HihO: accelerating artificial intelligence interpretability for medical imaging in IoT applications using hierarchical occlusion

Neural Computing and Applications - In the medical imaging domain, nonlinear warping has enabled pixel-by-pixel mapping of one image dataset to a reference dataset. This co-registration of data...     

Investigation of long term drift of an intraocular pressure sensor

Microsystem Technologies - “Zero drift” behavior of an optical intraocular pressure sensor is studied using an analytical model based on the deflection of a circular membrane. Results...     

Detection of tuberculosis using MEMS

Microsystem Technologies - Tuberculosis is a contagious disease caused by the bacteria, Mycobacterium Tuberculosis. Early detection of TB is vital to increase the chances of cure. In this paper, a...     

Co-located collaborative visual analytics around a tabletop display

Co-located collaboration can be extremely valuable during complex visual analytics tasks. We present an exploratory study of a system designed to support collaborative visual analysis tasks on a digital tabletop display. Fifteen participant pairs employed Cambiera, a visual analytics system, to solve a problem involving 240 digital documents. Our analysis, supported by observations, system logs, questionnaires, and interview data, explores how pairs approached the problem around the table. We contribute a unique, rich understanding of how users worked together around the table and identify eight types of collaboration styles that can be used to identify how closely people work together while problem solving. We show how the closeness of teams’ collaboration and communication influenced how they performed on the task overall. We further discuss the role of the tabletop for visual analytics tasks and derive design implications for future co-located collaborative tabletop problem solving systems.     

A digital library framework for heterogeneous music collections: from document acquisition to cross-modal interaction

In this paper, we present a digital library system for managing heterogeneous music collections. The heterogeneity refers to various document types and formats as well as to different modalities, e. g., CD-audio recordings, scanned sheet music, and lyrics. The system offers a full-fledged, widely automated document processing chain: digitization, indexing, annotation, access, and presentation. Our system is implemented as a generic and modular music repository based on a service-oriented software architecture. As a particular strength of our approach, the various documents representing aspects of a piece of music are jointly considered in all stages of the document processing chain. Our user interfaces allow for a multimodal and synchronized presentation of documents (WYSIWYH: what you see is what you hear), a score- or lyrics-based navigation in audio, as well as a cross- and multimodal retrieval. Hence, our music repository may be called a truly cross-modal library system. In our paper, we describe the system components, outline the techniques of the document processing chain, and illustrate the implemented functionalities for user interaction. We describe how the system is put into practice at the Bavarian State Library (BSB) Munich as a part of the German PROBADO Digital Library Initiative (PDLI).     

Higher-Order and Modal Logic as a Framework for Explanation-Based Generalization

Certain tasks, such as formal program development and theorem proving, fundamentally rely upon the manipulation of higher-order objects such as functions and predicates. Computing tools intended to assist in performing these tasks are at present inadequate in both the amount of knowledge they contain (i.e., the level of support they provide) and in their ability to learn (i.e., their capacity to enhance that support over time). The application of a relevant machine learning technique—explanation-based generalization (EBG)—has thus far been limited to first-order problem representations. We extend EBG to generalize higher-order values, thereby enabling its application to higher-order problem encodings.Logic programming provides a uniform framework in which all aspects of explanation-based generalization and learning may be defined and carried out. First-order Horn logics (e.g., Prolog) are not, however, well suited to higher-order applications. Instead, we employ Prolog, a higher-order logic programming language, as our basic framework for realizing higher-order EBG. In order to capture the distinction between domain theory and training instance upon which EBG relies, we extend Prolog with the necessity operator of modal logic. We develop a meta-interpreter realizing EBG for the extended language, Prolog, and provide examples of higher-order EBG.     

Bit transposition for very large scientific and statistical databases

Conventional access methods cannot be effectively used in large Scientific/Statistical Database (SSDB) applications. A file structure (called bit transposed file (BTF)) is proposed which offers several attractive features that are better suited for the special characteristics that SSDBs exhibit. This file structure is an extreme version of the (attribute) transposed file. The data are stored by vertical bit partitions. The bit patterns of attributes are assigned using one of several data encoding methods. Each of these encoding methods is appropriate for different query types. The bit partitions can also be compressed using a version of the run length encoding scheme. Efficient operators on compressed bit vectors have been developed and form the basis of a query language. Because of the simplicity of the file structure and query language, optimization problems for database design, query evaluation, and common subexpression removal can be formalized and efficient exact solution or near optimal solution can be achieved. In addition to selective power with low overheads for SSDBs, the BTF is also amenable to special parallel hardware. Results from experiments with the file structure suggest that this approach may be a reasonable alternative file structure for large SSDBs.     

Predicting fully plastic mode II crack growth from an asymmetric defect

Asymmetrically cracked specimens fail with considerably less ductility than symmetrically cracked ones, due to the crack progressing along a shear band into pre-damaged material. A formulation for the accumulation of damage ahead of an asymmetric crack is presented, based on strain increments following a power law relation. These results are integrated both numerically and approximately.The crack growth per unit displacement increases approximately as the logarithm of the total crack advance per inclusion spacing , and varies inversely as the critical fracture strain c. This provides a basis for predicting large-scale, fully plastic fracture from asymmetric weld defects, using small-scale laboratory specimens.

---

Résumé Les éprouvettes fissurées de manière asymétrique se rompent avec beaucoup moins de ductilité que celles qui sont fissurées de manières symétrique, en raison du fait que la fissure se développe suivant une bande de cisaillement dans un matériau préendommagé. On Présente une formulation de l'accumulation du dommage en avant d'une fissure asymétrique, en se basant sur des gradients de déformations répartis selon une loi quadratique. L'interprétation des résultats est effectuée par voie numérique et par approximation.On constate que le croissance d'une fissure par unité de déplacement est sensiblement proportionnelle au logarithme de l'avancement total de la fissure rapporté à l'espacement entre deux inclusions , et en raison inverse de la déformation critique à la rupture c. Ceci fournt une base pour prédire une rupture à plus grande échelle et en conditions tout à fait plastiques, qui résulteront de défauts de soudage asymétriques, et ce en utilisant des éprouvettes réduites de laboratoire.

     

Molecular Genetics of Conjunctival Melanoma and Prognostic Value of TERT Promoter Mutation Analysis

The aim of this study was exploration of the genetic background of conjunctival melanoma (CM) and correlation with recurrent and metastatic disease. Twenty-eight CM from the Rotterdam Ocular Melanoma Study group were collected and DNA was isolated from the formalin-fixed paraffin embedded tissue. Targeted next-generation sequencing was performed using a panel covering GNAQ, GNA11, EIF1AX, BAP1, BRAF, NRAS, c-KIT, PTEN, SF3B1, and TERT genes. Recurrences and metastasis were present in eight (29%) and nine (32%) CM cases, respectively. TERT promoter mutations were most common (54%), but BRAF (46%), NRAS (21%), BAP1 (18%), PTEN (14%), c-KIT (7%), and SF3B1 (4%) mutations were also observed. No mutations in GNAQ, GNA11, and EIF1AX were found. None of the mutations was significantly associated with recurrent disease. Presence of a TERT promoter mutation was associated with metastatic disease (p-value = 0.008). Based on our molecular findings, CM comprises a separate entity within melanoma, although there are overlapping molecular features with uveal melanoma, such as the presence of BAP1 and SF3B1 mutations. This warrants careful interpretation of molecular data, in the light of clinical findings. About three quarter of CM contain drug-targetable mutations, and TERT promoter mutations are correlated to metastatic disease in CM.     

Association of seed size with genotypic variation in the chemical constituents of soybeans

Ten soybean genotypes grown in 1992 with seed size ranging from 7.6 to 30.3 g/100 seeds and maturity group V or VI were selected and tested for oil and protein content and for fatty acid composition. In these germplasm, protein varied from 39.5 to 50.2%, oil, 16.3 to 21.6%, and protein plus oil, 59.7 to 67.5%. Percentages of individual fatty acids relative to total fatty acids varied as follows: palmitic, 11.0 to 12.8; stearic, 3.2 to 4.7; oleic, 17.6 to 24.2; linoleic, 51.1 to 56.3 and linolenic, 6.9 to 10.0. Seed size showed no significant correlations with individual saturated fatty acids, protein or oil content. However, significant correlations were found between seed size and individual unsaturated fatty acids: positive with oleic, and negative with linoleic and linolenic. Oil and protein content were negatively correlated with each other. Among the major fatty acids, only the unsaturated were significantly correlated with each other: negative between oleic and linoleic or linolenic, and positive between linoleic and linolenic. A subsequent study with soybeans grown in 1993 generally confirmed these findings. Variation in relative percentages of unsaturated fatty acids andr values for most pairs of relationships were even higher than those obtained from the 1992 crop. Presented at the 85th AOCS Annual Meeting and Expo, Atlanta, Georgia, May 8–12, 1994.