Artificial intelligence in gastroenterology: A state-of-the-art review

The development of artificial intelligence (AI) has increased dramatically in the last 20 years, with clinical applications progressively being explored for most of the medical specialties. The field of gastroenterology and hepatology, substantially reliant on vast amounts of imaging studies, is not an exception. The clinical applications of AI systems in this field include the identification of premalignant or malignant lesions (e.g., identification of dysplasia or esophageal adenocarcinoma in Barrett’s esophagus, pancreatic malignancies), detection of lesions (e.g., polyp identification and classification, small-bowel bleeding lesion on capsule endoscopy, pancreatic cystic lesions), development of objective scoring systems for risk stratification, predicting disease prognosis or treatment response [e.g., determining survival in patients post-resection of hepatocellular carcinoma), determining which patients with inflammatory bowel disease (IBD) will benefit from biologic therapy], or evaluation of metrics such as bowel preparation score or quality of endoscopic examination. The objective of this comprehensive review is to analyze the available AI-related studies pertaining to the entirety of the gastrointestinal tract, including the upper, middle and lower tracts; IBD; the hepatobiliary system; and the pancreas, discussing the findings and clinical applications, as well as outlining the current limitations and future directions in this field.     

Measuring disease activity and functional status in patients with scleroderma and Raynaud's phenomenon

OBJECTIVE: To document disease activity and functional status in patients with scleroderma (systemic sclerosis [SSc]) and Raynaud's phenomenon (RP) and to determine the sensitivity to change, reliability, ease of use, and validity of various outcome measures in these patients. METHODS: Patients with SSc and moderate-to-severe RP participating in a multicenter RP treatment trial completed daily diaries documenting the frequency and duration of RP attacks and recorded a daily Raynaud's Condition Score (RCS). Mean scores for the 2-week periods prior to baseline (week 0), end of trial (week 6), and posttrial followup (week 12) were calculated. At weeks 0, 6, and 12, physicians completed 3 global assessment scales and performed clinical assessments of digital ulcers and infarcts; patients completed the Health Assessment Questionnaire (HAQ), the Arthritis Impact Measurement Scales 2 (AIMS2) mood and tension subscales, 5 specific SSc/RP-related visual analog scales (VAS), and 3 other VAS global assessments. We used these measures to document baseline disease activity and to assess their construct validity, sensitivity to change, and reliability in trial data. RESULTS: Two hundred eighty-one patients (248 women, 33 men; mean age 50.4 years [range 18-82 years]) from 14 centers participated. Forty-eight percent had limited cutaneous SSc; 52% had diffuse cutaneous SSc. Fifty-nine patients (21%) had digital ulcers at baseline. Patients had 3.89 +/- 2.33 (mean +/- SD) daily RP attacks (range 0.8-14.6), with a duration of 82.1 +/- 91.6 minutes/attack. RCS for RP activity (possible range 0-10) was 4.30 +/- 1.92. HAQ scores (0-3 scale) indicated substantial disability at baseline (total disability 0.86, pain 1.19), especially among the subscales pertaining to hand function (grip, eating, dressing). AIMS2 mood and tension scores were fairly high, as were many of the VAS scores. Patients with digital ulcers had worse RCS, pain, HAQ disability (overall, grip, eating, and dressing), physician's global assessment, and tension, but no significant difference in the frequency of RP, duration of RP, patient's global assessment, or mood, compared with patients without digital ulcers. VAS scores for digital ulcers as rated by the patients were not consistent with the physician's ratings. Factor analysis of the 18 measures showed strong associations among variables in 4 distinct domains: disease activity, RP measures, digital ulcer measures, and mood/tension. Reliability of the RCS, HAQ pain and disability scales, and AIMS2 mood and tension subscales was high. The RP measures demonstrated good sensitivity to change (effect sizes 0.33-0.76). CONCLUSION: Our findings demonstrate that the significant activity, disability, pain, and psychological impact of RP and digital ulcers in SSc can be measured by a small set of valid and reliable outcome measures. These outcome measures provide information beyond the quantitative metrics of RP attacks. We propose a core set of measures for use in clinical trials of RP in SSc patients that includes the RCS, patient and physician VAS ratings of RP activity, a digital ulcer/infarct measure, measures of disability and pain (HAQ), and measures of psychological function (AIMS2).     

Comparison of False-Positive Reactions in Direct-Binding Anti-HIV ELISA Using Cell Lysate or Recombinant Antigens

In a 2-year study of false-positive anti-HIV-1 tests in blood donors at Manchester and Lancaster Blood Banks, the reactions associated with a HIV-infected cell lysate antigen were compared with those using recombinant-antigen-based tests. In year 1 (cell lysate test) at Manchester BTS 0.21% of 119.178 donations were repeatedly reactive, compared with 0.53% of 119,004 donations in year 2 (recombinant antigen). Reactive sera were tested at Manchester PHL by three different immunoassays. Referred specimens were classified as anti-HIV positive (95-100% reactive in all the assays), equivocal or negative (negative results in all three immunoassays). Two donors were confirmed to be anti-HIV positive over the 2-year period. Most sera were negative by confirmatory immunoassays in years 1 and 2. In year 1, a study of 60 referred sera with sex- and age-matched controls showed high correlation between a reactive anti-HIV-1 screening test and indeterminate anti-HIV-1 patterns on Western blot showing reactions with HIV gag-coded proteins. In year 2, less than 10% of referred sera were reactive by Western blot, and there was no correlation between a reactive screening anti-HIV test, the strength of signal in the test or a reactive Western blot. Follow-up showed that donors whose sera were reactive in years 1 and 2 by the anti-HIV-1 screening test formed almost two different populations. Four donors with equivocal anti-HIV-1 confirmatory tests had anti-HIV 'envelope' reactions.(ABSTRACT TRUNCATED AT 250 WORDS)     

The eukaryotic host factor that activates exoenzyme S of Pseudomonas aeruginosa is a member of the 14-3-3 protein family.

Exoenzyme S (ExoS), which has been implicated as a virulence factor of Pseudomonas aeruginosa, catalyzes transfer of the ADP-ribose moiety of NAD+ to many eukaryotic cellular proteins. Its preferred substrates include Ras and several other 21- to 25-kDa GTP-binding proteins. ExoS absolutely requires a ubiquitous eukaryotic protein factor, termed FAS (factor activating ExoS), for enzymatic activity. Here we describe the cloning and expression of a gene encoding FAS from a bovine brain cDNA library and demonstrate that purified recombinant FAS produced in Escherichia coli activates ExoS in a defined cell-free system. The deduced amino acid sequence of FAS shows that the protein (245 residues, calculated molecular mass 27,743 Da) belongs to a highly conserved, widely distributed eukaryotic protein family, collectively designated as 14-3-3 proteins. Various functions have been reported for members of the 14-3-3 family, including phospholipase A2 activity and regulation of tyrosine hydroxylase, tryptophan hydroxylase, and, possibly, protein kinase C activities. Identification of FAS as a 14-3-3 protein establishes an additional function for this family of proteins--the activation of an exogenous ADP-ribosyltransferase. Elucidation of the precise role of FAS in activating ExoS will contribute to understanding the molecular mechanisms by which P. aeruginosa causes disease.     

The effect of magnesium trisilicate and kaolin on the in vivo absorption of chloroquine

An in vitro study indicated that certain antacids and adsorbents may decrease the oral availability of th two widely used antimalarial agents chloroquine and pyrimethamine. To determine if this data was applicable to the clinical (in vivo) situation, plasma levels of one of the antimalarial agents (chloroquine) were followed in six Negro--Arab volunteers both when given alone and when taken with separate doses of two of the implicated interactants (magnesium trisilicate and kaolin). This in vivo work confirmed the in vitro findings; chloroquine area under the plasma concentration-time curve data were decreased by both magnesium trisilicate (18.2%) and kaolin (28.6%). Similar results could be expected for pyrimethamine. It is suggested therefore, to avoid loss of drug, that the antimalarials should not be taken with gastrointestinal medications of this type or that their administration should be separated by at least 4 h to reduce the risk of them interacting in the gut, thus preventing drug adsorption to the antacids/adsorbents, and loss of systemic availability.     

Expression analysis in a rat psychosis model identifies novel candidate genes validated in a large case–control sample of schizophrenia

Antagonists of the N-methyl-D-aspartate (NMDA)-type glutamate receptor induce psychosis in healthy individuals and exacerbate schizophrenia symptoms in patients. In this study we have produced an animal model of NMDA receptor hypofunction by chronically treating rats with low doses of the NMDA receptor antagonist MK-801. Subsequently, we performed an expression study and identified 20 genes showing altered expression in the brain of these rats compared with untreated animals. We then explored whether the human orthologs of these genes are associated with schizophrenia in the largest schizophrenia genome-wide association study published to date, and found evidence for association for 4 out of the 20 genes: SF3B1, FOXP1, DLG2 and VGLL4. Interestingly, three of these genes, FOXP1, SF3B1 and DLG2, have previously been implicated in neurodevelopmental disorders.     

Effects on the buoyant density of rabbit platelets of ADP and agents that increase the concentration of cyclic AMP

Rabbit platelets were aggregated by adenosine diphosphate (ADP), allowed to deaggregate and then separated into density subpopulations by centrifugation through discontinuous Stractan density gradients. Although ADP causes little or no release of the contents of the amine storage granules of rabbit platelets, ADP caused a decrease in platelet density as compared with control platelets subjected to the same procedures except for exposure to ADP. The density change persisted for at least four hours. The apparent size of platelets stimulated with ADP increased initially, but returned to control values during a one-hour period. A similar decrease in platelet density was observed with an albumin density gradient. Under conditions in which aggregation did not occur in response to ADP with ethylenediaminetetraacetic acid (EDTA) in the medium, little or no decrease in platelet density was observed. Agglutination with polylysine did not change platelet density. Thus, not only agents such as thrombin and plasmin that cause the release of the contents of the platelet granules decrease platelet density, but ADP also has this effect. Platelets would be exposed to all of these stimuli during thromboembolic processes, and their effect on platelets may account for the decrease in platelet density observed previously in experiments with rabbits with indwelling aortic catheters. Agents that increase the concentration of cyclic AMP (cAMP) in platelets (PGE1, adenosine, dibutyryl cAMP, forskolin, and papaverine) also decreased platelet density. This effect persisted when the platelets were washed and resuspended in fresh medium and was also demonstrable in plasma. Platelet size was gradually increased by prostaglandin E1 (PGE1) which maintains platelets in a disc shape and does not cause the release of granule contents, indicating that the decrease in platelet density caused by PGE1 may be attributable to platelet swelling.     

Specific DNA sequences associated with the nuclear matrix in synchronized mouse 3T3 cells.

Eukaryotic chromatin appears to be organized into arrays of supercoiled loops anchored to the scaffolding structure of the mitotic chromosome core or to the nuclear matrix of interphase nuclei. To reveal whether specific DNA sequences are involved in this level of chromatin organization, we isolated and cloned a population of DNA molecules [average length of 150 base pairs (bp)] closely associated with the nuclear matrix after exhaustive DNase digestion and subsequent extensive protease digestion. The nuclear matrix was obtained from murine BALB/c 3T3 cells synchronized at the G1/S border of the cell cycle. We report the structure of two sequences, designated G4 and G5, which are highly enriched in the matrix DNA. Sequence G4, of 152 bp, contains three 31-bp direct head-to-tail repeats. An 11-bp sequence at the end of each repeat is homologous to the first large tumor antigen recognition site of human papova virus. Sequence G5, of 135 bp, consists of two well-defined domains, in which the first domain is a fragment of the B1 repetitive sequence. The results suggest the possibility that the loops of histone-depleted chromatin are connected to the scaffold of the nuclear matrix, with specific DNA sequences at the anchorage sites.