Lungs

Filtration in pre-processing of data was done. Firstly, with Laplacian Sharpening Image Filter to make the edges of objects more sharp was used in pre-processing of data. Secondly, the Gradient Anisotropic Diffusion makes the data smoother by getting rid of the noise, while it preserves edges. The filtering of this CT data used both, simple and non-simple filtering. The resulting processed data shows the two tumours of carcinoma in the left upper lung and right upper lung.

Direct volume rendering was used for the lungs and the Region of Interest Applied to focus specifically on chest. The Transfer function used was a pre-set for CT-Lung. The Transform function was adjusted from green colour to red colour with gradient, opacity of the outer tissue of lung diminished and the inside structure with higher density, so that the lungs are lighter pink to show the location of the tumours, which have higher density in the lungs, as bright red, along with the vessels. To increase the visibility of location of the lungs, ROI was adjusted to show a slice of the lungs, as shown in the picture below.

Location of the tumours was therefore determined with surface volume rendering, slight adjustment of transfer function and the manipulation of ROI. To visualise the location of the tumours more clearly, indirect volume rendering was used on the tumours. The result is shown in the figure to the left.

Tumours can be located easily based on their different structure seen as a cluster of darker matter which contrasts to no matter (air), the light pink tubular structures (veins) and the surface of the lung.
Indirect volume rendering with segmentation of the trachea was done to show the difference between the structure of bronchitis (small tubular structures in lungs) and the main tube that brings air to the lungs.

The location can also be shown using indirect volume rendering after segmentation of the tumours with the paint effect. The combination of indirect and direct volume rendering is a good way to show these as the opacity of the lungs was low, the nods shown with darker shade of red showed approximate location of the tumours, and further segmentation of them showed them and their position clearly. Direct volume rendering allows the location of tumours in relation to the rest of the body, outside of the lungs.

The density of the tumour is comparable to the density of tissue.

Pelvis before and after surgery

The pelvis was rendered with Direct Volume Rendering because it showed the bones right away in Volume Rendering module. No part of the body needed to be shown other than the crack in the right side of the pelvic bone, which is visible in the model. By applying ROI to the region of interest- focusing on the pelvic bone, the table and some noisy data were removed. Use of CT-AA pre-set Transfer Function in Volume Rendering module required some adjustment of colour and opacity of the noisy data. Pictures clearly show the broken bone in the right side of the pelvic bone. CT data was clear; no filtration was needed.

Direct volume rendering was used as well to see the structure. ROI was used to focus on the pelvic bone and to cut off the table. Transfer function was used to add opacity to pelvis, add opacity to the metal and to remove noise from the data.
 Indirect Volume Rendering could have been used to segment the metal fixture with nails, while deleting the staples used to stitch the woman up, since the metal had a high density (1228- 3071) compared to the rest of the body. However, doing everything with volume rendering and adjusting transfer function was a simpler way to have cleaner data which is distorted and shown very shiny in the CT scan. Metal with higher density had very low opacity and opacity of the pelvis was increased to show where it was connected my metal plates and nails. This method also eliminated extraneous variables such as the contraceptive coil.

Brain Tumor

First Indication of tumour and it having a calcification since it lights up during segmentation of bones. It has therefore similar density as bones, since the threshold was aimed at segmenting them during indirect volume rendering.
Using TresholdEffect tool in editor to segment the tumour and bone and then using Paint Eraser tool to only keep the calcified part of the tumour as it helps show the structure and volume of it, after the soft tissue of the tumour will be segmented separately. The inside part of the tumour was segmented by PaintEffect.

The PET scan provides information about the metabolic activity in the body. The picture below represents a model made by the Treshold Paint tool in Editor of the tumour based on its activity. It is very visible due to the speed of breaking down materials inside of it compared to the rest of the brain. While cancer could be spread in other parts of the brain, because it is not shown metabolising as fast, it is not detected or not fully developed.

Metabolic function (PET) can serve to make a model with indirect volume rendering and cross it over with direct volume rendering of the skull and brain from MRI scan. After modifying the colours of a pre-set MR scan in Volume Rendering module, one can see the position of the tumour in the brain.

Volume rendering of the MRI data also sows the different densities of structures in the brain. As seen in indirect volume rendering during segmentation mentioned before, direct volume rendering (as shown on pictures below) shows the calcification of the tumour. It also clearly shows the structure of it, and its irregular, asymmetrical shape.

It is located on the brink of the two hemispheres, having the larger part in the right hemisphere and the smaller half in the left hemisphere of the brain. Indirect volume rendering allows for cross-section of the data to examine the insides of the tumour and showing the difference in its densities by colour.

he main strength of CT scan is providing their clear reference to objects with high density, such as ones. In the diagrams below, skull was made with Direct Rendering to make it see through. The surrounding matter, including the insides of the brain were made from MRI scan in CT scan provided a good segmentation of the skull

MRI was useful in visualising the surrounding tissues to the brain in volumetric rendering. The opacity was lowered, and ROI applied to focus on half of the brain to show a cross-section, where the tumour is.
Pet scan was used to use indirect volume rendering to show the metabolic reaction of tumour and its location within the brain in relation to the surrounding tissues- the grey matter and the skull. All 3 scans carry their own strengths: CT- identifying high density objects, PET showing metabolic function and MRI structure and its depth. The combination of all 3 along-side of using direct (MRI and CT) and indirect (PET) volume rendering allows to build a simple model which shows both, the depth of the surrounding and the location of the tumour in relation to it.
Spatial relationship between the tumours and other organs inside the brain can be visualised via segmentation in indirect volume rendering. The use of Threshold Paint tool allowed to segment cerebellum (dark blue), grey matter (light purple) and the tumour (red) based on their density. Afterwards, their opacity was adjusted in the Models module to make the grey matter more see through to show the general location of the tumour. Cerebellum and the tumour may seem similar in structure from the left side of the brain, however the front view shows cerebellum as a healthy symmetrical organ while the tumour is irregular and asymmetrical.