Ana Sánchez Martín, Juan A. Juanes, Andres Framiñan, Patricia Carreno-Moran, Paulino García Benedito, Begoña García Castaño.
Hospital Universitario de Móstoles
Universidad de Salamanca
Hospital Universitario de Salamanca
Technological Environments for Image Processing in Medical Training Settings
1. Entornos Tecnológicos para el Procesamiento
Médico de la Imagen con Fines Formativos
Ana Sánchez Martín1
Juan A. Juanes Méndez2
Andrés Framiñán de Miguel3
Patricia Carreño Moran3
Paulino García Benedito1
Begoña García Castaño1
1 Hospital Universitario de Móstoles
2 Universidad de Salamanca
2. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
Importante mejora de la imagen radiológica en base a la física y la informática.
Los nuevos software e interfaz del usuario de las distintas casas comerciales, cada vez más
simples e intuitivos, han conseguido un desarrollo y mejora de la imagen radiológica
facilitando su posproceso y por lo tanto la comprensión , la interpretación y el aprendizaje
radiológico tanto de la anatomía como de la patología.
Visor Syngo de imágenes DICOM de Siemens con la aplicación InSpace 4D.
Reconstrucciones volumétricas con la eliminación de estructuras no deseadas.
3. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
Adquisición:
-20 pacientes sanos entre 20 y 40 años
en el hospital Universitario de Móstoles.
Consentimiento informado.
-Resonancia Phillips de 1,5 Teslas:
Secuencias
-3D T1 TSE ISOvoxel
-T1 IR ISO SENSE
-FLAIR and T2 TSE
4. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
Adquisición:
-20 pacientes sanos entre 20 y 40 años
en el hospital Universitario de Móstoles.
Consentimiento informado.
-TC Siemens Somaton Sensation de
40 coronas:
TC helicoidal de abdomen en fase venosa
5. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
Procesamiento de la imagen:
Visor Syngo de Siemens en la estación LEONARDO
6. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
Procesamiento de la imagen:
Aplicación InSpace 4D/Áreas de tarjeta de tarea. Interfaz del Usuario.
(1) Barra de menú principal
(2) Iconos de modo
(3) Pila superior de tarjeta de subtareas
(4) Área de control de clasificación
(5) Ventana de imagen
(6) Pila inferior de tarjeta de subtareas
(7) Iconos de entrada/salida
(8) Barra de estado para los mensajes del sistema
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
7. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Procesamiento de la imagen:
Aplicación InSpace 4D/vision MPR y volumétrica.
VTR (volume rendering technique)
MIP (maximum intensity projection)
MinIP (minimum intensity projection)
MPR (multiplanar reconstruction).
Formativos
8. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
-Eliminación de la mesa.
-Cortar y manipular un volumen (planos de corte).
-Extracción de un volumen.
-Eliminación de estructuras óseas.
9. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
-Eliminación de la mesa:
Optimizado para los accesorios de Siemens.
Manual o automático.
Fundamental para la realización de reconstrucciones
volumétricas y evitar malinterpretaciones.
12. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Segmentar o eliminar partes de un volumen:
Formativos
-Eliminación de la mesa.
-Cortar y manipular un volumen (planos de corte).
-Extracción de un volumen.
-Eliminación de estructuras óseas.
13. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
-Cortar y manipular un volumen (planos de corte).
Un plano de corte es un MPR (isovoxel) de un volumen que
podemos arrastrar rápidamente en un determinado plano.
Los planos de corte muestran y ocultan diferentes partes de
un volumen al desplazarlos.
Se pueden manejar a la vez 4 ó 6 planos de corte.
14. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
-Cortar y manipular un volumen (planos de corte).
Mas eficaces y rápidos que manipular un volumen al realizar
movimiento, rotación y zoom de las distintas estructuras. No
es posible cambiar la distancia entre ellos mientras están
rotando.
Así, añadir o eliminar de forma interactiva diferentes
estructuras utilizando los planos de corte nos permite una
mejor compresión 3D y una adecuada visualización de la
localización y relación anatómica de las mismas.
15. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
Cortar y manipular un volumen (planos de corte).
16. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
Cortar y manipular un volumen (planos de corte).
17. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
Cortar y manipular un volumen (planos de corte).
19. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
-Eliminación de la mesa.
-Cortar y manipular un volumen (planos de corte).
-Extracción de un volumen.
-Eliminación de estructuras óseas.
20. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
-Extracción de un volumen.
Uso de ROIs (Regions Of Interest) y VOIs (Volume Of Interest).
Mano alzada o polígonos dibujando ROIs para extraer VOIs.
Consume mayor cantidad de tiempo y no se consigue una
compresión 3D tan eficaz como los planos de corte.
22. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
-Eliminación de la mesa.
-Cortar y manipular un volumen (planos de corte).
-Extracción de un volumen.
-Eliminación de estructuras óseas.
23. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Eliminación de estructuras óseas.
Formativos
24. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
Eliminación de estructuras óseas.
27. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
-La aplicación de Syngo InSpace 4D representa una
importante mejora en el manejo de la imagen médica
3D con una interfaz de usuario que facilita el trabajo de
forma intuitiva y sin interrupciones.
-Permite el procesamiento de todas las modalidades de
imagen (radiología simple, TC, RM y medicina nuclear)
pudiendo procesarse grandes cantidades de datos en
tiempo real en el estación LEONARDO.
28. Introducción Metodología Resultados Conclusión
Entornos Tecnológicos para el Procesamiento Médico de la Imagen con Fines
Formativos
-Técnicas de renderización: MIP, MiniIP, VTR, MPR.
Facilitan la interpretación de la imagen 3D.
-Con la eliminación de las estructuras no deseadas se
consigue una mejor visualización y comprensión
anatómica, lo que no solo facilita en estudio radiológico
sino también el aprendizaje de la anatomía.
Notas del editor
Radiology image has improved significantly since its beginning due to new softwares and user interfaces, which are more intuitive and simple than before.
Post-processing image is considered a useful tool for a proper understanding of anatomy and pathology in hospital day to day basis, specifically being a helpful device for radiologists, clinicians and medical training.
Syngo DICOM viewer together with InSpace 4D application is currently one of the most sophisticated program in the market. InSpace 4D provides a new feature which can remove certain anatomical regions or non-critical structures. Hence, we have an easy 3D view of a given anatomical structure, its location and its relationship with any other one. This helps to a better understanding of anatomical complexity inherent to any part of the human body and, mainly neuroimaging. This article ́s aim is to describe how to process a volume in order to obtain a 3D visualization and hence a better volumetric anatomy understanding.
The images of this study were obtained from 20 Healthy patients between 20 and 40 years old at Mostoles University Hospital. All patients were provided and agreed with the informed consent to participate in the study, which is approved by the World Medical Association Declaration of Helsinki on ethical principles, allowing medical research involving humans. A head MRI on a 1.5 Tesla Phillips equipment with volumetric adquisition of 3D T1 TSE , T1 IR ISO SENSE, FLAIR and T2 TSE secuences and abdomen CT at venous phase using a Siemens 40 CT were carried out.
The images of this study were obtained from 20 Healthy patients between 20 and 40 years old at Mostoles University Hospital. All patients were provided and agreed with the informed consent to participate in the study, which is approved by the World Medical Association Declaration of Helsinki on ethical principles, allowing medical research involving humans. A head MRI on a 1.5 Tesla Phillips equipment with volumetric adquisition of 3D T1 TSE , T1 IR ISO SENSE, FLAIR and T2 TSE secuences and abdomen CT at venous phase using a Siemens 40 CT were carried out.
Images were taken by a Siemens SOMATOM Sensation 40 CT and handled with a Siemens Syngo Viewer at a LEONARDO station using InSpace 4D application. (see Figure 1 and 2). Inspace is the Syngo interactive module volume rendering in real time. It interactively shows patient ́s data sets that are used for diagnosis, surgical and treatment planning, subsequent studies and obviously this information can be used for educational purposes, providing a better 3D anatomical understanding. It displays 3D volumes renderings using VTR (volume rendering technique), MIP (maximum intensity projection), MinIP (minimum intensity projection) or MPR (multiplanar reconstruction). Colored VTR helps to identify different kinds of tissues. 3D data set often includes anatomy structures or parts of little interest. InSpace 4D provides with several ways of segmenting or volume parts removing, as for example table removal, cutting planes use and bone segment extraction.
InSpace es el nódulo interactivo de Syngo que renderiza volúmenes en tiempo real. Muestra de forma interactiva los datos del paciente que van a ser usados para el diagnóstico, la planificación quirúrgica. L
"Table removal" function is optimized for Siemens accessories and cannot be used with any other sort of mattress. Hence, table removal can be either done manually or automatically, InSpace 4D works either way. “Table removal” function allows us to create a volume hiding the table and any other flat surfaces. Clicking on “table removal” icon, located in the tool menu, table will be hidden. Having the table off the studies is essential and facilitates the volumetric reconstruction, and avoids overlaps and misinterpretations of the table as an anatomical structure. Hence, "Table removal" function facilitates radiological and anatomical interpretation and, therefore, the anatomy volumetrically understanding. Locate and remove patient ́s table from the CT image is important and desirable for a better image rendering. Literature about table removal is scarce. Some methods have been reported only recently [3, 4] and researchers and vendors alike have included them in software applications [5]. These methods can be broadly classified as manual or automatic. In a manual method, the user draws a contour or plane to localize the table [6], which is time consuming and lacks reproducibility. Automatic algorithms can be template based [3] or connected component analysis based [4, 7]. The template method requires a preacquired template, which limits its applicability. The connected component analysis methods sometime fail to isolate the table, particularly when the patient body touches the table edges. Development of a robust and automatic method for identification and removal of any table, regardless of its characteristics, is today a challenge. Table ́s shape identification methods have been developed by different commercial brands, as is the case of Siemens with InSpace 4D application. Recently it has been reported an automatic table identification and removal method which has been shown to be independent of table characteristics. It is based on a simple observation that, in sagittal planes, the top of the table essentially forms a vertical line.
Cut planes show different parts of a volume which can be hidden away dragging over them. Cut planes are more accurate than capturing a volume and means the fastest way to move through it. The maximum number of cut planes that can be used at once are either four (Volume Pro on) or six (Volume Pro off) planes depending on the selected mode. The main constrain for a selected portion of volume located between two cut planes is that the mentioned two planes will always rotate parallel and synchronized, not being able to change the distance between them while they are rotating. A cut plane is defined as a MPR view (1 isovoxel wide) of a volume that can be quickly dragged over in a particular plane. Handling volumes quiet intuitive and logical using InSpace4D cut planes. In order to get an optimal view the image can be rotated, moved and in and out zoomed. "Manipulate all objects" is an option to work with the image using the right mouse button or clicking on the different icons on the toolbar. Hence, adding and removing interactively different anatomical structures using cut planes is an easy way to display, study and understand the anatomy from a radiology point of view.
Using cut planes with 3D images allows views in any orientation and depth, removing unwanted structures and showing the ones within the volume that otherwise would be hidden. Moreover, we use several cut planes at the same time for an optimal structure view from different angles and directions.
Cut planes show different parts of a volume which can be hidden away dragging over them. Cut planes are more accurate than capturing a volume and means the fastest way to move through it. The maximum number of cut planes that can be used at once are either four (Volume Pro on) or six (Volume Pro off) planes depending on the selected mode. The main constrain for a selected portion of volume located between two cut planes is that the mentioned two planes will always rotate parallel and synchronized, not being able to change the distance between them while they are rotating. A cut plane is defined as a MPR view (1 isovoxel wide) of a volume that can be quickly dragged over in a particular plane. Handling volumes quiet intuitive and logical using InSpace4D cut planes. In order to get an optimal view the image can be rotated, moved and in and out zoomed. "Manipulate all objects" is an option to work with the image using the right mouse button or clicking on the different icons on the toolbar. Hence, adding and removing interactively different anatomical structures using cut planes is an easy way to display, study and understand the anatomy from a radiology point of view.
Using cut planes with 3D images allows views in any orientation and depth, removing unwanted structures and showing the ones within the volume that otherwise would be hidden. Moreover, we use several cut planes at the same time for an optimal structure view from different angles and directions.
Drawning ROIs (Regions Of Interest) is a more complex way to remove a certain volume. ROIs can be created either by freehand (dragging the mouse) or drafting polygons (single mouse click). There is also an option to switch straight lines into curved ones.
-Using ROIs in order to extract VOIs (Volume Of Interest):
Clicking on “VOI extraction” icon in the lower pillar card subtasks will open a VOI dialog box. Once we are in this dialog box, a ROI can be created by clicking on button "new". We will have the final ROI by clicking on the last dot.
ROIs can be edited by placing the pointer inside the area. As the pointer moves into the area it switches to a cross shape and then we can either change the shape of the ROI selecting a fixed point on the contour line and moving it or select and move the whole ROI to the desired position by dragging and dropping.
There are three different options working with VOI; "recover", "activate" and "undo". "Recover" function restores the original volume. Unchecking "activate" restores the initial volume. Finally selecting "undo" the effect of the most recent ROI is removed.
-Remove a VOI:
We can either select the VOI inside of the contour line (“hold inside” button) or retain the external area of the contour line (“keep off” button).
-Delete a ROI:
Individual fix points of a ROI contour line or the whole ROI, can be deleted by hitting the delete button on the keyboard.
ROI and VOI provide us with a concrete 3D view of a volume which is of our interest, ensuring a better radiology and anatomy understanding.
Drawning ROIs (Regions Of Interest) is a more complex way to remove a certain volume. ROIs can be created either by freehand (dragging the mouse) or drafting polygons (single mouse click). There is also an option to switch straight lines into curved ones.
-Using ROIs in order to extract VOIs (Volume Of Interest):
Clicking on “VOI extraction” icon in the lower pillar card subtasks will open a VOI dialog box. Once we are in this dialog box, a ROI can be created by clicking on button "new". We will have the final ROI by clicking on the last dot.
ROIs can be edited by placing the pointer inside the area. As the pointer moves into the area it switches to a cross shape and then we can either change the shape of the ROI selecting a fixed point on the contour line and moving it or select and move the whole ROI to the desired position by dragging and dropping.
There are three different options working with VOI; "recover", "activate" and "undo". "Recover" function restores the original volume. Unchecking "activate" restores the initial volume. Finally selecting "undo" the effect of the most recent ROI is removed.
-Remove a VOI:
We can either select the VOI inside of the contour line (“hold inside” button) or retain the external area of the contour line (“keep off” button).
-Delete a ROI:
Individual fix points of a ROI contour line or the whole ROI, can be deleted by hitting the delete button on the keyboard.
ROI and VOI provide us with a concrete 3D view of a volume which is of our interest, ensuring a better radiology and anatomy understanding.
Drawning ROIs (Regions Of Interest) is a more complex way to remove a certain volume. ROIs can be created either by freehand (dragging the mouse) or drafting polygons (single mouse click). There is also an option to switch straight lines into curved ones.
-Using ROIs in order to extract VOIs (Volume Of Interest):
Clicking on “VOI extraction” icon in the lower pillar card subtasks will open a VOI dialog box. Once we are in this dialog box, a ROI can be created by clicking on button "new". We will have the final ROI by clicking on the last dot.
ROIs can be edited by placing the pointer inside the area. As the pointer moves into the area it switches to a cross shape and then we can either change the shape of the ROI selecting a fixed point on the contour line and moving it or select and move the whole ROI to the desired position by dragging and dropping.
There are three different options working with VOI; "recover", "activate" and "undo". "Recover" function restores the original volume. Unchecking "activate" restores the initial volume. Finally selecting "undo" the effect of the most recent ROI is removed.
-Remove a VOI:
We can either select the VOI inside of the contour line (“hold inside” button) or retain the external area of the contour line (“keep off” button).
-Delete a ROI:
Individual fix points of a ROI contour line or the whole ROI, can be deleted by hitting the delete button on the keyboard.
ROI and VOI provide us with a concrete 3D view of a volume which is of our interest, ensuring a better radiology and anatomy understanding.
Another way to process a volume is using ROIs and VOIs. Unlike the use of cut planes, the handling of VOIs and ROIs is more time consuming involving greater amounts of time in image processing (between 30 and 60 minutes per case), making use of cut planes a preferred option
Sometimes is useful to mask certain structures of a volume for a better diagnosis. At the bottom of the card stack subtasks select the "delete bone" icon and a new window pops up with 3 further steps to take.
Step 1 allows us choosing among body, head or fracture bones. For higher accuracy, Hounsfield Units (HU) threshold predefined values can be adjusted depending on the selected volume.
Step 2 helps us to refine the image we have just obtained using “markers” for a correct segmentation.
-"Mark" defines as bone a mistakenly structure not previously considered as a bone.
-"Unmark" defines as no-bone a mistakenly structure previously considered as a bone.
-"Change" just switches a structure definition as "bone" or "no- bone".
Once we have all the bone structures properly identify we move forward to step 3, where we can choose between different display options:
-Marked: displays just bone structures.
-Highlight: displays the colored mask.
-Not marked: hides bone structures away.-Both: displays both bone and no-bone structures.
Bone Removal is another InSpace4D application, which recognizes the bone structures by setting a certain threshold UH. Regarding studies without intravenous contrast and venous phase ones, the recognition and differentiation between bone and vascular structures is optimal based on the UH differentiation, hence we can differentiate properly both structures. On the other hand, angio CT vascular studies, where attenuation of both structures is pretty similar and they are often located close one to each other, is necessary for a proper differentiation using dual energy CT (DECT). This device acquires data with different X-ray spectrum and characterizes different substances (iodine, calcium, uric acid...), so unwanted structures can be removed, shortening the post-processing time and facilitating the interpretation of the acquired images.