Support of deep learning to classify vocal fold images in flexible laryngoscopy

PURPOSE

To collect a dataset with adequate laryngoscopy images and identify the appearance of vocal folds and their lesions in flexible laryngoscopy images by objective deep learning models.

METHODS

We adopted a number of novel deep learning models to train and classify 4549 flexible laryngoscopy images as no vocal fold, normal vocal folds, and abnormal vocal folds. This could help these models recognize vocal folds and their lesions within these images. Ultimately, we made a comparison between the results of the state-of-the-art deep learning models, and another comparison of the results between the computer-aided classification system and ENT doctors.

RESULTS

This study exhibited the performance of the deep learning models by evaluating laryngoscopy images collected from 876 patients. The efficiency of the Xception model was higher and steadier than almost the rest of the models. The accuracy of no vocal fold, normal vocal folds, and vocal fold abnormalities on this model were 98.90 %, 97.36 %, and 96.26 %, respectively. Compared to our ENT doctors, the Xception model produced better results than a junior doctor and was near an expert.

CONCLUSION

Our results show that current deep learning models can classify vocal fold images well and effectively assist physicians in vocal fold identification and classification of normal or abnormal vocal folds.

Histamine-releasing and allergenic properties of opioid analgesic drugs: resolving the two

Opioid analgesics are amongst the most commonly administered drugs in hospitals. Whether natural or synthetic, they show some common structural features, morphine-like pharmacological action and binding specificity for complementary opioid receptors. Tramadol differs from the other opioid analgesics in possessing monoaminergic activity in addition to its affinity for the µ opioid receptor. Many opioids are potent histamine releasers producing a variety of haemodynamic changes and anaphylactoid reactions, but the relationship of the appearance of these effects to the histamine plasma concentration is complex and there is no direct and invariable relationship between the two. Studies of the histamine-releasing effects, chiefly centred on morphine, reveal variable findings and conclusions often due to a range of factors including differences in technical measurements, dose, mode of administration, site of injection, the anatomical distribution of histamine receptors and heterogeneity of patient responses. Morphine itself has multiple direct effects on the vasculature and other haemodynamically-active mediators released along with histamine contribute to the variable responses to opioid drug administration. Despite their heavy use and occasional apparent anaphylactic-like side-effects, immunoglobulin E antibody-mediated immediate hypersensitivity reactions to the drugs are not often encountered. Uncertainties associated with skin testing with these known histamine-releasers, and the general unavailability of opioid drug-specific immunoglobulin E antibody tests contribute to the frequent failure to adequately investigate and establish underlying mechanisms of reactions by distinguishing anaphylactoid from true anaphylactic reactions. Clinical implications for diagnosis of reactions and some speculations on the rarity of true Type 1 allergies to these drugs are presented.

Drug-specific cyclodextrins with emphasis on sugammadex, the neuromuscular blocker rocuronium and perioperative anaphylaxis: implications for drug allergy

Cyclodextrins, oligosaccharides linked in a circular arrangement around a central cavity, are used extensively in the pharmaceutical industry to improve drug delivery. Their usefulness depends on their capacity to form a drug inclusion, or host-guest, complex within the cavity. In an attempt to improve the delivery of the widely used neuromuscular blocking drug (NMBD) rocuronium, a rocuronium inclusion complex was formed with a chemically modified γ-cyclodextrin. The high binding affinity and specificity of the modified carrier (named sugammadex) for rocuronium (and other aminosteroid NMBDs) led to its use in anaesthesia as an innovative and useful agent for rapid reversal of rocuronium-induced neuromuscular block by sequestering the drug as an inclusion complex. This, in turn, led to the suggestion that sugammadex might be useful to remove the NMBD from the circulation of patients experiencing rocuronium-induced anaphylaxis, a suggestion subsequently supported in case reports where traditional treatment had failed. Successful resuscitations suggested that sugammadex might be a valuable new treatment for such intractable cases but, given the inappropriateness of clinical trials, confirmation or refutation will have to await the slow accumulation of results of individual case reports. Important questions related to antibody accessibility of drug allergenic structures on the rocuronium-sugammadex inclusion complex, and the competition between sugammadex and IgE antibodies (both free and cell bound) for rocuronium, also remain and can be investigated in vitro. The sugammadex findings indicate that the use of carrier molecules such as the cyclodextrins to improve drug delivery will sometimes give rise to changed immunologic and allergenic behaviour of some drugs and this will have to be taken into account in preclinical drug safety assessments of drug-carrier complexes. The possibility of encapsulating and removing other allergenic drugs, e.g., penicillins and cephalosporins, in cases of difficult-to-reverse anaphylaxis to these drugs is discussed.

Studies on the mechanism of multiple drug allergies. Structural basis of drug recognition

The multiple drug allergy syndrome, that is, allergic recognition of a variety of drugs that may be both pharmacologically and structurally different, has been little studied and, consequently, the underlying mechanism(s) is unknown. The molecular basis of drug recognition by IgE antibodies found in the sera of subjects exhibiting multiple allergic drug sensitivities was studied by direct binding and quantitative hapten inhibition assays in experiments employing a wide range of carefully selected drugs and other chemicals. Drug recognition was shown to be related to the presence of tertiary and quaternary mono-, di- and trialkyl amino groups, but only if the alkyl groups were 'small' viz., methyl or, perhaps, ethyl. Primary, secondary, and tertiary (with R = 'large' alkyl) groups showed no direct antibody binding or antibody inhibitory activities. Near-neighbour effects of amide and hydroxyl groups appeared to promote weaker antigenic recognition. Results indicate that the antibody recognition and clinical drug allergy spectra of at least some subjects with multiple drug allergies are due to wide ranging immunological cross-reactivities with drugs containing tertiary amino and quaternary ammonium groups which are present in many different pharmacologically active agents. Separate populations of antibodies to other non-cross reacting drugs, for example, beta-lactam antibiotics, may also be present in the sera of such subjects.

Chemistry of drug allergenicity

Of the very large number and variety of drugs used in medicine, those that are frequently implicated in immediate allergic reactions are relatively small in number and include neuromuscular blocking drugs used in anaesthesia, beta-lactam antibiotics, some other antibacterial agents including broad-spectrum antibiotics and quinolones and, less often, some narcotics. Structure-activity and immunochemical investigations have been most numerous and detailed for neuromuscular blocking drugs and beta-lactams, particularly penicillins. For the former group of drugs, morphine is proving to be a useful agent for the in-vitro detection of clinically relevant neuromuscular blocking drug-- as well as morphine- and fentanyl-reactive IgE antibodies. The employment of so-called 'major' and 'minor' determinants for a range of different penicillins and cephalosporins has revealed previously unsuspected heterogeneity in patient recognition responses, and has reinforced findings on the allergenic importance of side-chain groups. Many reports have been published on anaphylaxis to chlorhexidine, and progress in identifying allergenic determinants is reviewed together with the still inadequately understood subject of IgE antibody recognition of quinolone antibacterial agents.

Structural determinants of antibiotic allergy

Allergies to antibiotics, mainly the beta-lactam antibiotics (penicillins and cephalosporins), are a common, costly, and potentially dangerous clinical problem encountered in everyday practice. Although studies on the role of non-beta-lactam antibiotics in allergic diseases, particularly the development of specific diagnostic tests and the immunochemical identification of allergenic structures, have been too few and relatively superficial, the situation with the beta-lactam antibiotics is much more advanced. Good progress has been made in identifying the spectra of allergenic determinants recognized by IgE antibodies in the sera of subjects sensitized to penicillins and cephalosporins, and this is aiding the development of an appropriate battery of drug conjugates for use as diagnostic agents. Patient-sensitivity responses to the beta-lactam antibiotics are frequently heterogeneous, and this factor must be taken into account for any diagnostic strategy or future therapy with a penicillin or cephalosporin.

Anaphylaxis to chlorhexidine. Case report. Implication of immunoglobulin E antibodies and identification of an allergenic determinant

BACKGROUND

There are many reports of allergic reactions, including anaphylaxis, following exposure to chlorhexidine. Reactions may occur via contact with the skin and mucous membranes or from catheters treated with the antibacterial agent. Apart from implicating chlorguanide in immunoglobulin (Ig) E antibody-binding studies on serum from an anaphylactic patient, little work has been done on the molecular basis of recognition of the agent in sensitive subjects.

OBJECTIVES

The molecular basis of IgE-binding to chlorhexidine was closely examined with the view of defining its fine structural recognition features by antibodies from a subject who experienced anaphylaxis following contact with the antiseptic.

METHODS

Tryptase determinations, different drug-solid phases, immunoassays and quantitative hapten inhibition studies with chlorhexidine and selected structural analogues were employed together with serum from the anaphylactic patient. Results were analysed to define the complete drug allergenic determinant and to identify the important structural features complementary to the IgE antibody combining sites.

RESULTS

The subject's serum tryptase levels sampled after the reaction were elevated and employment of a chlorhexidine-EA Sepharose solid phase showed the presence of serum IgE antibodies to the drug. Lack of inhibition by 4-chlorophenol and other selected substituted phenyl compounds showed that the terminal groups at each end of the chlorhexidine molecule, alone, did not account for antibody recognition of the antibacterial agent. Although chlorguanide and alexidine, the structures of which each comprise part of the chlorhexidine molecule, showed significant inhibition of the binding of IgE antibodies to chlorhexidine, neither compound was as potent an inhibitor as chlorhexidine itself. Two molecules of chlorguanide make up the symmetrical molecule of chlorhexidine while the interior structure of alexidine (that is excluding the terminal 2-ethylhexyl groups) is identical to part of the chlorhexidine molecule.

CONCLUSIONS

Taken together, for this patient, these results lead to the conclusion that the whole chlorhexidine molecule is complementary to the IgE antibody combining sites and that the 4-chlorophenol, biguanide and hexamethylene structures together comprise the allergenic determinant. Hence, like one of the trimethoprim determinants identified, but unlike most drug allergenic determinants identified so far, the chlorhexidine allergenic determinant identified here encompasses the entire molecule.

Fine structural specificity differences of trimethoprim allergenic determinants

BACKGROUND

Adverse reactions, including immediate hypersensitivity, to the widely used antibacterial agent trimethoprim occur quite frequently. In recent years some progress has been made in developing an immunoassay to aid diagnosis of type 1 allergic reactions to trimethoprim and to define the basis of IgE antibody recognition of the drug.

OBJECTIVES

The molecular basis of IgE binding to trimethoprim was examined more closely with a view to defining the fine structural recognition differences between patient's sera. Utilization of such information may lead to immunoassays that are more specific and sensitive and of greater diagnostic value.

METHODS

Immunoassays for specific IgE antibodies and quantitative hapten inhibition studies with trimethoprim and selected structural analogues were employed, together with sera from eight subjects clearly defined clinically as allergic to trimethoprim.

RESULTS

Three different allergenic determinant structures have been identified on the trimethoprim molecule. Identification of the 3,4-dimethoxybenzyl group as a determinant was achieved on the basis of inhibitory activities of diaveridine, 3,4-dimethoxyphenylethylamine, 3,4-dimethoxybenzoic acid and 3,4,5-trimethoxycinnamic acid. Evidence that the opposite end of the trimethoprim molecule was not being recognized was obtained from results with some pyrimidine derivatives, each of which showed no activity. Identification of the second determinant, the 2,4-diamino-5-(3',4'-dimethoxybenzyl) pyrimidine group, rested mainly on the superior inhibitory potency of diaveridine, which differs from trimethoprim by just one methoxy group. With sera from some trimethoprim-allergic subjects, only trimethoprim was active, suggesting that the entire molecule was a third IgE-binding determinant structure.

CONCLUSION

As with other drug allergenic determinants defined so far, heterogeneity of trimethoprim IgE-binding determinants exists, and fine structural differences between determinants may be as small as a single methoxy group. Identification of the 2,4-diamino-5-(3',4'-dimethoxybenzyl) pyrimidine group as an allergenic determinant increases the number of known trimethoprim determinants to three, and suggests that the number and heterogeneity of determinants will be a reflection of the number of allergic subjects studied.

beta-Lactam drug allergens: fine structural recognition patterns of cephalosporin-reactive IgE antibodies

Lack of experimental findings on the spectrum of cephalosporin allergenic determinants has hindered diagnosis of adverse reactions to these drugs and retarded understanding of allergenic cross-reactions between cephalosporins and between cephalosporins and penicillins. Subjects allergic to the widely used cephalosporin antibiotic cefaclor have serum immuno globulin (Ig) E antibodies that react with the drug. Quantitative hapten inhibition studies employing sera from subjects allergic to cefaclor revealed fine structural recognition differences between the combining site specificities of cefaclor-reactive IgE antibodies in the sera of different subjects. Unlike penicillins, where discrete side chain or thiazolidine ring determinants alone may be recognized, IgE binding determinants on cefaclor encompassed the entire molecule. Fine structural recognition specificity differences at positions R1 (side-chain) and R2 (substituent attached to dihydrothiazine ring) were detected between IgE antibodies in different sera. Some antibodies showed clear preferential recognition of the aminobenzyl group at position R1 and Cl at R2 while with others, a greater degree of recognition tolerance was seen at R1 where, for example, the aminohydroxybenzyl or aminodihydrobenzyl groups were recognized, and at R2 where a methyl or even an ester group was tolerated. As with the penicillins, cephalosporins as allergens cannot simply be considered as a group of compounds with a common allergenic determinant structure. IgE antibodies that bind to cefaclor show great heterogeneity indicated by clear, fine structural differences in recognition of the R1 and R2 groups on the drug.

Allergenic relationship between taxonomically diverse pollens

BACKGROUND

Skin tests and tests for IgE antibodies show that subjects are usually sensitive to a number of different pollens, frequently from taxonomically diverse species which are assumed to be allergenically non-crossreactive. This suggests that the presence of IgE antibody-reactivity to an individual pollen may not necessarily have resulted from contact with that pollen or even with a taxonomically closely related species.

OBJECTIVE

Since this has important consequences for allergen avoidance and desensitization of patients, we attempted to define allergenic relationships between diverse pollen species.

METHODS

Sera from subjects were examined in direct IgE antibody binding experiments and by quantitative inhibition, protein blotting and adsorption and elution studies.

RESULTS

Sera from subjects diagnosed as allergic to white cypress pine, Italian cypress, ryegrass or birch pollen were shown to have IgE antibodies that reacted with pollens from these four species and from cocksfoot, couch grass, lamb's quarter, wall pellitory, olive, plantain and ragweed. These reactions were confirmed in protein blotting and adsorption and elution studies where numerous IgE-binding bands were detected in all 11 different pollen extracts with sera from each of the different allergic categories. Further evidence of allergenic (i.e. IgE-binding crossreactivity between the different pollens was provided by inhibition studies in which clear-cut inhibitions of IgE binding to the different pollen allergen discs were obtained with comparable amounts of the different pollen extracts.

CONCLUSION

We conclude that the presence of pollen reactive IgE antibodies may not necessarily be a true reflection of sensitizing pollen species.

Detection and side-chain specificity of IgE antibodies to flucloxacillin in allergic subjects

Unlike studies on the antigenicity of penicillins in laboratory animals, limited information is available on the allergenicity of penicillins in man, especially with regard to fine structural allergenic differences between the many different penicillins. Inconsistent with the earlier conclusions of others, our studies suggest that side-chain structures on the penicillin molecule are the major allergenic determinants in many reactions. Immediate allergic reactions to flucloxacillin were observed in a number of patients where diagnosis was confirmed by skin testing and detection of flucloxacillin-reactive IgE antibodies. Quantitative hapten inhibition studies revealed potent inhibition by flucloxacillin and three structurally related penicillins: oxacillin, cloxacillin and dicloxacillin. Analysis of the inhibition results showed that the side-chain group of flucloxacillin, 3-(2-chloro-6-fluorophenyl)-5-methyl-4-isoxazolyl, is recognized by some antibodies and that the 5-methyl-3-phenyl-4-isoxazolyl group, with or without halogen substituents, accounts for the reactivity of other antibodies and for the cross-reactions seen with some other penicillins. Since it is the side-chain group that distinguishes the many different penicillins, and since the side-chain groups are recognized by IgE antibodies in many of the allergic reactions, it is becoming clear that specific assays are required for the detection of IgE antibodies to each of the different penicillins.

Structure--activity studies on drug-induced anaphylactic reactions

As a result of recent research and development, an expanding range of in vitro assays for the detection of drug-reactive IgE antibodies is now available for the diagnosis of individual drug allergies. Continuing immunochemical studies on drugs implicated in allergic reactions via further development of specific immunoassays and production of drug-protein conjugates will allow us to build up a picture of the repertoire of drug and drug-derived B-cell allergenic determinant structures and to identify and predict cross-reactivities. Although we have made good progress in identifying some drug B-cell determinants, we remain ignorant of drug T-cell determinants and, indeed, of the nature of T-cell recognition of all nonpeptide molecules. Demonstration that drugs specifically recognize T-cells from drug-allergic patients may reveal associations with HLA phenotypes, the nature and location of interaction between drug and MHC molecules, and the nature and identity of drug or drug-derived T-cell antigens. A knowledge of the molecular nature of T-cell determinants inducing allergic responses is fundamental to therapeutic attempts to modulate these deleterious reactions. The possibility now exists for the allergenic screening of drugs as part of their toxicological evaluation, and identification of allergenic structures on drugs also has implications for the testing of allergenic activity and sensitizing potential of other chemicals in our environment. Obvious and important areas for the application of our methods and findings are the pharmaceutical and cosmetic industries.

Cypress pollen allergy. Identification of allergens and crossreactivity between divergent species

Studies employing sera from 34 subjects allergic to white cypress pine (Callitris glaucophylla) pollen identified 18 IgE antibody-binding components in the pollen of this species, five of which (MWs approximately 94, 68, 64, 43 and 34 kDa) were recognized by all of the sera. Protein blotting and quantitative inhibition studies revealed clear cross-reactivity between C. glaucophylla and Cupressus sempervirens pollen proteins and striking similarities in the IgE recognition band patterns of the two pollens. Inhibition experiments with other pollen extracts revealed that sera from C. glaucophylla pollen-allergic subjects can be divided into two groups--those inhibited only by extracts from the two Cupressaceae pollens and those inhibited both by these pollen proteins and by pollen extracts from other species. Most of the crossreactions in the latter group cannot be explained on the basis of taxonomic relationships or separate sensitizations. As with previous studies on birch and olive pollens, we conclude that pollen allergenic crossreactivity is much more wide-ranging than generally believed.