same method can be of value for the detection of early signs of lung cancer in patients with peribronchial changes of obscure etiology.
Asunto(s)
Procesamiento de Imagen Asistido por Computador , Neoplasias Pulmonares/diagnóstico por imagen , Tomografía Computarizada por Rayos X , Humanos , Masculino , Persona de Mediana EdadRESUMEN
The digital methods of processing the tomograms and the methods of measurements in accordance with the processed tomograms have been elaborated to determine the quantitative characteristics, enabling one to solve the problem of early preclinical differential diagnosis of peribronchial lesions in the lungs. A linear filter discriminating the diagnostically important structures has been designed and the quantitative characteristics and their threshold values have been established.
Asunto(s)
Enfermedades Pulmonares/diagnóstico por imagen , Intensificación de Imagen Radiográfica , Tomografía por Rayos X , Bronquitis/diagnóstico por imagen , Diagnóstico Diferencial , Humanos , Neoplasias Pulmonares/diagnóstico por imagen , Modelos Biológicos , Lesiones Precancerosas/diagnóstico por imagenRESUMEN
X-ray and morphological signs were compared in 50 patients with histologically verified bronchioloalveolar cancer (BAC). Particular attention was drawn to a study of an x-ray picture after computer processing of a tomographic image using a special technique. The most characteristic x-ray signs of BAC were subpleural localization, the presence of a fragmented scar in a tumor, and ill defined outlines. The peculiarities of an x-ray picture of small-size BAC were determined by the morphological structure of a tumor, a degree of its differentiation that could be detected in most cases only after image processing by the linear filtration technique.