Adam Goode

A Boosting Framework for Visuality-Preserving Distance Metric Learning and Its Application to Medical Image Retrieval

Similarity measurement is a critical component in content-based image retrieval systems, and learning a good distance metric can significantly improve retrieval performance. However, despite extensive study, there are several major shortcomings with the existing approaches for distance metric learning that can significantly affect their application to medical image retrieval. In particular, ldquosimilarityrdquo can mean very different things in image retrieval: resemblance in visual appearance (e.g., two images that look like one another) or similarity in semantic annotation (e.g., two images of tumors that look quite different yet are both malignant). Current approaches for distance metric learning typically address only one goal without consideration of the other. This is problematic for medical image retrieval where the goal is to assist doctors in decision making. In these applications, given a query image, the goal is to retrieve similar images from a reference library whose semantic annotations could provide the medical professional with greater insight into the possible interpretations of the query image. If the system were to retrieve images that did not look like the query, then users would be less likely to trust the system; on the other hand, retrieving images that appear superficially similar to the query but are semantically unrelated is undesirable because that could lead users toward an incorrect diagnosis. Hence, learning a distance metric that preserves both visual resemblance and semantic similarity is important. We emphasize that, although our study is focused on medical image retrieval, the problem addressed in this work is critical to many image retrieval systems. We present a boosting framework for distance metric learning that aims to preserve both visual and semantic similarities. The boosting framework first learns a binary representation using side information, in the form of labeled pairs, and then computes the distance as a weighted Hamming distance using the learned binary representation. A boosting algorithm is presented to efficiently learn the distance function. We evaluate the proposed algorithm on a mammographic image reference library with an interactive search-assisted decision support (ISADS) system and on the medical image data set from ImageCLEF. Our results show that the boosting framework compares favorably to state-of-the-art approaches for distance metric learning in retrieval accuracy, with much lower computational cost. Additional evaluation with the COREL collection shows that our algorithm works well for regular image data sets.

GRACE: an autonomous robot for the AAAI Robot challenge

In an attempt to solve as much of the AAAI Robot Challenge as possible, five research institutions representing academia, industry, and government integrated their research into a single robot named GRACE. This article describes this first-year effort by the GRACE team, including not only the various techniques each participant brought to GRACE but also the difficult integration effort itself.

Competition and representation during memory retrieval: Roles of the prefrontal cortex and the posterior parietal cortex

In this functional-MRI study we examined the hypothesis that the prefrontal cortex responds differently to the extent of competition during retrieval, whereas the parietal cortex is responsible for problem representation that should not be directly related to the competition. Participants mastered arbitrary person–location pairs, and their recognition memory was tested in a functional-MRI session. The pairs were constructed such that a person was associated with one, two, or three different locations and vice versa. The recognition time increased with the number of associations, reflecting increased competition. A confirmatory analysis of imaging data with prespecified prefrontal and parietal regions showed that, although both regions were highly involved during memory retrieval, only the prefrontal region responded to the levels of competition. This result was consistent with predictions of an information-processing model as well as with an exploratory identification of regions of interest.

Predicting the practice effects on the blood oxygenation level-dependent (BOLD) function of fMRI in a symbolic manipulation task.

Based on adaptive control of thought-rational (ACT-R), a cognitive architecture for cognitive modeling, researchers have developed an information-processing model to predict the blood oxygenation level-dependent (BOLD) response of functional MRI in symbol manipulation tasks. As an extension of this research, the current event-related functional MRI study investigates the effect of relatively extensive practice on the activation patterns of related brain regions. The task involved performing transformations on equations in an artificial algebra system. This paper shows that the base-level activation learning in the ACT-R theory can predict the change of the BOLD response in practice in a left prefrontal region reflecting retrieval of information. In contrast, practice has relatively little effect on the form of BOLD response in the parietal region reflecting imagined transformations to the equation or the motor region reflecting manual programming.

An information-processing model of three cortical regions: evidence in episodic memory retrieval

ACT-R (Anderson, J.R., et al., 2003. An information-processing model of the BOLD response in symbol manipulation tasks. Psychon. Bull. Rev. 10, 241-261) relates the inferior dorso-lateral prefrontal cortex to a retrieval buffer that holds information retrieved from memory and the posterior parietal cortex to an imaginal buffer that holds problem representations. Because the number of changes in a problem representation is not necessarily correlated with retrieval difficulties, it is possible to dissociate prefrontal-parietal activations. In two fMRI experiments, we examined this dissociation using the fan effect paradigm. Experiment 1 compared a recognition task, in which representation requirement remains the same regardless of retrieval difficulty, with a recall task, in which both representation and retrieval loads increase with retrieval difficulty. In the recognition task, the prefrontal activation revealed a fan effect but not the parietal activation. In the recall task, both regions revealed fan effects. In Experiment 2, we compared visually presented stimuli and aurally presented stimuli using the recognition task. While only the prefrontal region revealed the fan effect, the activation patterns in the prefrontal and the parietal region did not differ by stimulus presentation modality. In general, these results provide support for the prefrontal-parietal dissociation in terms of retrieval and representation and the modality-independent nature of the information processed by these regions. Using ACT-R, we also provide computational models that explain patterns of fMRI responses in these two areas during recognition and recall.

OpenSlide: A vendor-neutral software foundation for digital pathology

Although widely touted as a replacement for glass slides and microscopes in pathology, digital slides present major challenges in data storage, transmission, processing and interoperability. Since no universal data format is in widespread use for these images today, each vendor defines its own proprietary data formats, analysis tools, viewers and software libraries. This creates issues not only for pathologists, but also for interoperability. In this paper, we present the design and implementation of OpenSlide, a vendor-neutral C library for reading and manipulating digital slides of diverse vendor formats. The library is extensible and easily interfaced to various programming languages. An application written to the OpenSlide interface can transparently handle multiple vendor formats. OpenSlide is in use today by many academic and industrial organizations world-wide, including many research sites in the United States that are funded by the National Institutes of Health.

Behavioral equivalence, but not neural equivalence—neural evidence of alternative strategies in mathematical thinking

In a functional magnetic resonance imaging study, we investigated how people solve mathematically equivalent problems presented in two alternative formats: verbal, story format or symbolic, equation format. Although representation format had no effect on behavior, anterior prefrontal activation was greater in the story condition and posterior parietal activation was greater in the equation condition. These results show that there exist alternative neural pathways that implement different and yet equally efficient problem-solving strategies.

Differential Fan Effect and Attentional Focus

As people study more facts about a concept, it takes longer to retrieve a particular fact about that concept. This fan effect (Anderson, 1974) has been attributed to competition among associations to a concept. Alternatively, the mental-model theory (Radvansky & Zacks, 1991) suggests that the fan effect disappears when the related concepts are organized into a single mental model. In the present study, attentional focus was manipulated to affect the mental model to be constructed. One group of participants focused on the person dimension of personlocation pairs, whereas the other group focused on the location dimension. The result showed that the fan effect with the focused dimension was greater than the fan effect with the nonfocused dimension, which is contrary to the mental-model theory. The number of associations with a concept is indeed crucial during retrieval, and the importance of the information seems to be accentuated with attentional focus.

The Case for Content Search of VM Clouds

The success of cloud computing can lead to large, centralized collections of virtual machine~(VM) images. The ability to interactively search these VM images at a high semantic level emerges as an important capability. This paper examines the opportunities and challenges in creating such a search capability, and presents early evidence of its feasibility.

Distributed online anomaly detection in high-content screening

This paper presents an automated, online approach to anomaly detection in high-content screening assays for pharmaceutical research. Online detection of anomalies is attractive because it offers the possibility of immediate corrective action, early termination, and redesign of assays that may require many hours or days to execute. The proposed approach employs assay-specific image processing within an assay-independent framework for distributed control, machine learning, and anomaly reporting. Specifically, we exploit coarse-grained parallelism to distribute image processing over several computing nodes while efficiently aggregating sufficient statistics across nodes. This architecture also allows us to easily handle geographically-distributed data sources. Our results from two applications, adipocyte quantitation and neurite growth estimation, confirm that this online approach to anomaly detection is feasible, efficient, and accurate.