Incorporating Domain Knowledge Into the Fuzzy Connectedness Framework: Application to Brain Lesion Volume Estimation in Multiple Sclerosis

Abstract

Abstract-A method for incorporating prior knowledge into the fuzzy connectedness image segmentation framework is presented. This prior knowledge is in the form of probabilistic feature distribution and feature size maps, in a standard anatomical space, and "intensity hints" selected by the user that allow for a skewed distribution of the feature intensity characteristics. The fuzzy affinity between pixels is modified to encapsulate this domain knowledge.The method was tested by using it to segment brain lesions in patients with multiple sclerosis, and the results compared to an established method for lesion outlining based on edge detection and contour following. With the fuzzy connections (FC) method, the user is required to identify each lesion with a mouse click, to provide a set of seed pixels. The algorithm then grows the features from the seeds to define the lesions as a set of objects with fuzzy connectedness above a pre-set threshold. The FC method gave improved inter-observer reproducibility of lesion volumes, and the set of pixels determined to be lesion was more consistent compared to the contouring method. The operator interaction time required to evaluate one subject was reduced from an average of 111 minutes with contouring to 16 minutes with the FC method.

Estimating Brain Lesion Volume Change in Multiple Sclerosis by Subtraction of Magnetic Resonance Images

“…We used a template that was generated as part of a previously-developed method for segmenting MS lesions. 15 Because the template is only used to find the approximate head position for the seriallyacquired images, it is likely that other generic templates 32,33 could also be used without compromising the quality of the final registration.…”

“…Details of the production of the target image are given elsewhere. 15 All image time points are registered to this template to obtain a template image in the average position of the input images. Because the template and input images may have very different contrast, Stage 1 uses mutual information as the cost function.…”

Improving Multiple Sclerosis Plaque Detection Using a Semiautomated Assistive Approach

“…9 To facilitate time-efficient, reproducible, and accurate lesion-load detection, many algorithms have been proposed for fully automated computer-assistive solutions. 3,18 These methods use different principles, including intensity-gradient features, 19 intensity thresholding, 20 intensity-histogram modeling of expected tissue classes, [21][22][23] fuzzy connectedness, 24 identification of nearest neighbors in a feature space, 25,26 or a combination of these. Methods such as Bayesian inference, expectation maximization, support-vector machines, k-nearest neighbor majority voting, and artificial neural networks are algorithmic approaches used to op- Comparing the number of study pairs improved with demyelinating lesions detected by both readers when using the newly developed assistive software to the issued radiology report.…”

Uso de volumetría y carga lesional en el seguimiento de pacientes con esclerosis múltiple: Experiencia local y revisión de la literatura

“…Estas variaciones se derivan de los diferentes fabricantes y modelos de resonadores, las diferencias en los protocolos de adquisición (espesor de corte, etc. ), también por ser realizados en diferentes fases de la enfermedad o la presencia de la patología concomitante que puede afectar el comportamiento de la intensidad de los tejidos de manera significativa (19,20) . Debido a lo anterior, algunos softwares como TOADS-CRUISE, además de utilizar los niveles de gris para detectar las lesiones de EM, utilizan atlas anatómicos para eliminar falsos positivos (21) .…”