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 253 records · Page 14Linked to original sources

PubMedDate not supplied

Automatic speech recognition and training for severely dysarthric users of assistive technology: the STARDUST project.

The STARDUST project developed robust computer speech recognizers for use by eight people with severe dysarthria and concomitant physical disability to access assistive technologies. Independent computer speech recognizers trained with normal speech are of limited functional use by those with severe dysarthria due to limited and inconsistent proximity to "normal" articulatory patterns. Severe dysarthric output may also be characterized by a small mass of distinguishable phonetic tokens making the acoustic differentiation of target words difficult. Speaker dependent computer speech recognition using Hidden Markov Models was achieved by the identification of robust phonetic elements within the individual speaker output patterns. A new system of speech training using computer generated visual and auditory feedback reduced the inconsistent production of key phonetic tokens over time.

Artificial Intelligence↗

PubMed2005

Toward the neurocomputer: image processing and pattern recognition with neuronal cultures.

Information processing in the nervous system is based on parallel computation, adaptation and learning. These features cannot be easily implemented on conventional silicon devices. In order to obtain a better insight of how neurons process information, we have explored the possibility of using biological neurons as parallel and adaptable computing elements for image processing and pattern recognition. Commercially available multielectrode arrays (MEAs) were used to record and stimulate the electrical activity from neuronal cultures. By mapping digital images, i.e., arrays of pixels, into the stimulation of neuronal cultures, a low and bandpass filtering of images could be quickly and easily obtained. Responses to specific spatial patterns of stimulation were potentiated by an appropriate training (tetanization). Learning allowed pattern recognition and extraction of spatial features in processed images. Therefore, neurocomputers, (i.e., hybrid devices containing man-made elements and natural neurons) seem feasible and may become a new generation of computing devices, to be developed by a synergy of Neuroscience and Material Science.

Animals↗

PubMed2007

Joint optimization of word alignment and epenthesis generation for Chinese to Taiwanese sign synthesis.

This work proposes a novel approach to translate Chinese to Taiwanese sign language and to synthesize sign videos. An aligned bilingual corpus of Chinese and Taiwanese Sign Language (TSL) with linguistic and signing information is also presented for sign language translation. A two-pass alignment in syntax level and phrase level is developed to obtain the optimal alignment between Chinese sentences and Taiwanese sign sequences. For sign video synthesis, a scoring function is presented to develop motion transition-balanced sign videos with rich combinations of intersign transitions. Finally, the maximum a posteriori (MAP) algorithm is employed for sign video synthesis based on joint optimization of two-pass word alignment and intersign epenthesis generation. Several experiments are conducted in an educational environment to evaluate the performance on the comprehension of sign expression. The proposed approach outperforms the IBM Model 2 in sign language translation. Moreover, deaf students perceived sign videos generated by the proposed method to be satisfactory.

Algorithms↗

PubMed2002

Slow feature analysis: unsupervised learning of invariances.

Invariant features of temporally varying signals are useful for analysis and classification. Slow feature analysis (SFA) is a new method for learning invariant or slowly varying features from a vectorial input signal. It is based on a nonlinear expansion of the input signal and application of principal component analysis to this expanded signal and its time derivative. It is guaranteed to find the optimal solution within a family of functions directly and can learn to extract a large number of decorrelated features, which are ordered by their degree of invariance. SFA can be applied hierarchically to process high-dimensional input signals and extract complex features. SFA is applied first to complex cell tuning properties based on simple cell output, including disparity and motion. Then more complicated input-output functions are learned by repeated application of SFA. Finally, a hierarchical network of SFA modules is presented as a simple model of the visual system. The same unstructured network can learn translation, size, rotation, contrast, or, to a lesser degree, illumination invariance for one-dimensional objects, depending on only the training stimulus. Surprisingly, only a few training objects suffice to achieve good generalization to new objects. The generated representation is suitable for object recognition. Performance degrades if the network is trained to learn multiple invariances simultaneously.

Algorithms↗

PubMed2004

A functional hierarchical organization of the protein sequence space.

BACKGROUND: It is a major challenge of computational biology to provide a comprehensive functional classification of all known proteins. Most existing methods seek recurrent patterns in known proteins based on manually-validated alignments of known protein families. Such methods can achieve high sensitivity, but are limited by the necessary manual labor. This makes our current view of the protein world incomplete and biased. This paper concerns ProtoNet, a automatic unsupervised global clustering system that generates a hierarchical tree of over 1,000,000 proteins, based solely on sequence similarity. RESULTS: In this paper we show that ProtoNet correctly captures functional and structural aspects of the protein world. Furthermore, a novel feature is an automatic procedure that reduces the tree to 12% its original size. This procedure utilizes only parameters intrinsic to the clustering process. Despite the substantial reduction in size, the system's predictive power concerning biological functions is hardly affected. We then carry out an automatic comparison with existing functional protein annotations. Consequently, 78% of the clusters in the compressed tree (5,300 clusters) get assigned a biological function with a high confidence. The clustering and compression processes are unsupervised, and robust. CONCLUSIONS: We present an automatically generated unbiased method that provides a hierarchical classification of all currently known proteins.

Algorithms↗

PubMed2005

Molecular diagnosis of lymphoma: outcome prediction by gene expression profiling in diffuse large B-cell lymphoma.

This chapter describes the illness diffuse large B-cell lymphoma (DLBCL) and why research has and continues to focus on creating accurate predictors of response to treatment to allow individual risk assessment for a patient and individualization of treatment choice to maximize the chances of cure. Microarray technology has the promise to bring these objectives within reach. The first papers attempting to identify molecular signatures of response and outcome using microarray technology were generated using DLBCL samples and are described. The different types of microarray platform and data analysis tools are reviewed followed by a detailed step-by-step guide to data generation using the Affymetrix chip system from RNA extraction to laser scanning of the hybridized and stained chips.

Artificial Intelligence↗

PubMed2025

Cost-Effectiveness Analysis of 3D Total-Body Photography for People at High Risk of Melanoma.

IMPORTANCE: Greater use of novel digital technologies could be associated with improved health outcomes and save health care costs by detecting smaller melanomas earlier (needing less treatment) or benign tumors (needing no treatment). OBJECTIVE: To compare costs and health effects of 3-dimensional (3D) total-body photography (TBP) and sequential digital dermoscopy imaging (SDDI) vs usual care for early detection of melanoma. DESIGN, SETTING, AND PARTICIPANTS: This prespecified cost-effectiveness analysis using randomized clinical trial (n = 309) data with 2 years of follow-up was conducted at a research hospital in Brisbane, Australia, and took a health system perspective. It included adults 18 years or older at high risk of developing a primary or subsequent melanoma. INTERVENTION: The intervention group received usual care plus clinical skin examinations by junior clinicians at baseline and 6, 12, 18, and 24 months with 3D TBP-SDDI reviewed by a teledermatologist. The control group continued to receive usual care and completed online surveys every 6 months. MAIN OUTCOMES AND MEASURES: Government health care costs, patient out-of-pocket costs, numbers of benign and malignant skin tumor excisions, and quality-adjusted life-years. Skin biopsy, excisions, pathology, and their costs were collected using administrative claims data. Quality of life was collected using the EuroQol-5D-5L. RESULTS: The trial included 314 participants (mean [SD] age, 51.6 [12.8] years; 194 female individuals [62%]) who completed all of the study procedures (158 in the intervention and 156 in the control groups). Compared with controls, intervention group participants had fewer melanoma excisions, more keratinocyte carcinomas and benign excisions, and more biopsy specimens. Over 24 months, mean per-person costs (analyzed in Australian dollars and converted to US$) for the intervention group were $1708 (95% CI, $1455-$1961) vs $763 (95% CI, $655-$870) for controls, an incremental cost of $945 (95% CI, $738-$1157) to provide the intervention. Total quality-adjusted life-years per person were similar for the intervention (1.84; 95% CI, 1.82-1.86) and control groups (1.84; 95% CI, 1.83-1.86). The incremental cost per additional malignant skin tumor excised was $40 (95% CI, $34-$48). CONCLUSIONS AND RELEVANCE: Over 2 years of the trial, the 3D TBP-SDDI model by junior clinicians and teledermatologist review generated higher costs and detected similar numbers of malignant tumors than usual care in a high-risk melanoma cohort. Cost-effectiveness is a necessary but not sufficient consideration for implementation. Other benefits of 3D TBP-SDDI may arise once artificial intelligence clinician support systems are integrated, and more research is needed to understand factors associated with costs and whether there are other benefits of 3D TBP-SDDI.

Adult↗

PubMed2026

Pragmatic gynecologic cancer clinical trials: statements and roadmap from the Gynecologic Cancer InterGroup Chicago Brainstorming Meeting.

Randomized controlled trials remain fundamental to evidence generation in oncology but are increasingly complex, costly, and often misaligned with real-world practice. Traditional explanatory trials, designed under ideal, controlled conditions, frequently enroll highly selected populations, limiting generalizability and underrepresenting key groups such as older adults, patients with comorbidities, and those from low- and middle-income countries. Pragmatic clinical trials offer an alternative by evaluating interventions under routine care conditions, with broader eligibility, simplified procedures, and patient-centered outcomes. To address these challenges, the Gynecologic Cancer InterGroup convened an international brainstorming meeting in May 2025 with multi-disciplinary experts, patients, and advocates to define priorities and develop a roadmap for pragmatic trials in gynecologic oncology. Key discussions emphasized embedding trial design within routine care, aligning eligibility criteria and procedures with standard practice, minimizing non-essential data collection, and prioritizing outcomes meaningful to patients, including quality of life. Innovative designs such as registry-based randomized trials, trials-within-cohorts, and cluster randomization were highlighted as feasible approaches to improve efficiency while preserving internal validity. Integration of patient-reported outcomes and real-world data was considered achievable when carefully streamlined. Major challenges identified included regulatory heterogeneity, consent complexity, data interoperability, and funding limitations, particularly in multi-national settings. Proposed solutions include simplified consent models, centralized ethics processes, hybrid funding strategies, and the responsible use of artificial intelligence to enhance patient identification, recruitment, and potential development of synthetic control arms. Patient engagement was recognized as essential to ensure relevance, feasibility, and equity. Incorporation of patient-reported outcomes was discussed as key to informing acceptance and tolerability. In summary, pragmatic trials within Gynecologic Cancer InterGroup represent a critical pathway to generate efficient, inclusive, and practice-changing evidence in gynecologic cancers across diverse health care settings.

Humans↗

PubMed2004

Information-dependent acquisition-mediated LC-MS/MS screening procedure with semiquantitative potential.

The development of a LC-MS/MS general unknown screening procedure for toxicologically relevant substances in blood samples by means of information-dependent acquisition on a Q-TOF is reported. IDA is an artificial intelligence-based product ion scan mode providing automatic "on-the-fly" MS to MS/MS switching. By performing information-dependent scanning at two different fragmentation energies, two collision-induced dissociation product ion spectra for each of the detected compounds are generated. As such, information-rich MS/MS spectra are obtained from precursor ions not known beforehand. In addition, limitation of the MS/MS acquisition time to an acceptable minimum resulted in an almost instantaneous switch back to the MS mode. As such, this approach provided MS chromatograms that still could be of use for semiquantitative purposes. Since the switching intensity threshold, unequivocally related to the background noise, proved a critical parameter, the solid-phase extraction procedure, the liquid chromatographic conditions, and the mass spectrometric parameters all were optimized to the advantage of information-dependent acquisition. Finally, the screening procedure we developed was benchmarked, on one hand, qualitatively against the results obtained from traditional GUS approaches in a number of routine toxicological laboratories (20 samples) and, on the other hand, quantitatively with respect to its potential against established LC-MS/MS methods (7 samples). The procedure performed very well from a qualitative point of view; almost all of the drugs detected by the conventional techniques were identified, as well as additional drugs that were not previously reported. The procedure proved well-suited for an initial semiquantitative assessment, as is customary in, for example, forensic toxicology before accurate intoxication levels are determined using targeted analytical analyses.

Chromatography, Liquid↗

PubMed1989

A knowledge-based system for automatic interpretation of an analytical profile of complement factors.

A comprehensive assay to evaluate the complement system includes functional tests of both classical and alternative pathways and immunochemical measurements of C3, C4, B, C1-INA, and C3d. The purpose of this analytical profile is to screen for rare hereditary deficiencies and acquired abnormalities of complement and to define the activation pathway in cases of complement consumptive processes. Based on several years' experience, a routine was established in our laboratory to report the data to the clinician, together with a computer-generated interpretive statement. This routine was formulated into a knowledge base by specifying a series of decision rules for each of the complement disorders. After the rules were tested and updated against some 400 complement profile analyses, reasonable analytical comments were produced by the system. This knowledge-based system for reporting and interpreting complement results offers several advantages: the interpretative work is facilitated and made more reliable; a consistent interpretative comment is generated that is recognized and therefore more meaningful for the clinician; the communication of analytical procedures and policy is enhanced.

Artificial Intelligence↗

PubMed1997

A case-based consiliarius for therapy recommendation (ICONS): computer-based advice for calculated antibiotic therapy in intensive care medicine.

We report here on the system ICONS which utilizes case-based reasoning for medical decision support. As an application domain we have chosen the medical field of 'calculated antibiotic therapy' in an intensive care medicine setting. The system ICONS which runs on a personal computer suggests adequate antibiotic therapy regimen satisfying medical and economic conditions. To speed up the process of finding an adequate antibiotic therapy for a current patient, case-based reasoning is used for finding previously documented similar cases and for modifying them according to the requirements of the current patient. To reduce the memory capacity for the documentation of cases, collections of similar cases are clustered to prototypes. Medical knowledge is represented within a hierarchy of such prototypes and cases and an additional context-sensitive background knowledge-base. A knowledge acquisition tool was programmed that allows revisions of the background medical knowledge-base by simple and comprehensive methods. In addition to the advantage of producing site-specific and time-dependent knowledge, case-based reasoning is a practical method for speeding-up the process of generating and evaluating hypotheses in medical classification tasks.

Anti-Bacterial Agents↗

PubMed2004

Retinally-generated saccadic suppression of a locust looming-detector neuron: investigations using a robot locust.

A fundamental task performed by many visual systems is to distinguish apparent motion caused by eye movements from real motion occurring within the environment. During saccadic eye movements, this task is achieved by inhibitory signals of central and retinal origin that suppress the output of motion-detecting neurons. To investigate the retinally-generated component of this suppression, we used a computational model of a locust looming-detecting pathway that experiences saccadic suppression. This model received input from the camera of a mobile robot that performed simple saccade-like movements, allowing the model's response to simplified real stimuli to be tested. Retinally-generated saccadic suppression resulted from two inhibitory mechanisms within the looming-detector's input architecture. One mechanism fed inhibition forward through the network, inhibiting the looming-detector's initial response to movement. The second spread inhibition laterally within the network, suppressing the looming-detector's maintained response to movement. These mechanisms prevent a looming-detector model response to whole-field visual stimuli. In the locust, this mechanism of saccadic suppression may operate in addition to centrally-generated suppression. Because lateral inhibition is a common feature of early visual processing in many organisms, we discuss whether the mechanism of retinally-generated saccadic suppression found in the locust looming-detector model may also operate in these species.

Animals↗

PubMed2006

A robust method for spike sorting with automatic overlap decomposition.

Spike sorting is the mandatory first step in analyzing multiunit recording signals for studying information processing mechanisms within the nervous system. Extracellular recordings usually contain overlapped spikes produced by a number of neurons adjacent to the electrode, together with unknown background noise, which in turn induce some difficulties in neural signal identification. In this paper, we propose a robust method to deal with these problems, which employs an automatic overlap decomposition technique based on the relaxation algorithm that requires simple fast Fourier transforms. The performance of the presented system was tested at various signal-to-noise ratio levels based on synthetic data that were generated from real recordings.

Action Potentials↗

PubMed2006

Identifying interaction sites in "recalcitrant" proteins: predicted protein and RNA binding sites in rev proteins of HIV-1 and EIAV agree with experimental data.

Protein-protein and protein nucleic acid interactions are vitally important for a wide range of biological processes, including regulation of gene expression, protein synthesis, and replication and assembly of many viruses. We have developed machine learning approaches for predicting which amino acids of a protein participate in its interactions with other proteins and/or nucleic acids, using only the protein sequence as input. In this paper, we describe an application of classifiers trained on datasets of well-characterized protein-protein and protein-RNA complexes for which experimental structures are available. We apply these classifiers to the problem of predicting protein and RNA binding sites in the sequence of a clinically important protein for which the structure is not known: the regulatory protein Rev, essential for the replication of HIV-1 and other lentiviruses. We compare our predictions with published biochemical, genetic and partial structural information for HIV-1 and EIAV Rev and with our own published experimental mapping of RNA binding sites in EIAV Rev. The predicted and experimentally determined binding sites are in very good agreement. The ability to predict reliably the residues of a protein that directly contribute to specific binding events--without the requirement for structural information regarding either the protein or complexes in which it participates--can potentially generate new disease intervention strategies.

Amino Acid Sequence↗

PubMed2006

Knowledge acquisition for computation of semantic distance between WHO-ART terms.

Computation of semantic distance between adverse drug reactions terms may be an efficient way to group related medical conditions in pharmacovigilance case reports. Previous experience with ICD-10 on a semantic distance tool highlighted a bottleneck related to manual description of formal definitions in large terminologies. We propose a method based on acquisition of formal definitions by knowledge extraction from UMLS and morphosemantic analysis. These formal definitions are expressed with SNOMED International terms. We provide formal definitions for 758 WHO-ART terms: 321 terms defined from UMLS, 320 terms defined using morphosemantic analysis and 117 terms defined after expert evaluation. Computation of semantic distance (e.g. k-nearest neighbours) was implemented in J2EE terminology services. Similar WHO-ART terms defined by automated knowledge acquisition and ICD terms defined manually show similar behaviour in the semantic distance tool. Our knowledge acquisition method can help us to generate new formal definitions of medical terms for our semantic distance terminology services.

Adverse Drug Reaction Reporting Systems↗

PubMed2000

John von Neumann and the evolutionary growth of complexity: looking backward, looking forward...

In the late 1940s John von Neumann began to work on what he intended as a comprehensive "theory of [complex] automata." He started to develop a book length manuscript on the subject in 1952. However, he put it aside in 1953, apparently due to pressure of other work. Due to his tragically early death in 1957, he was never to return to it. The draft manuscript was eventually edited, and combined for publication with some related lecture transcripts, by Burks in 1966. It is clear from the time and effort that von Neumann invested in it that he considered this to be a very significant and substantial piece of work. However, subsequent commentators (beginning even with Burks) have found it surprisingly difficult to articulate this substance. Indeed, it has since been suggested that von Neumann's results in this area either are trivial, or, at the very least, could have been achieved by much simpler means. It is an enigma. In this paper I review the history of this debate (briefly) and then present my own attempt at resolving the issue by focusing on an analysis of von Neumann's problem situation. I claim that this reveals the true depth of von Neumann's achievement and influence on the subsequent development of this field, and further that it generates a whole family of new consequent problems, which can still serve to inform - if not actually define - the field of artificial life for many years to come.

Artificial Intelligence↗

PubMed2026

How advances in machine learning drive early detection and risk prediction of early-onset colorectal cancer.

Early-onset colorectal cancer (EOCRC), defined as colorectal cancer diagnosed before age 50, is rising across high- and middle-income settings whilst organised screening stays anchored to older age thresholds. Blood-based liquid biopsy, combined with machine learning, is the most plausible route to early detection in this group because it does not depend on bowel preparation, endoscopy capacity, or adherence to stool-based testing. The gap is structural: incidence climbs fastest in the population below the age at which any guideline-endorsed modality is offered. The analytical challenge is that early-stage tumour-derived signals in plasma are low in abundance and distributed across heterogeneous molecular layers: circulating tumour DNA mutations, aberrant methylation, cfDNA fragmentomics, and small non-coding RNA. Machine learning converts these into a single calibrated probability. This review examines where artificial intelligence (AI)-driven liquid biopsy genuinely adds diagnostic value in EOCRC, distinguishes components in which learned models are decorative from those in which they are mechanistically necessary, and identifies the validation deficit separating research cohorts from deployable clinical tools. It summarises the first-generation tools used clinically for early detection and post-treatment monitoring, then considers analytes from exosome-bound microRNAs to long-read whole-genome sequencing of circulating plasma DNA, which reads cytosine modification natively, resolves methylation and fragmentation on single molecules, and characterises structural events short reads cannot anchor. Any analyte can feed a learned model, but more diverse input yields better discrimination. The central argument is that approved, guideline-included blood tests were validated in populations aged 45 and above, and their performance in younger patients cannot be assumed.

cfDNA fragmentomics↗

PubMed1996

Artificial intelligence in pediatrics: important clinical signs in newborn syndromes.

New methods are warranted in the field of syndromology. This study is an exploration into whether an artificial intelligence method (ID3) could provide a new angle for approaching syndromes. Diagnosing syndromes in the newborn is difficult. The accepted approach is to look for individual clinical signs that add up to a syndrome diagnosis. Of all possible clinical signs, one would want to extract the signs with the strongest predictive power. I used the ID3 algorithm to extract predictive clinical signs from a catalogue of syndromes (Birth Defects Encyclopedia Online; BDEO). Using information from BDEO, files of randomly generated "patients" were created. The signs consistently high in the identification tree were long philtrum, short palpebral fissures, low-set ears, and hepatosplenomegaly. The program used featured a crude "expert system" based on the ID3 algorithm. When using one-half of the data set as a training set and the other half as a testbed, a correct classification rate of 92.1-98.1% was attained. When the ID3 expert system was tested against cases from a clinical database (Pictures of Standard Syndromes and Undiagnosed Malformations), the correct classification rate was less than 20%. This may not necessarily reflect faults with the ID3 approach, but possibly biases in the clinical database. In syndromology no "criterion standards" exist that can confirm a diagnosis. The statistical method of cluster analysis does not require prior knowledge of diagnoses and will make a tree of syndromes based upon clinical signs. A cluster analysis was performed as a validity check to provide a tree for comparison with the ID3 tree. There was a reasonable degree of agreement between the two. Applying artificial intelligence methods to this field highlights problems with basic assumptions and philosophical aspects of syndrome diagnosis.

Algorithms↗