David Oczka

Evaluation Application for Tracking and Statistical Analysis of Patient Data from Hospital Real Time Location System

In the recent time, an importance of the Real-time Locating systems (RTLS) in the hospital environment is increasingly important. Such systems are essentially utilized for the localization of either unmovable objects, or patients in the hospital environment. It is supposed that patients who underwent some trauma and surgery may have influenced their cognitive functions in a sense of lost concentration, and orientation is space. Such disorders may endanger the patient’s health. In cooperation with the Trauma Center of University Hospital we have designed a localization system constituting the patients IR tags which are detected by the IR anchors placed in every hospital room. Data from such system represents the patient’s movement and localization completed by sophisticated system of alarms. In this context we have developed the SW application which is connected with the RTLS database where the patient’s localization data are stored. This application is fully integrated into the RTLS system. The SW application evaluates time data represents spent time in every room. The patient’s data are consecutively refreshed and updated in the SW application to receive current information.

Fuzzy Segmentation Driven by Modified ABC Algorithm Using Cartilage Features Completed by Spatial Aggregation: Modeling of Early Cartilage Loss

In a clinical practice of the orthopedics, the articular cartilage assessment is one of the major clinical procedures serving as a predictor of the future cartilage loss development. The early stage of the cartilage osteoarthritis is badly observable from the native MR records due to weak contrast between the physiological cartilage and the osteoarthritic spots. Therefore, the cartilage regional modeling would reliably differentiate the physiological cartilage from the early cartilage deterioration, and can serve as an effective clinical tool. In a comparison with the conventional segmentation methods based on the hard thresholding, the soft fuzzy thresholding based on the histogram separation into segmentation classes via the fuzzy triangular functions represents a sensitive regional segmentation even in the non-contrast environment. We have proposed the soft segmentation where the fuzzy sets are driven by the ABC genetic algorithm to optimal fuzzy class’s distribution regarding the knee tissues characteristics. Consequently, the spatial aggregation is employed to taking advantage the spatial dependences which allows for modification the original fuzzy membership function. This procedure ensures the correct pixel’s classification especially when the noise pixels are present. Such multiregional segmentation makes a mathematical model well separating the physiological cartilage from the early osteoarthritic spots which are highlighted in the model.

Detection and Dynamical Tracking of Temperature Facial Distribution Caused by Alcohol Intoxication with Using of Modified OTSU Regional Segmentation

The alcohol intoxication is a significant issue influencing many social areas. There are commonly utilized standards for the alcohol measurement like is the breath and blood analyzers. Nevertheless such methods require direct contact, and agreement of tested person. New potential trend of the alcohol measurement reflect a fact that the alcohol cause thermal facial changes which can be observable from the IR image records. In this regard, we have performed analysis dealing with the alcohol effect modeling via multiregional segmentation algorithm based on the modified OTSU method. This method autonomously extracts areas exhibiting significant temperature variations whilst person drinking. By this way we have tracked the alcohol effect over the time which corresponded with the gradual alcohol consumption. Significant part of the analysis deals with comparison and verification of the facial temperature distribution caused by the alcohol against breath analysis.

Multiregional Soft Segmentation Driven by Modified ABC Algorithm and Completed by Spatial Aggregation: Volumetric, Spatial Modelling and Features Extraction of Articular Cartilage Early Loss

In a clinical practise of the orthopaedics and medical imaging systems, the early cartilage loss, and cartilage lesions are challenging tasks. Due to an insufficient contrast, such pathologies are badly observable by naked eyes. Furthermore, objectification and quantification of those pathological findings are usually only subjectively estimated without the SW support. We propose a multiregional segmentation model based on the histogram classification with using of a sequence of triangular fuzzy functions where each such function represents specific knee area. To ensure a robustness of the model, respective fuzzy class location is driven by the ABC (Artificial Bee Colony) genetic algorithm respecting statistical features of the physiological cartilage. In the second step of the algorithm, a spatial aggregation is applied in order to consider spatial relationships in every region to prevent the image noise deterioration. Such multiregional segmentation model allows for an extraction of significant features well corresponding with the early cartilage loss like is the cartilage volume.

Analysis and Modelling of Heel Bone Fracture with Using of Active Contour Driven by Gaussian Energy and Features Extraction

In a clinical practice of the traumatology, the heel bone fracture is a routine procedure which is commonly diagnostic via the X-ray imaging. A reliable indicator of the particular heel bone fracture is the periosteal callus having dynamical geometrical features over the time. Unfortunately, the clinical imaging techniques allow for only the basic visualization without further post processing. Thus, such kind of the diagnosis may be affected by the subjective error depending on the particular experience of the clinical physician. In cooperation with the clinical experts, we have developed a mathematical model which is based on the active contours utilizing the Gaussian fitting energy to autonomously specify an area of the periosteal callus. Consequently, the periosteal callus features are figured out to objectively measure and track the heel bone fracture.

Modeling and objectification of blood vessel calcification with using of multiregional segmentation

AbstractIn a clinical practice of the angiography, the blood vessel analysis is substantially important mainly in a sense of an objectification and modeling of the pathological spots such as the blood vessel calcifications. An amount of the calcification is commonly just estimated by naked eyes; therefore, the automatic modeling may be beneficial in a context of an extraction of the blood vessel features well representing a level of the blood vessel deterioration. In this work, we have proposed a fully automatic software environment (BloodVessCalc) for processing the blood vessel images acquired by the CT (computer tomography). The main function of the SW is the multiregional image segmentation allowing for an extraction of the physiological blood vessel location from the calcification spots. This model offers the calcium score calculation in a form of amount of the calcification. In the last part of our analysis, the predictive intervals of the average value and median for calcium score are calculated.

Design and Implementation of an Algorithm for System of Exposure Limit in Radiology

This thesis describes the implementation of application for effective planning process of CyberKnife system by processing CT images. CyberKnife system, which focuses therapeutic target using two X-ray tubes and takes images of the patient at an angle of 45°, allows localization of lung irradiated tumor bearing based on the difference in density between the bearing and the surrounding tissue. Some tumor bearings are difficult to localize due to summation and overlay of monitoring area with other high densities structures. The application is designed as a client created in C#, connected to the COM server of computing system MATLAB, which provides most of the calculations. The application was tested on a single core chip system and speed of acquisition and processing is in average around 1 s. The application was developed, implemented and now successfully tested on the Oncology Clinic at FN Ostrava.