Computing, School of
http://hdl.handle.net/1974/771
2017-09-20T00:23:51ZFractional Labelmap Representation of Anatomical Structures
http://hdl.handle.net/1974/22677
Fractional Labelmap Representation of Anatomical Structures
Sunderland, Kyle
INTRODUCTION: In medical imaging software, structures are often represented using image volumes known as labelmaps. Conversion of structures to labelmap representations results in a loss of information which impacts medical imaging algorithms, including those in radiation therapy (RT) treatment planning. RT treatment planning systems are used to optimize radiation delivery to structures, which are represented as planar contours. Errors from conversion affect metrics such as dose volume histograms (DVHs) that are the primary metric for RT plan optimization. The goal of this thesis is to develop fractional labelmaps as a structure representation that preserves more structural information and reduces conversion errors. METHODS: The effect of voxelization on treatment planning metrics was tested by comparing DVHs calculated using varying voxel sizes. An algorithm was implemented that triangulates planar contours to closed surface mesh and handles features such as branching, keyhole contours, and end-capping. Closed surfaces were converted into fractional labelmap representations, where each voxel represents occupancy between 0% and 100%. Existing segmentation methods were modified to allow fractional labelmaps to be edited directly. RESULTS: Algorithms were implemented in the open-source medical imaging platform 3D Slicer, and the SlicerRT toolkit. Voxel size was found to affect DVH accuracy for structures with small features or in regions with a high dose gradient. The planar contour to closed surface triangulation algorithm produced qualitatively good surfaces. Fractional labelmaps from closed surfaces were found to be up to 19.1% better at representing structure volume than binary labelmaps, with an average improvement of 6.8%. DVHs calculated from fractional labelmaps were found to be more accurate than DVHs calculated using binary labelmaps. Fractional segmentation methods were found to create good quality segmentations. CONCLUSION: Labelmap voxel size was found to be a contributing factor for DVH accuracy. Accurate conversion algorithms were implemented for planar contour to closed surface and closed surface to fractional labelmap conversions. When used for structure representation and DVH calculation, fractional labelmaps were more accurate than binary labelmaps at the same resolution. Fractional labelmaps were found to be an effective tool for structure representation in radiotherapy that could be expanded to other use cases.
CISC 220 Course Notes: Linux and C
http://hdl.handle.net/1974/22632
CISC 220 Course Notes: Linux and C
Lamb, David Alex Lamb
There are many Linux books on the market, but many are not suitable for a
single-term university-level textbook. Typically they provide far too much
detail, or far too little, and donâ€™t tie the material to basic concepts taught
in other undergraduate computing courses. These notes are meant to address
these issues.
Course notes for CISC 220. An overview of Linux, its file system, shell scripting, command line tools, and some aspects of C programming.
2017-08-28T00:00:00ZContext Sensitive and Secure Parser Generation for Deep Packet Inspection of Binary Protocols
http://hdl.handle.net/1974/22040
Context Sensitive and Secure Parser Generation for Deep Packet Inspection of Binary Protocols
El Shakankiry, Ali
Network protocol parsers constantly dissect a large number of network data to place into internal data structures for further processing by traffic analysis systems. Many network protocol parsers are hand-written for performance reasons, and lack the security required to run on mission-critical networks. We propose an approach that automatically generates custom protocol parsers to process network traffic to be used as part of an Intrusion Detection System. The user is provided a specification language in which they can define the protocols they need to analyse. This thesis looks at command and control/industrial control networks that are characterized by a limited number of known protocols. We present a robust, secure, and high-performing solution that deals with the issues that have only partially been addressed in this domain.
Distances Between Languages: Algorithms and Descriptional Complexity
http://hdl.handle.net/1974/22018
Distances Between Languages: Algorithms and Descriptional Complexity
Ng, Timothy
We consider the descriptional complexity of neighbourhoods of regular languages and the computational complexity of computing the distance between languages. Distance measures are defined on words to describe their similarity. These measures can be extended to languages, which are sets of words. We consider several different notions of distance measures on languages.
The neighbourhood of a language L is the set of words within some fixed distance of a word in L. We show that we can construct weighted finite automata for neighbourhoods of regular languages with respect to additive quasi-distances. We consider the state complexity of the neighbourhoods and the state complexity of approximate pattern matching. We establish a tight state complexity lower bound for the edit distance neighbourhood of a language recognized by a deterministic finite automaton.
We consider the deterministic and nondeterministic state complexity of prefix, suffix, and subword distance neighbourhoods. We establish tight deterministic state complexity bounds for the prefix and suffix distance neighbourhoods of regular languages and of finite languages. For prefix distance neighbourhoods, we consider prefix-convex subclasses of regular languages. We also establish the state complexity of suffix neighbourhoods of suffix-closed languages. We give an upper bound for the state complexity of the subword distance neighbourhood of a regular language.
Based on the edit distance algorithms of Mohri and Han and co-authors, we give algorithms for computing the prefix distance between two regular languages, a context-free language and a regular language, and two visibly pushdown languages. We extend these algorithms to compute the inner prefix distance of regular languages and visibly pushdown languages.
We give algorithms to compute the relative prefix distance of two regular languages. For a given integer k, we show that for a deterministic context-free language and a regular language or for two visibly pushdown languages, it is decidable whether their relative prefix distance is at most k. This is very close to the borderline between decidable and undecidable since, on the other hand, we show that deciding the same question for a deterministic context-free language and a visibly pushdown language is undecidable.