SEARCH · PubMed Health

Results for “Explainable AI”

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 55 records · Page 3Linked to original sources

PubMed2026

The Impact of Chatbot Type and Normative Messaging on Chatbot Usage Intention Based on the Health Technology Acceptance Model: Randomized Controlled Trial.

BACKGROUND: Digital health tools, such as health chatbots, may improve access to scalable health support, but adoption remains inconsistent. Existing models do not fully integrate technology acceptance factors with health motivation factors relevant to digital health use. OBJECTIVE: This study proposed and tested the health technology acceptance model and examined whether normative message framing and chatbot type were associated with health motivation, technology acceptance, and intention to use a health chatbot. METHODS: In October 2025, we conducted a 4 × 2 between-participants online experiment with 1000 US adults recruited from a nationally representative YouGov panel. Participants were randomized to 1 of 8 conditions varying norm message type (self-oriented, peer-oriented, expert-oriented, or family-oriented) and chatbot type (AI-powered or rule-based) in a cancer prevention and genetic risk information scenario. Outcomes included descriptive norms, injunctive norms, perceived susceptibility, perceived severity, perceived benefits, self-efficacy, perceived ease of use, trust, privacy concerns, and usage intention. Data were analyzed using a multivariate ANOVA with Bonferroni-adjusted post hoc tests and multiple linear regression. RESULTS: Peer-oriented and family-oriented messages produced higher usage intention than expert-oriented messages, and peer-oriented messages also increased descriptive norms, injunctive norms, self-efficacy, and trust. AI-powered chatbots were associated with higher usage intention (P=.02) and greater trust (P=.008) than rule-based chatbots. In regression analyses, the model explained 50.8% of the variance in usage intention. Usage intention was positively associated with descriptive norms (β=0.087; P=.003), injunctive norms (β=0.078; P=.009), perceived susceptibility (β=0.051; P=.03), perceived benefits (β=0.253; P<.001), and trust (β=0.33; P<.001), and negatively associated with perceived severity (β=-0.047; P=.049) and privacy concerns (β=-0.11; P<.001). Perceived ease of use and self-efficacy were not significant predictors. CONCLUSIONS: The health technology acceptance model was a useful framework for explaining the intention to use a health chatbot by combining technology acceptance and health motivation constructs. Both social design features and chatbot design features shaped adoption-related beliefs, with peer-oriented and family-oriented framing and AI-powered chatbots showing particular promise. Trust and privacy concerns remained central determinants of intended use.

Humans↗

PubMed1991

Understanding the speech-understanding problems of the hearing impaired.

This article presents a tutorial overview of the speech-recognition difficulties of the hearing impaired. Much recent research indicates that the primary problem underlying the speech-recognition difficulties of the hearing impaired is the loss of hearing sensitivity and accompanying loudness recruitment. This tutorial demonstrates why loss of hearing sensitivity plays such an important role and why hearing aids fitted with contemporary fitting strategies provide only limited benefit in noise for many of these individuals. In addition, low-pass noise is often said to present the hearing-impaired listener with greater difficulties than broad-band noise. This observation is also explained quite simply by careful consideration of the loss of hearing sensitivity. Finally, two clinical articulation index (AI) calculation schemes, designed to quantify the effects of hearing loss on speech understanding, are reviewed and evaluated.

Audiometry, Speech↗

PubMed1992

Polymorphisms in the apolipoprotein (apo) AI-CIII-AIV gene cluster: detection of genetic variation determining plasma apo AI, apo CIII and apo AIV concentrations.

We have examined the associations between levels of plasma apolipoprotein (apo) AI, apo CIII and apo AIV and genetic variation in the apo AI-CIII-AIV gene cluster in 162 boys and young men from Belgium aged from 7 to 23 years. Genotypes were determined for six restriction enzymes XmnI, PstI, SstI, PvuIIA-CIII, PvuIIB-AIV and XbaI, and for the G to A substitution at -75 bp in the 5' region of the apo AI gene. The polymorphism most strongly associated with apo AI levels was the G to A substitution (P = 0.025, R2 x 100 = 3.6%) confirming previous observations. The polymorphism most strongly associated with apo CIII levels was that of PvuIIA-CIII (P = 0.023, R2 x 100 = 2.9%) in the apo CIII gene. This novel association must be interpreted with caution until it has been confirmed in an independent sample. The polymorphism associated with the largest effect on apo AIV levels was that detected with XbaI in the apo AIV gene, but this association was not statistically significant. Previously reported associations between the SstI polymorphism and triglyceride levels, and between the PstI polymorphism and apo AI levels, were weakly detected in the present sample. Our results show that variation associated with some of the polymorphisms in the apo AI-CIII-AIV cluster makes a small, but statistically significant, contribution to the determination of apo AI and apo CIII levels in this sample of young men and boys. These observations may, in part, explain reported associations between polymorphisms in this gene cluster, differences in plasma lipid and lipoprotein levels, and prevalence of coronary artery disease.

Adolescent↗

PubMed1991

Apolipoprotein E polymorphism and plasma lipid, lipoprotein, and apolipoprotein levels in Italian children.

We have investigated the effect of apolipoprotein (apo) E polymorphism on serum lipid, lipoprotein, and apolipoprotein levels in a sample of 195 children, aged 8-11 years, from Sezze, Central Italy. The relative frequencies of e2, e3, and e4 alleles were 0.062, 0.867, and 0.072, respectively. Variation at the apo E gene locus explained 5.1% of the sample variance in serum total cholesterol levels, 7.6% in low-density lipoprotein (LDL) cholesterol levels, 7.3% in apo B levels, and 14.1% in high-density lipoprotein-apo E (HDL-E) levels. The effect of the e2 allele was to lower levels of total cholesterol, LDL-cholesterol, and apo B and to raise levels of HDL-E, while the effect of the e4 allele was the opposite. Variation at the apo E gene locus was not associated with differences in serum triglyceride, HDL-cholesterol, or apo AI levels. The effects of common apo E polymorphisms and genetic variation associated with the PvuII RFLP of the apo B gene on serum apo B levels were additive, explaining 11.3% of the phenotypic variance in this sample. When the effect of apo E polymorphism on serum lipid traits was estimated in boys and girls separately, variation at the apo E gene locus explained 10.4, 13.3, 13.3, and 13.5% of the phenotypic variance in serum total cholesterol, LDL-cholesterol, apo B, and HDL-E levels, respectively, in boys, while in girls only the effect on HDL-E levels (19.3%) reached statistical significance. This study has demonstrated that genetic variations at the apo E locus contribute to the determination of serum lipid, lipoprotein, and apolipoprotein levels in youths and that the effects are gender specific.

Alleles↗

PubMed1975

Spermatogenetic clones developing from repopulating stem cells surviving a high dose of an alkylating agent.

We investigated stem cell renewal and differentiation in 10- and 15-days-old spermatogonial clones developing in mouse seminiferous epithelium after an extremely large cell loss, inflicted by high doses of the alkylating agent Myleran. The spermatogonial clones arise from cells that resemble the Ais spermatogonia but have a larger nuclear diameter. In spite of their mitotic activity these 'repopulating stem cells' lie mainly isolated or in pairs. This explained by migration and differentiation. Migration appeared to occur at random in all directions along the basement membrane of the seminiferous tubule. After one or more divisions of the stem cells, a second type of cell appears, which is called the 'differentiating spermatogomium'. The time elapsing before this type of cell appears, depends on the dose of Myleran: the larger the dose the later differentiation starts. A relation could be demonstrated between the stage of the cycle of the seminiferous epithelium and the start of differentiation. Differentiating cells were found isolated or in groups of two, four, eight or sixteen cells. Hence we concluded that at least up to their fourth division differentiating cells divide synchronously without degenerations. Three types of division of repopulating stem cells were distinguished, producing (1) two repopulating stem cells, (2) one repopulating stem cell and one cell starting spermatogonial differentiation, or (3) two differentiating cells. Type 1 divisions were found most frequently.

Animals↗

PubMed1989

Angiotensin II receptors labelled with 125I-[Sar1, Ile8]-AII in albino rabbit ocular tissues.

High affinity binding sites for the angiotensin II antagonist 125I-[Sar1,Ile8]-AII have been identified and characterized in membrane suspensions of ocular tissues of albino rabbits. Scatchard analysis of the binding indicated a single class of sites with Kd values of 186, 92, 152, 50, 102 pM for the iris + ciliary body, choroid, ciliary process, retina and cornea, respectively. The corresponding concentrations of binding sites were 22, 68, 35, 22 and 4 fmole/mg of protein. The order of potency for several AII analogs to compete with 125I-[Sar1,Ile8]-AII at its binding sites in iris + ciliary body membranes ([Sar1,Leu8]-AII = [Sar1,Ile8]-AII greater than AII = [Sar1, Ala8]-AII greater than AIII greater than AI) resembled the order of potency found for AII receptors in other tissues. The competition curves for this tissue using AII and AIII were best explained by the existence of two populations of binding sites. The addition of the guanine nucleotide, GppNHp, to the assay resulted in a 6.7-fold and 2.3-fold decrease in the respective affinities of AII and AIII for 125I-[Sar1,Ile8]-AII binding sites without a change in the slope of the competition curves. The GppNHp-induced effect was also observed in ciliary process membranes but not in retinal or choroidal membranes. These results indicate the presence of AII receptors regulated by a GTP-binding protein in both the ciliary process and the iris + ciliary body of the rabbit. They also suggest a difference in the guanine nucleotide regulation of AII receptors in different ocular tissues.

Angiotensin II↗

PubMed1992

Lipids and lipoproteins in persons with Down's syndrome.

This study was designed to investigate whether the observed decreased prevalence of coronary artery disease in individuals with Down's syndrome may be explained by their serum lipid and lipoprotein profiles. Twenty-seven persons with Down's syndrome and 23 non-affected control individuals were enrolled in this study. Their fasting venous blood was analysed for total cholesterol, triglyceride, LDH cholesterol, HDL cholesterol, apo B and apo AI. The results revealed no significant differences between the study and control group with regard to total cholesterol, LDL cholesterol, apo B and the apo B:apo AI ratio. However, triglyceride levels were significantly increased, and serum HDL cholesterol, apo AI and HDL cholesterol:total cholesterol ratio were significantly decreased in patients with Down's syndrome when compared with the control group. The latter observations are all associated with an increased risk for coronary artery disease. Therefore, it is concluded that the decreased prevalence of coronary artery disease in individuals with Down's syndrome cannot be explained by the lipid and lipoprotein levels observed in this study population.

Adolescent↗

PubMed2026

Precision Medicine in Transfusion-Dependent and Non-Transfusion-Dependent β-Thalassemia: Toward Personalized Diagnosis and Therapy.

β-thalassemia comprises a clinically heterogeneous group of disorders in which anemia severity, transfusion exposure, iron loading, and organ complications vary widely among individuals. This structured narrative review summarizes practical applications of precision medicine in transfusion-dependent thalassemia (TDT) and non-transfusion-dependent thalassemia (NTDT), with explicit attention to which strategies apply to each clinical category. Literature indexed in PubMed and Scopus from 2000 to 2025 was reviewed using terms related to thalassemia, precision medicine, magnetic resonance imaging (MRI), chelation tailoring, next-generation sequencing (NGS), fetal hemoglobin (HbF) modifiers, luspatercept, mitapivat, hepcidin, gene therapy, gene editing, and artificial intelligence (AI). Evidence was synthesized descriptively because interventions, outcomes, and populations were heterogeneous, and no pooled meta-analysis was performed. In TDT, precision care is centered on individualized transfusion planning, extended red-cell antigen matching, MRI-guided cardiac and hepatic iron monitoring, organ-directed chelation intensification, and selection of disease-modifying or curative approaches. In NTDT, precision care emphasizes accurate phenotype classification, MRI liver iron concentration, because serum ferritin may underestimate iron burden, selective chelation, surveillance for NTDT-specific complications, and individualized use of agents that improve anemia. Personalized chelation should include deferiprone, either alone or in combination, when cardiac iron is increased. Comprehensive molecular diagnosis should include HBB together with HBA1 and HBA2 assessment, while secondary and tertiary modifiers help explain phenotypic variability and complication risk. Hepcidin and growth differentiation factor 15 (GDF-15) are discussed as investigational biomarkers; transferrin saturation is not recommended for routine iron-overload assessment in thalassemia. AI currently has its strongest role in screening and diagnosis, whereas risk-stratification models remain exploratory. Equitable implementation requires standardized TDT/NTDT pathways, regional MRI and genomics access, longitudinal registries, and multidisciplinary interpretation.

Humans↗

PubMed1979

Quantitative estimation of the photosynthetic proton binding inside the thylakoids by correlating internal acidification to external alkalinisation and to oxygen evolution in chloroplasts.

The external alkalinisation delta pHe, or the rate of oxygen evolution vO2, of a suspension of envelope-free chlorplasts was correlated with their internal acidification, estimated from the transmembrane delta pHei. Knowing the external buffer value, the concentration of the total protons moved Hi was calculated from the delta pHe, measured with a glass electrode ([Hi] was also obtained from vO2), and the free proton concentration [Hi+] was determined from delta pHei, measured with 9-aminoacridine. This gives a ratio gamma i = theta [Hi]/theta [Hi+], which is independent of the thylakoids internal volume. Within a large pHi range, scanned by varying the light intensity, gamma i was kept reasonably constant; it was hardly sensitive to pHi. This apparent invariability implies a continuous change of the internal buffer value beta i with pHi, since beta i/gamma i = -2.3.....10pHi, a relationship which inlcudes neither the total concentration of protonizable groups [Ai] nor pKi. As gamma i approximately Ki[Ai]/(Ki + [Hi+i]2, to keep gamma i constant when pHi drops, pKi and [Ai] must increase. This may be achieved by a progressive unmasking of anionic functions, initially inaccessible in the membrane. The relative slowness of this process may explain why gamma i calculated from the initial kinetics was sometimes smaller in high than in low light, where it always equalled that measured from the steady-state amplitude at all intensities. A small deficit of [Hi+] deduced from what could have been expected from delta pHe may reflect a limited binding of protons in the membrane itself, about 1 H+ for 30--130 chlorophylls (gamma i could be between 70 and 240, more frequently around 100); these numbers varied depending on the samples, but were constant for a given preparation.

Chloroplasts↗

PubMed1992

[Artificial intelligence methods for support of medial patient education before surgical interventions in the region of the neck-nose-ear].

As a rule, curative operations require the patient's consent. Determined by the expansion of surgical possibilities, the kind and frequency of specific complications are subjected to constant changes. The physician is encouraged to explain therapeutic methods as well as the probability of complications within the patient's grasp. It has been investigated to what extent methods of artificial intelligence (AI) are suited for assisting the physician in this task. For this purpose, a comprehensive list of surgical complications as reported in research literature has been compiled. The list has been transferred into a hierarchical structure which can be depicted as a rule tree classified according to topographic aspects. In each otolaryngological operation, the reported complications can be classed with these rules. By employing an expert system (Fig. 1), the physician is capable of compiling an individualized document of agreement (Fig. 2) which serves as a basis for the explanatory talk with the patient.

Artificial Intelligence↗

PubMed2025

Modeling unknowns: A vision for uncertainty-aware machine learning in healthcare.

The integration of machine learning (ML) into healthcare is accelerating, driven by the proliferation of biomedical data and the promise of data-driven clinical support. A key challenge in this context is managing the pervasive uncertainty inherent in medical reasoning and decision-making. Despite its recognized importance, uncertainty is often underrepresented in the design and evaluation of clinical AI systems. Here we report an editorial overview of a special issue dedicated to uncertainty modeling in medical AI, which gathers theoretical, methodological, and practical contributions addressing this critical gap. Across these works, authors reveal that fewer than 4% of studies address uncertainty explicitly, and propose alternative design principles-such as optimizing for clinical net benefit or embedding explainability with confidence estimates. Notable contributions include the RelAI system for real-time prediction reliability, empirical findings on how uncertainty communication shapes clinical interpretation, and benchmarks for out-of-distribution detection in tabular data. Furthermore, this issue highlights the use of causal reasoning and anomaly detection to enhance system robustness and accountability. Together, these studies argue that representing, communicating, and operationalizing uncertainty are essential not only for clinical safety but also for building trust in AI-driven care. This special issue thus repositions uncertainty from a limitation to a foundational asset in the responsible deployment of ML in healthcare.

Machine Learning↗

PubMed1992

Estimates of genetic trend in an artificial insemination progeny test program and their association with herd characteristics.

Individual lactation records from Holstein cows in 3449 herds participating in an AI stud's young sire sampling program from 1971 to 1987 were used to characterize the sampling program and to estimate genetic merit and trend. Average genetic merit of cows in sampling program herds was consistently superior to the average genetic merit of cows in the US population. Genetic trend of sires of first-crop cows was 58 kg of milk and 1.5 kg of fat/yr from 1971 to 1978 and 176 kg of milk and 5.5 kg of fat/yr from 1979 to 1987. The average genetic merit of sires of first-crop cows born after 1983 was equivalent to or exceeded the genetic level of sires of other cows in the herd. Within-herd-year means and standard deviations of yield, genetic evaluation, and management traits (herd-year characteristics) were computed for a subset of 341 herds contributing first-crop daughters for at least 10 yr. The average of each herd-year characteristic during 10 or more years was used to predict within-herd genetic trend. Herd characteristics explained up to 51% of differences in within-herd genetic trends. Average sire genetic merit of daughters other than first-crop daughters accounted for up to 80% of the explained differences. Other herd characteristics suggested that herds with larger within-herd standard deviation milk yields, a larger number of young sires represented, younger cows, and greater percentage of cows sired by AI sires made greater genetic improvement. Results indicated that the average genetic merit of cows and the rate of within-herd genetic improvement are higher in herds that participate in a young sire sampling program.

Animals↗

PubMed1972

Comparison of the multiple deoxyribonucleic acid-dependent ribonucleic acid polymerase forms of whole rat liver and a minimal-deviation rat hepatoma cell line.

To investigate the possibility that the pattern of multiple DNA-dependent RNA polymerases of an animal cell exerts a controlling influence on its nature, the activities of these enzymes were compared in differentiated rat liver and in a rapidly growing minimal-deviation rat hepatoma cell line by using established techniques of enzyme extraction, separation and determination. Relative to the DNA content of the tissues, RNA polymerase activities of forms AI, AII and B were approx. ninefold, twofold and twofold higher respectively in the cell line than in the liver. Tests indicated that these results could not be explained by differences in extraction efficiency or by the presence of unbound inhibitors or stimulators of polymerase activity in the final enzyme preparations. New forms of the enzyme were not detected in either tissue. The significance of these findings with respect to the possible role of multiple RNA polymerases in the control of cellular activities is discussed.

Animals↗

PubMed1977

Unexplained hyperinsulinemia in normal and "prediabetic" Pima Indians compared with normal Caucasians. An example of racial differences in insulin secretion.

The pattern of insulin response to oral and/or intravenous glucose has been claimed to be characteristic of diabetes and even prediabetes. To determine if differences in insluin secretion might explain the exceptionally high prevalence of diabetes in the Pima Indians, 26 genetically normal Pimas (nondiabetic offspring of nondiabetic parents), 32 genetically prediabetic Pimas (nondiabetic offspring of diabetic parents), 10 diabetic Pimas, and 29 normal Caucasians were studied. All subjects received an intravenous glucose tolerance test (IVGTT) to examine the acute-phase insulin response, and all nondiabetic subjects received an oral glucose tolerance test (OGTT) and arginine infusion (AI). The prediabetics also received a cortisone-primed oral glucose tolerance test (CGTT) and were classified by the result of this test. While acute-phase insulin release during the IVGTT was absent in the diabetics, there was a rapid response in all nondiabetics. Prediabetic Pimas with normal or abnormal CGTT had insulin levels similar to normal Indians during the IVGTT, OGTT, and AI. Thus, no evidence of impairment of acute- or late-phase insulin release was found. The normal and prediabetic Indians had fasting and stimulated insulin levels during all the tests two-to-threefold greater than the Caucasians. Differences in insulin levels between the two races could not be explained by differences in glucose level, age, or obesity.

Adolescent↗

PubMed1985

Changes in human high density lipoproteins in patients with extra-hepatic biliary obstruction.

Plasma high density lipoproteins (HDL) from patients with obstruction of the common bile duct were studied by crossed immunoelectrophoresis and isoelectric focusing. All cholestatic HDL fractions were rich in phospholipids (51.5 +/- 9%) with high proportions of free cholesterol (13.8 +/- 1.5%). Moreover, crossed immunoelectrophoresis of sera against anti-Apo A revealed the presence of multiple immunoprecipitates sharply contrasting with the pattern formed by normal sera. Tandem crossed immunoelectrophoresis against anti-Apo A and anti-Apo B was performed with whole serum and with the HDL fraction from various cholestatic subjects. Crossed identity was observed for two of these precipitates, which could be explained by the decrease in HDL stability due to the detergent effect of bile salts. The most noteworthy changes found in cholestatic patients appeared to be the apolipoprotein pattern of HDL. Both Apo AI (48%) and Apo AII (5.5%) were greatly diminished and Apo E was present in remarkably high amounts (39%) with two additional isoforms (Apo E'1 and Apo E'2).

Apolipoproteins A↗

PubMed1992

The von Rosen splint compared with the Frejka pillow. A study of 408 neonatally unstable hips.

101 children in Tromsö, Norway, treated with the Frejka pillow for 4.5 months because of neonatal hip instability (NHI) were compared with 307 children in Malmö, Sweden, treated with the von Rosen splint for 3 months. The pelvic radiographs, taken when the treatment was terminated, were assessed by the acetabular index (AI) and the cases of failure were evaluated. The AI showed no difference between the two groups. The Frejka group had 4 patients who received further treatment because of remaining acetabular dysplasia and/or subluxation while the von Rosen group had none. The difference in risk of failure might partly be explained by different criteria for failure.

Female↗

PubMed1992

Plasma lipid changes in psoriatic children.

Plasma lipid, lipoprotein and apoprotein concentrations were determined in psoriatic children and in healthy controls. The plasma total cholesterol (TC) was higher than in controls (p = 0.03). The higher cholesterol levels were explained by an almost significant increase in cholesterol associated with high-density lipoproteins and by the tendency towards higher values of cholesterol associated with low-density lipoproteins and very-low-density lipoproteins. No significant changes of plasma triglycerides were observed. Levels of apoproteins (apo) AI and apo B were not significantly different in psoriatic children; however, the levels of apo B were correlated differently with plasma TC in psoriatic children, and the ratio TC/apo B was significantly increased in patients with respect to the controls, suggesting some differences of plasma lipoprotein lipid/apoprotein relationship in psoriasis.

Adolescent↗

PubMed1985

Apolipoprotein AIMilano. Accelerated binding and dissociation from lipids of a human apolipoprotein variant.

The lipid binding properties of apolipoprotein (apo) AIMilano, a molecular variant of human apolipoprotein AI, characterized by the Arg173----Cys substitution, was investigated by the use of dimyristoylphosphatidylcholine liposomes. Both the variant AIMilano and normal AI are incorporated to the same extent in stable complexes isolated by gel filtration, showing similar dimensions and stoichiometries. A higher affinity of apo-AIMilano for dimyristoylphosphatidylcholine is suggested by the faster association rate of the variant apoprotein compared to normal AI; similarly, apo-AIMilano is more readily displaced by guanidine hydrochloride from the isolated dimyristoylphosphatidylcholine-apoprotein complexes. When the secondary structure of apo-AIMilano was investigated by spectrofluoroscopy and circular dichroism, a higher fluorescence peak wavelength and a lower alpha-helical content were detected in the variant apoprotein compared to normal AI. The substitution Arg173----Cys in the AIMilano dramatically alters the amphipathic nature of the modified alpha-helical fragment of apoprotein AI. The association rate with lipids is accelerated by an increased exposure of hydrophobic residues. The reduced stability of the lipid-apoprotein particles is possibly mediated by a reduction in the number of helical segments involved in lipid association. The high flexibility of the AIMilano apolipoprotein in the interaction with lipids may explain its accelerated catabolism and the possibly improved uptake capacities for tissue lipids.

Amino Acid Sequence↗