SEARCH · PubMed Health

Results for “generative artificial intelligence”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 631 records · Page 35Linked to original sources

PubMed2004

Generalized parallel-perspective stereo mosaics from airborne video.

In this paper, we present a new method for automatically and efficiently generating stereoscopic mosaics by seamless registration of images collected by a video camera mounted on an airborne platform. Using a parallel-perspective representation, a pair of geometrically registered stereo mosaics can be precisely constructed under quite general motion. A novel parallel ray interpolation for stereo mosaicing (PRISM) approach is proposed to make stereo mosaics seamless in the presence of obvious motion parallax and for rather arbitrary scenes. Parallel-perspective stereo mosaics generated with the PRISM method have better depth resolution than perspective stereo due to the adaptive baseline geometry. Moreover, unlike previous results showing that parallel-perspective stereo has a constant depth error, we conclude that the depth estimation error of stereo mosaics is in fact a linear function of the absolute depths of a scene. Experimental results on long video sequences are given.

Aircraft↗

PubMed1986

Data storage and retrieval.

The entire face of modern medical and surgical practice is being significantly affected by the application of technologic developments to the practice of surgery--developments that will tie together such areas as information management and processing, robotics, communication networks, and computerized surgical equipment. The achievements in these areas will create a sophisticated, fully automatic system that will assist the plastic surgeon in many aspects of work, such as regular office activities, doctor-patient interaction, professional updating, communication, and even assistance during the operational process itself. It will be as simple as dialing a telephone today. When it is necessary to consult with other colleagues, a combined vocal and visual consulting network in other medical centers as well as consulting computerized expert systems will be available all day and night as part of the communication services. The plastic surgical expert systems will store valuable information, based on the knowledge of the best human experts, on any important subtopics and will be accessed in a very friendly way. This will be an invaluable tool for the residents in training, for emergency room work, and for just getting a second opinion, even for the more experienced practitioner. All the electronic mail, professional magazines, and any other required professional information will flow between central and personal retrieval systems. The doctor, at a desired time in the privacy and comfort of his or her own home or office, can read the mail, make required changes to suit his or her needs, and store, send back, or distribute information, all in a speedy and efficient manner. The simulation of a planned surgery will give the surgeon the ability to prepare and will prevent difficulties during complicated procedures through the luxury of a dry run, without any sequelae if certain expected outcomes fail to materialize. The preprogrammed control of sophisticated surgical equipment and the use of robotics would generate new operational possibilities for more complicated surgeries, which are now prevented owing to the surgeon's physical limitations. Information urgently required during the operation as a result of an unexpected situation will be available immediately from storage and retrieval systems, and real-time vocal and visual consulting with expert colleagues, often in remote locations, will bring the operations process itself to a new era.(ABSTRACT TRUNCATED AT 400 WORDS)

Artificial Intelligence↗

PubMed2005

Statistical representation and simulation of high-dimensional deformations: application to synthesizing brain deformations.

This paper proposes an approach to effectively representing the statistics of high-dimensional deformations, when relatively few training samples are available, and conventional methods, like PCA, fail due to insufficient training. Based on previous work on scale-space decomposition of deformation fields, herein we represent the space of "valid deformations" as the intersection of three subspaces: one that satisfies constraints on deformations themselves, one that satisfies constraints on Jacobian determinants of deformations, and one that represents smooth deformations via a Markov Random Field (MRF). The first two are extensions of PCA-based statistical shape models. They are based on a wavelet packet basis decomposition that allows for more accurate estimation of the covariance structure of deformation or Jacobian fields, and they are used jointly due to their complementary strengths and limitations. The third is a nested MRF regularization aiming at eliminating potential discontinuities introduced by assumptions in the statistical models. A randomly sampled deformation field is projected onto the space of valid deformations via iterative projections on each of these subspaces until convergence, i.e. all three constraints are met. A deformation field simulator uses this process to generate random samples of deformation fields that are not only realistic but also representative of the full range of anatomical variability. These simulated deformations can be used for validation of deformable registration methods. Other potential uses of this approach include representation of shape priors in statistical shape models as well as various estimation and hypothesis testing paradigms in the general fields of computational anatomy and pattern recognition.

Algorithms↗

PubMed1998

[Voice prostheses with sound-producing metal reed element--an experimental study and initial clinical results].

BACKGROUND: Following total laryngectomy the voice is produced by esophageal speech as well as with voice prostheses by vibrations of pharyngeal mucosal folds. This pharyngeal sound normally has a significantly lower fundamental frequency than the healthy voice (men about 120 Hz, women about 240 Hz, pharyngeal voice about 70 Hz), which is a handicap especially for female laryngectomy patients. In order to improve the postlaryngectomy voice, a new type of voice prostheses containing an integrated sound-producing metallic reed element was developed (ADEVA Company, Lübeck, Germany). METHODS/PATIENTS: Thirty-five of these new sound-producing voice prostheses were tested in vitro for different prosthesis-specific physical parameters (pressure, flow, sound pressure, flow resistance, frequency range). In 15 voice prosthesis speakers, a sound-producing prosthesis was introduced during a routine outpatient visit. Besides measurement of the above mentioned physical parameters in patients with conventional and sound-producing prostheses, the resulting voice as also evaluated by means of a video recording. RESULTS: In vitro all prostheses with the metallic reed element produced a clear sound. Flow resistance of the prostheses was slightly elevated by the reed element. Insertion of the prostheses was hindered by the reed element. Period of uninterrupted sound production was prolonged after insertion of a sound-producing prosthesis and patients could speak on a lower pressure level, but the sound of the reed element was permanently distinguishable only in 6 of 15 patients. CONCLUSIONS: In principle a variation of the pharyngeal voice by means of a sound producing element, which is integrated into a voice prosthesis, is possible. The current design of the metallic reed element tested is not yet suitable for routine clinical use: 1. The reed element is too sensitive and is easily damaged during insertion, so the insertion device has to be improved. 2. The sound producing element is blocked by small amounts of tracheal secretions, so that this element should be replaceable separately without requiring removal of the silicone value (if possible by the patient himself). Prior to insertion of the sound producing voice prosthesis the maximum air flow through the shunt should be measured to determine if the patient can produce the necessary air flow for activation of the reed element. A further improvement for these special types of voice prostheses would be a sound producing element, which generates a variable frequency of sound. Limiting the patient to only one fundamental frequency creates a monotone, which does not sound naturally. Initial progress toward a sound-producing voice prostheses has been made. This should be followed by the necessary improvements in order to improve the feasibility of this design for routine clinical use.

Equipment Failure Analysis↗

PubMedDate not supplied

A new paradigm for explaining and linking knowledge in diagnostic problem solving.

Medical expert systems frequently use causal models to capture knowledge and diagnostic-problem-solving expertise. A significant obstacle confronting these systems is providing informative explanations without prohibitive computational expense. The explanations should allow the user to understand the decisions of the expert system and obtain additional details when needed. A new method, called HyperExplain, has been devised to flexibly link explanations with conclusions generated by a causal reasoning system. This approach creates a patient specific explanatory (PSE) model for the medical expert system that provides decision support from a variety of perspectives. A key feature of this method is the ability to alter the focus of explanations depending upon the problem-solving context and patient manifestations. The method has been implemented in a program that provides diagnostic assistance to physicians in the domain of neurophysiology.

Artificial Intelligence↗

PubMed2006

Sigma: multiple alignment of weakly-conserved non-coding DNA sequence.

BACKGROUND: Existing tools for multiple-sequence alignment focus on aligning protein sequence or protein-coding DNA sequence, and are often based on extensions to Needleman-Wunsch-like pairwise alignment methods. We introduce a new tool, Sigma, with a new algorithm and scoring scheme designed specifically for non-coding DNA sequence. This problem acquires importance with the increasing number of published sequences of closely-related species. In particular, studies of gene regulation seek to take advantage of comparative genomics, and recent algorithms for finding regulatory sites in phylogenetically-related intergenic sequence require alignment as a preprocessing step. Much can also be learned about evolution from intergenic DNA, which tends to evolve faster than coding DNA. Sigma uses a strategy of seeking the best possible gapless local alignments (a strategy earlier used by DiAlign), at each step making the best possible alignment consistent with existing alignments, and scores the significance of the alignment based on the lengths of the aligned fragments and a background model which may be supplied or estimated from an auxiliary file of intergenic DNA. RESULTS: Comparative tests of sigma with five earlier algorithms on synthetic data generated to mimic real data show excellent performance, with Sigma balancing high "sensitivity" (more bases aligned) with effective filtering of "incorrect" alignments. With real data, while "correctness" can't be directly quantified for the alignment, running the PhyloGibbs motif finder on pre-aligned sequence suggests that Sigma's alignments are superior. CONCLUSION: By taking into account the peculiarities of non-coding DNA, Sigma fills a gap in the toolbox of bioinformatics.

Algorithms↗

PubMed1993

Toward multistrategy parallel and distributed learning in sequence analysis.

Machine learning techniques have been shown to be effective in sequence analysis tasks. However, current learning algorithms, which are typically serial main-memory-based, are not capable of handling the vast amounts of information being generated by the Human Genome Project. The multistrategy parallel learning approach presented in this paper is an attempt to scale existing learning algorithms. Learning speed is improved through running multiple learning processes in parallel and prediction accuracy is improved through multiple learners. Our approaches are independent of the learning algorithms used. This paper focuses on one of the MSPL approaches and preliminary empirical results that we present are encouraging.

Algorithms↗

PubMed1995

Proper staging techniques in testicular cancer patients.

Testicular cancer has become a highly curable neoplasm, and research efforts in the 1990s are focusing on ways to improve staging and treatment so as to limit cost and morbidity. Our group has performed a number of recent studies that help to answer a number of important clinical questions. First, do we need to order computed tomography of the chest (CCT) to stage all newly diagnosed patients? Second, how accurate is contemporary era abdominal CT to stage the retroperitoneum in low-stage nonseminoma patients, and are there techniques that may improve accuracy? Third, can histological primary tumor factors be useful to predict stage in low-stage nonseminoma patients? In a study of 201 testicular cancer patients [117 (58%) NSGCT, 84 (42%) seminoma] who had both CCT and chest X-ray (CXR) in initial staging, CXR alone was found to be sufficient initial chest staging in all seminoma patients and in NSGCT patients who had a negative abdominal (CTA). For low-stage patients without retroperitoneal adenopathy, CCT had unacceptable false-positive rates, which precipitated additional invasive maneuvers. For higher stage NSGCT patients with retroperitoneal disease on initial CTA, CXR alone missed a significant number of occult thoracic metastases and CCT remains indicated. In a study of 57 clinical stage 1 NSGCT all having negative staging CTA followed by surgical staging, third and fourth generation CT had a 67% accuracy in predicting retroperitoneal metastases. This contemporary experience shows a 33% false-negative abdominal CT staging rate. Consideration of any nodes, regardless of size, in the primary echelon retroperitoneal areas as indicative of retroperitoneal metastases may hold promise for improving accuracy of CTA in low-stage NSGCT testis cancer. In a study of 92 clinical stage 1 NSGCT patients, determination of primary tumor vascular invasion (VI) and percentage of tumor composed of embryonal carcinoma component (%EMB) was found to be a useful staging tool. A multivariate model using VI and %EMB was able to predict correct stage in 86% of the study cohort, and a probability table with these two variables was created. Using these histological variables in a neural network artificial intelligence program, an expert correctly predicted stage in 92% of patients. In the 1990s chest staging should be tailored to tumor cell type and retroperitoneal disease status. Staging of the retroperitoneum utilizing abdominal CT remains problematic due to the inability to detect microscopic metastases. Primary tumor histological factors, particularly vascular invasion and quantitation of embryonal carcinoma, are clinically useful staging tools.

Carcinoma, Embryonal↗

PubMed1989

The implementation of a knowledge-based Pathology Hypertext under HyperCard.

A knowledge-based Hypertext of Pathology integrating videodisc-based images and computer-generated graphics with the textual cognitive information of an undergraduate pathology curriculum has been developed. The system described in this paper was implemented under HyperCard during 1988 and 1989. Three earlier versions of the system that were developed on different platforms are contrasted with the present system. Strengths, weaknesses, and future extensions of the system are enumerated. The conceptual basis and organizational principles of the knowledge base are also briefly discussed.

Artificial Intelligence↗

PubMedDate not supplied

Model control of image processing: pupillometry.

Smart instruments require on-line computers, special purpose hardware, or both. A pupillometer is described that relies only on a general purpose microcomputer with a frame grabber to process infrared video camera pictures of the human eye. An essential feature of the instrument is that a top-down model controls the image processing algorithms. The model generates regions of interest, ROIs, positioned from knowledge of anatomy and optics of the eye and information from previously analyzed frames. Within these adaptively controlled ROIs, fast, run-time algorithms automatically calculate local thresholds, area measurements, moments for centroid position information, and use pyramiding to shorten calculation time for large pupils. Outputs are precise measurements of pupil size and eye position in real time, with adequate bandwidth for most purposes.

Artificial Intelligence↗

PubMed2006

Pattern generation using likelihood inference for cellular automata.

Cellular automata are discrete dynamical systems which evolve on a discrete grid. Recent studies have shown that cellular automata with relatively simple rules can produce highly complex patterns. We develop likelihood-based methods for estimating rules of cellular automata aimed at the re-generation of observed regular patterns. Under noisy data, our approach is equivalent to estimating the local map of a stochastic cellular automaton. Direct computations of the maximum likelihood estimates are possible for regular binary patterns. The likelihood formulation of the problem is congenial with the use of the minimum description length principle as a model selection tool. We illustrate our method with a series of examples using binary images.

Algorithms↗

PubMed2005

Automated DNA fragments recognition and sizing through AFM image processing.

This paper presents an automated algorithm to determine DNA fragment size from atomic force microscope images and to extract the molecular profiles. The sizing of DNA fragments is a widely used procedure for investigating the physical properties of individual or protein-bound DNA molecules. Several atomic force microscope (AFM) real and computer-generated images were tested for different pixel and fragment sizes and for different background noises. The automated approach minimizes processing time with respect to manual and semi-automated DNA sizing. Moreover, the DNA molecule profile recognition can be used to perform further structural analysis. For computer-generated images, the root mean square error incurred by the automated algorithm in the length estimation is 0.6% for a 7.8 nm image pixel size and 0.34% for a 3.9 nm image pixel size. For AFM real images we obtain a distribution of lengths with a standard deviation of 2.3% of mean and a measured average length very close to the real one, with an error around 0.33%.

Algorithms↗

PubMed2006

A generative sketch model for human hair analysis and synthesis.

In this paper, we present a generative sketch model for human hair analysis and synthesis. We treat hair images as 2D piecewise smooth vector (flow) fields and, thus, our representation is view-based in contrast to the physically-based 3D hair models in graphics. The generative model has three levels. The bottom level is the high-frequency band of the hair image. The middle level is a piecewise smooth vector field for the hair orientation, gradient strength, and growth directions. The top level is an attribute sketch graph for representing the discontinuities in the vector field. A sketch graph typically has a number of sketch curves which are divided into 11 types of directed primitives. Each primitive is a small window (say 5 x 7 pixels) where the orientations and growth directions are defined in parametric forms, for example, hair boundaries, occluding lines between hair strands, dividing lines on top of the hair, etc. In addition to the three level representation, we model the shading effects, i.e., the low-frequency band of the hair image, by a linear superposition of some Gaussian image bases and we encode the hair color by a color map. The inference algorithm is divided into two stages: 1) We compute the undirected orientation field and sketch graph from an input image and 2) we compute the hair growth direction forthe sketch curves and the orientation field using a Swendsen-Wang cut algorithm. Both steps maximize a joint Bayesian posterior probability. The generative model provides a straightforward way for synthesizing realistic hair images and stylistic drawings (rendering) from a sketch graph and a few Gaussian bases. The latter can be either inferred from a real hair image or input (edited) manually using a simple sketching interface. We test our algorithm on a large data set of hair images with diverse hair styles. Analysis, synthesis, and rendering results are reported in the experiments.

Algorithms↗

PubMed2004

A comparative study of fuzzy classification methods on breast cancer data.

In this paper, we examine the performance of four fuzzy rule generation methods on Wisconsin breast cancer data. The first method generates fuzzy if-then rules using the mean and the standard deviation of attribute values with 92.2% correct classification rate. The second approach generates fuzzy if-then rules using the histogram of attributes values with 86.7% correct classification rate. The third procedure generates fuzzy if-then rules with certainty of each attribute into homogeneous fuzzy sets with 99.73% correct classification rate. In the fourth approach, only overlapping areas are partitioned with 62.57% correct classification rate. The first two approaches generate a single fuzzy if-then rule for each class by specifying the membership function of each antecedent fuzzy set using the information about attribute values of training patterns. The other two approaches are based on fuzzy grids with homogeneous fuzzy partitions of each attribute. The performance of each approach is evaluated on breast cancer data sets. Simulation results show that the simple grid approach has a high classification rate of 99.73%.

Algorithms↗

PubMed1998

A development environment for knowledge-based medical applications on the World-Wide Web.

The World-Wide Web (WWW) is increasingly being used as a platform to develop distributed applications, particularly in contexts, such as medical ones, where high usability and availability are required. In this paper we propose a methodology for the development of knowledge-based medical applications on the web, based on the use of an explicit domain ontology to automatically generate parts of the system. We describe a development environment, centred on the LISPWEB Common Lisp HTTP server, that supports this methodology, and we show how it facilitates the creation of complex web-based applications, by overcoming the limitations that normally affect the adequacy of the web for this purpose. Finally, we present an outline of a system for the management of diabetic patients built using the LISPWEB environment.

Artificial Intelligence↗

PubMed2001

The TRANSPATH signal transduction database: a knowledge base on signal transduction networks.

UNLABELLED: TRANSPATH is an information system on gene-regulatory pathways, and an extension module to the TRANSFAC database system (Wingender et al., Nucleic Acids Res., 28, 316-319, 2000). It focuses on pathways involved in the regulation of transcription factors in different species, mainly human, mouse and rat. Elements of the relevant signal transduction pathways like complexes, signaling molecules, and their states are stored together with information about their interaction in an object-oriented database. The database interface provides clickable maps and automatically generated pathway cascades as additional ways to explore the data. All information is validated with references to the original publications. Also, references to other databases are provided (TRANSFAC, SWISS-PROT, EMBL, PubMed and others). AVAILABILITY: The database is available over (http://transpath.gbf.de) for interactive perusal. As an exchange format for the data, eXtensible Markup Language (XML) flatfiles and a Document Type Definition (DTD) are provided.

Algorithms↗

PubMed2004

Approaching the Ocean Color problem using fuzzy rules.

In this paper, we propose a fuzzy logic-based approach which exploits remotely sensed multispectral measurements of the reflected sunlight to estimate the concentration of optically active constituents of the sea water. The relation between the concentrations of interest and the subsurface reflectances is modeled by a set of fuzzy rules extracted automatically from the data through a two-step procedure. First, a compact initial rule base is generated by projecting onto the input variables the clusters produced by a fuzzy clustering algorithm. Then, a genetic algorithm is applied to optimize the rules. Appropriate constraints maintain the semantic properties of the initial model during the genetic evolution. Results of the application of the fuzzy model obtained from data simulated with an ocean color model over the channels of the Medium Resolution Imaging Spectrometer are shown and discussed.

Algorithms↗

PubMed2005

Synchronized oscillation in a modular neural network composed of columns.

The columnar organization is a ubiquitous feature in the cerebral cortex. In this study, a neural network model simulating the cortical columns has been constructed. When fed with random pulse input with constant rate, a column generates synchronized oscillations, with a frequency varying from 3 to 43 Hz depending on parameter values. The behavior of the model under periodic stimulation was studied and the input-output relationship was non-linear. When identical columns were sparsely interconnected, the column oscillator could be locked in synchrony. In a network composed of heterogeneous columns, the columns were organized by intrinsic properties and formed partially synchronized assemblies.

Animals↗