Spotting malignancies from gastric endoscopic images using deep learning.

BACKGROUND: Gastric cancer is a common kind of malignancies, with yearly occurrences exceeding one million worldwide in 2017. Typically, ulcerous and cancerous tissues develop abnormal morphologies through courses of progression. Endoscopy is a routinely adopted means for examination of gastrointestinal tract for malignancy. Early and timely detection of malignancy closely correlate with good prognosis. Repeated presentation of similar frames from gastrointestinal tract endoscopy often weakens attention for practitioners to result in true patients missed out to incur higher medical cost and unnecessary morbidity. Highly needed is an automatic means for spotting visual abnormality and prompts for attention for medical staff for more thorough examination. METHODS: We conduct classification of benign ulcer and cancer for gastrointestinal endoscopic color images using deep neural network and transfer-learning approach. Using clinical data gathered from Gil Hospital, we built a dataset comprised of 200 normal, 367 cancer, and 220 ulcer cases, and applied the inception, ResNet, and VGGNet models pretrained on ImageNet. Three classes were defined-normal, benign ulcer, and cancer, and three separate binary classifiers were built-those for normal vs cancer, normal vs ulcer, and cancer vs ulcer for the corresponding classification tasks. For each task, considering inherent randomness entailed in the deep learning process, we performed data partitioning and model building experiments 100 times and averaged the performance values. RESULTS: Areas under curves of respective receiver operating characteristics were 0.95, 0.97, and 0.85 for the three classifiers. The ResNet showed the highest level of performance. The cases involving normal, i.e., normal vs ulcer and normal vs cancer resulted in accuracies above 90%. The case of ulcer vs cancer classification resulted in a lower accuracy of 77.1%, possibly due to smaller difference in appearance than those cases involving normal. CONCLUSIONS: The overall level of performance of the proposed method was very promising to encourage applications in clinical environments. Automatic classification using deep learning technique as proposed can be used to complement manual inspection efforts for practitioners to minimize dangers of missed out positives resulting from repetitive sequence of endoscopic frames and weakening attentions.

[Peptic ulcer disease and stress]. / A fekélybetegség és a stressz.

The discovery that Helicobacter pylori infection is the major cause of peptic ulcer disease revolutionised our views on the etiology and treatment of the disease. This discovery has tempted many experts to conclude that psychological factors and, specifically, stress are unimportant. However, Helicobacter pylori infection alone does not explain fully the incidence and prevalence of peptic ulcer disease. It has been demonstrated that stress can cause peptic ulcer disease even in the absence of Helicobacter pylori infection, supporting a multicausal model of peptic ulcer etiology. Psychological stress among other risk factors can function as a cofactor with Helicobacter pylori infection.

Etiopathogenetic principles and peptic ulcer disease classification.

Ulceration corresponds to tissue loss, breaching the muscularis mucosae. When ulcers develop in the acid-peptic environment of the gastroduodenum, they are traditionally called peptic ulcer (PUD). Ulcers never develop spontaneously in a healthy gastroduodenal mucosa. Ulceration is the ultimate consequence of a disequilibrium between aggressive injurious factors and defensive mucosa-protective factors. The dominant aggressors are strong acid and high proteolytic (pepsin) activity in gastric secretions. The dominant defensors are the phospholipid surfactant layer, covering the mucus bicarbonate gel, the mucus bicarbonate layer covering the epithelium, the tight junctional structures between the epithelial cells, restricting proton permeability, and the epithelial trefoil peptides, contributing to healing after injury. Initially, acid-peptic aggression was considered the overwhelming cause of PUD, supported by the pioneering work of Schwartz, launching the dictum 'no acid, no ulcer'. This led to the universal therapy directed against intragastric acidity, also interfering with peptic activity when the pH was >4. The therapeutic sequence went from large doses of antacids to H(2)-receptor antagonists and finally to proton pump inhibitors (PPIs). The longer the intragastric pH was >3, the quicker ulcer healing was seen. Unfortunately, ulcers often recurred after stopping therapy, demanding maintenance therapy to keep the ulcers healed and to prevent the need for surgery (vagotomy, partial gastric resection). Later on, the emphasis gradually shifted to weakening/failing of the defensive factors, raising the vulnerability of the gastroduodenal mucosa to luminal secretions. Leading injurious mechanisms jeopardizing the mucosal integrity are numerous: infections, especially Helicobacter pylori, drug-induced injury, particularly acetylsalicylic acid (ASA) and non-steroidal anti-inflammatory drugs (NSAIDs), physicochemical and caustic injury, vascular disorders, interfering with perfusion, etc. Currently the leading cause of PUD is H. pylori infection. Standard triple eradication therapy is losing interest in favor of quadruple therapy (PPI, bismuth, tetracycline, metronidazole). H. pylori-induced PPI is rapidly disappearing in the Western world, in contrast to drug-induced ulcer disease and what is called idiopathic PUD. Partial prophylaxis of ASA/NSAID-induced ulceration is possible with PPI maintenance therapy, but novel ways to strengthen the mucosal defense are urgently awaited.

Short- and long-term results of transcatheter embolization for massive arterial hemorrhage from gastroduodenal ulcers not controlled by endoscopic hemostasis.

BACKGROUND AND AIM: Severe bleeding from gastrointestinal ulcers is a life-threatening event that is difficult to manage when endoscopic treatment fails. Transcatheter embolization has been suggested as an alternative treatment in this situation. The present study reports on the efficacy and long-term outcomes of transcatheter embolization after failed endoscopic treatments were assessed in high operative- risk patients. METHODS: A retrospective review of 60 consecutive emergency embolization procedures in hemodynamically unstable patients (41 men, 19 women; mean [+/-SD] age 69.4+/-15 years) was conducted. Patients were referred for selective angiography between 1999 and 2008 after failed endoscopic treatment of massive bleeding from gastrointestinal ulcers. Mean follow-up was 22 months. RESULTS: Embolization was feasible and successful in 57 patients. Sandwich coiling of the gastroduodenal artery was used in 34 patients, and superselective occlusion of the terminal feeding artery (with glue, coils or gelatin particles) was used in 23 patients. Early rebleeding occurred in 16 patients and was managed with endoscopy (n=8), reembolization (n=3) or surgery (n=5). No major embolization-related complications occurred. Sixteen patients died within 30 days after embolization (including three who died from rebleeding) and 11 died thereafter. No late bleeding recurrences were reported. CONCLUSIONS: Selective angiographic embolization is safe and effective for controlling life-threatening bleeding from gastroduodenal ulcers. The procedure usually obviates the need for emergency surgery in these high-risk patients. Survival depends chiefly on underlying conditions.

Formulation, Optimization, in vitro and in-vivo evaluation of levofloxacin hemihydrate Floating Tablets

Abstract The objective of the present investigation was to design, optimize and characterize the gastro retentive floating levofloxacin tablets and perform in-vivo evaluation using radiographic imaging. The floating tablets were prepared by using polymers i.e hydroxy propyl methyl cellulose (HPMC-K4M) and carbopol-940 individually and in combination by nonaquous granulation method. All the Formulations were evaluated for swelling index (S.I), floating behavior and in-vitro drug release kinetics. The compatibility study of levofloxacin with other polymers was investigated by FTIR, DSC, TGA and XRD. Results from FTIR and DSC revealed no chemical interaction amongst the formulation components. The optimized formulation (F11) showed floating lag time (FLT), total floating time (TFT) swelling index (S.I) of 60 sec, >16h and approximately 75 %, respectively. Moreover, F11 showed zero order levofloxacin release in simulated gastric fluid over the period of 6 h. X-ray studies showed that total buoyancy time was able to delay the gastric emptying of levofloxacin floating tablets in rabbits for more than 4 hours. In conclusion the optimized formulation (F11) can be used for the sustained delivery of levofloxacin for the treatment of peptic ulcer.

[On correlations between the clinical course of peptic ulcer and the patients, compliance and quality of life].

The article is dedicated to changes in life quality (LQ) and mental status, and to forming of various types of response to the disease in patients with benign and unfavorable course of peptic ulcer (PU). The subjects were 140 patients in acute phase of PU, in 66 of whom the course of the disease was benign, and in 74--unfavorable. In both groups there were patients with low and satisfactory LQ. High level of nervousness was typical of patients with low LQ and benign course of the disease. Patients with satisfactory LQ and unfavorable course of the disease were characterized by minor personality changes, but developed an agnostic response to the disease much more frequently. Reduction of compliance was found mainly in patients with low LQ and an agnostic response; this should be taken into consideration while correcting the therapy.

Management of bleeding peptic ulcer: current status of intravenous proton pump inhibitors.

Over 80% of ulcer bleeding stops spontaneously, but associated mortality and morbidity remain high. Occurrence of re-bleeding increases mortality 10-fold. Endoscopic findings in those that have bled predict the risk of recurrent bleeding. Patients whose ulcers show a 'flat dot' or clean base (Forrest Class 3) rarely rebleed or need hospitalization. However, actively bleeding ulcers or those with evidence of recent hemorrhage (Forrest Classes 1 and 2) are likely to re-bleed and may need intensive care. Meta-analyses indicate that endoscopic hemostasis has reduced re-bleeding and surgical intervention by over 60% and mortality by 45%. Beyond these reductions, data indicate that (unlike H2-receptor antagonists use, largely devoid of benefits in this area) continuous intravenous infusions of high doses of proton pump inhibitors reduce rebleeding, re-endoscopy, blood transfusion and surgical intervention, but have little effect (beyond endoscopic therapy) on associated mortality, much of it due to conditions other than rebleeding.

Site of peptic ulcer. comparison of hydrochloric acid output, pepsinogen and gastrin serum levels.

BACKGROUND/AIMS: Basal (BAO) and maximum (PAO) hydrochloric acid output after Histalog stimulation, basal pepsinogen (SPL-B), at 60 (SPL-60) and at 90 minutes (SPL-90), and basal gastrin (BG) levels were measured and compared in different gastric (GU) and duodenal (DU) ulcer sites. MATERIAL AND METHODS: Fifty nine patients with peptic ulcer were grouped according to Johnson's classification for gastric ulcers: type I (15), type II (16) type III (12) GU and (16) DU. Fifteen normal subjects were studied as controls. RESULTS: The BAO was greater in the DU than in the control or GU groups. No significant difference was noted in the production of hydrochloric acid after stimulation with Histalog. The SPL-B, at 60 and at 90 minutes was higher in type II GU than in the DU group and controls. The SPL-60 was higher in type II GU patients than in type III GU. Basal gastrin was higher in group DU and types II and III GU compared to the type I GU patients and controls. CONCLUSION: The topographic criteria for differentiating peptic ulcers has low discrimination capacity based on comparison of mean values of HCl acid production, pepsinogen and gastrin serum levels both basal and after stimulation with Histalog due to heterogeneity of these variables in group studies. In these studies, peptic ulcers from different sites should not be grouped as distinct entities except for type II gastric ulcers.

Gastric cyto-secretory correlations in peptic ulcer.

BACKGROUND/AIMS: Maximal acid output, parietal cell mass, serum pepsinogen A (PGA) and total peptic activity (TPA) in gastric juice were studied and compared in duodenal ulcer and in different gastric ulcer sites. METHODOLOGY: 152 peptic ulcer patients were studied. 64 cases of gastric ulcer (GU) were subdivided according to Johnsons's classification and compared with 88 duodenal ulcer (DU) patients diagnosed for the first time. 40 normal subjects were studied as controls. RESULTS: Duodenal ulcer is characterized by normo-hyperparietalism, normo-hyperchloridria and an increase in peptic activity. In cases of GU, such correlation is not only conditioned by the topographic seat of the ulcer, but by the histological condition of the gastric mucosa too. Body GU is characterized by hypoparietalism, hypochloridria, hyper-PGA and hyper-TPA. Pre-pyloric GU is characterized by normo-hyperparietalism, normo-hyperchloridria, hyper-PGA and hyper-TPA. In GU the cyto-secretory behavior is characterized by the histology of the body mucosa with prevalence of preatrophic-atrophic gastritis in case of body GU and prevalence of superficial gastritis in case of GU type II and III. CONCLUSIONS: The results confirm the anatomic-functional analogy between DU and type II and III GU. If considered from the functional point of view, these conditions differ considerably from those that are characteristic of type I GU (as they closely follow the chronic gastritis pattern).

[Cushing's ulcer. Various physiopathological studies and considerations]. / L'ulcera di Cushing Alcune osservazioni e considerazioni fisiopatologiche

28 cases of acute Cushing's ulcer observed out of 2670 necropsies at the S. Spirito Hospital of Rome are reported. Special attention is paid to the most recent physiopathological aspects of the condition.

[Therapeutic relevance of the classification of gastritis]. / Therapeutische Relevanz der Gastritis-Klassifikation.

The histological gastritis classification according to the Sydney system is a standardisation of different classification systems used so far. It is based on the principles of etiology, topography and course of the disease, as well as morphological changes of the gastric mucosa in gastritis. Histological examination represents the practical gold standard of Helicobacter pylori detection and can be used on routinely formaldehyde-fixed biopsy material and hematoxylin-eosin stained slides. Histology is therefore also of practical diagnostic value for the control of the eradication therapy of Helicobacter pylori. The sensitivity of histological Helicobacter detection is on average higher than that of other methods including microbiology. Nevertheless, additional bacterial cultures are useful in cases of therapy resistant Helicobacter pylori infections. For routine diagnosis histological Helicobacter pylori detection is often combined with the so-called rapid urease test, which can be used in the endoscopy outpatient department, frequently offering an interim diagnosis of Helicobacter pylori infection. Furthermore, histological examination enables not only exact gastritis classification with Helicobacter pylori detection, but also the diagnosis of precancerous lesions and gastric carcinomas, as well as primary gastric lymphomas. Biopsy material for histological examination can be taken during the routinely necessary gastroscopic examination of patients with gastric symptoms without much additional burden.

[Evaluation of long-term results in 254 patients surgically treated by gastric resection for peptic ulcer]. / Valutazione dei risultati a distanza in 254 pazienti operati di resezione gastrica per ulcera peptica.

The authors checked 254 patients, operated on with gastric resection according to Billroth II for peptic ulcer, at an interval ranging from 7 up to 20 years from the operation, and, on the basis of the results recorded, these patients were subdivided according to Visick's grading (15,7% in grade 1; 30,6% in grade 2; 26,4% in grade 3 S; 4,7% in grade 3 I; 14,6% in grade 4). On the whole, the remote results were considered satisfactory in 205 patients (80,7%) and unsatisfactory in 49 (19,3%): the latter represent those included in Visick's grades 3 I and 4, wherein the most serious post-resection syndromes had been ascertained; their incidence appeared absolutely higher (91,8%) in the patients subjected to gastric resection for non-complicate peptic ulcer. The Authors therefore conclude the failures of such type of operation are to be ascribed, in high percentage, to uncorrect indications thereto.

[Blood and urine prostaglandin E2 and prostaglandin F2 alpha in patients with chronic gastritis and peptic ulcer].

The blood and urine prostaglandin E2 (PGE2), Prostaglandin F2 alpha (PGF2 alpha) in 106 cases of chronic gastritis and peptic ulcer were investigated by RIA. Meanwhile, the relationship among PGE2, PGF2 alpha and the Syndromes of TCM were approached. The result showed: In comparing with the normal control, the blood and urine PGE2 of 106 cases were obviously higher (P < 0.01), but PGF2 alpha was not (P > 0.05). The urine PGE2 and PGF2 alpha of moderate gastritis were markedly higher than those of mild gastritis (P < 0.05), but there were no significant difference between blood PGE2, PGF2 alpha of moderate gastritis and those of mild gastritis (P > 0.05). The blood PGE2, PGE2/PGF2 alpha ratio of Dampness-Heat in Spleen-Stomach Syndrome and the blood PGE2/PGF2 alpha ratio of incoordination between Liver and Stomach Syndrome were higher than those of Spleen Stomach Deficiency Syndrome in all the cases (P < 0.05). Compared with the normal control, both the decreased amplitude of blood PGE2/urine PGE2 and increased amplitude of blood PGF2 alpha/urine PGF2 alpha ratio showed as following: Spleen-Stomach Deficiency Syndrome > incoordination between Liver and stomach Syndrome > Dampness-Heat in Spleen-Stomach Syndrome. This study suggested: (1) There was a close relation between PGE2 and chronic gastritis and peptic ulcer; (2) There was no correlation between blood PGE2, PGF2 alpha and urine PGE2, PGF2 alpha; (3) PG was possibly a useful objective parameter to the Syndrome Differentiation in TCM.

[Peptic ulcer in the elderly--general considerations, definition and classification].

People over age 60 constitute a substantial and growing proportion of the population in Japan. So it is important that caregivers need to keep in mind the approach to acid related disease in the elderly. We review the physiologic changes of the gastrointestinal tract and related organs with age, epidemiology, unique clinical features, and special considerations in management of digestive disorders in the elderly. Identifying physiologic changes that can be attributed to aging has proven to be quite difficult. Because normal changes of aging have produced conflicting results of changes due to diseases often found in elderly. In the elderly, the manifestation of peptic ulcer is more likely to be silent. The risk of complications, particularly when consuming NSAIDs, is increased in the elderly. Elderly individuals also poorly tolerate ulcer complications, in large part secondary to a higher risk of associated illness.

[Erosive gastroduodenal ulcers]. / Ob érozivno-iazvennykh porazheniiakh zheludka i dvenadtsatiperstnoí kishki.

The authors discuss classification and treatment of erosive lesions of the stomach and duodenum. For after-care and prevention of recurrences it is important to make regular endoscopic check-ups, staged differentiated treatment of patients with recurrent erosive gastroduodenal lesions.

[Classification of peptic ulcer]. / O klassifikatsii iazvennoi bolezni.

To make more simple and useful classification of peptic ulcer the following criteria of the above classification are proposed: nosological entity (ulcer), form (primary, recurrent), site (subcardial, mediogastric, antral, pyloric. duodenal bulbar and duodenal retrobulbar), activity (A-III--a severe exacerbation, A-II--a moderate exacerbation, A-I--clinical remission, A-0--complete remission), clinical course (severe, moderate, light, latent), complications (perforation, hemorrhage, penetration, malignant transformation, pyloric stenosis compensated, pyloric stenosis decompensated).

[Chronic gastroduodenal erosions: clinical and pathogenetic characteristics, classification, differential treatment]. / Khronicheskie gastroduodenal'nye érozii: kliniko-patogeneticheskaia kharakteristika, klassifikatsiia, differentsirovannoe lechenie.

The etiology of chronic (complete) gastroduodenal erosions is not quite clear and is apparently multifactorial. A total of 100 patients with chronic gastroduodenal erosions were examined and divided into 5 representative groups, 20 pts each. For objective evaluation of the significance of various (probable) etiological and pathogenetic factors in development of chronic erosions, monotherapies with drugs with known pharmacodynamic effects were used (ranitidine, double and triple antihelicobacter protocols, dalargin, trental). Triple antihelicobacter protocol and trental proved to be the most effective in the treatment of chronic erosions (90 and 85%, respectively). Dalargin and double antihelicobacter protocol caused disappearance of complete erosions in 50% patients and ranitidine in only 15%. Analysis of the frequency of erosion epithelialization and Helicobacter pylori eradication under the effect of antibiotics showed a clear-cut positive correlation (r = 0.9) between these parameters, which indicated an important role of Helicobacter pylori in the pathogenesis of chronic gastroduodenal erosions. Moderate antihelicobacter activity of dalargin was revealed for the first time. The results of this study indicate that the majority of complete erosions are an individual nosological entity. Microcirculatory disorders in the antral part of the gastric mucosa and its contamination with Helicobacter pylori play an important role in the disease development; decreased cytoprotection and local immunity in the antral part of the stomach are also. The acidopeptic factor cannot cause complete erosions but is one of conditions for their formation.

[The classification of peptic ulcer in the interests of military medical expertise]. / K voprosu o klassifikatsii iazvennoi bolezni v interesakh voenno-vrachebnoi ékspertizy.

The author studies new approaches to the ulcer disease classification conducted in the interests of medical rating board. It is proposed to distinguish acute and chronic forms of this illness in order to diagnose the real prior disease and to determine the category of patients in which ulcer disease was not diagnosed at the preceding stages of medical care. It is reasonable to classify four course phases of illness instead of two, and determine the gravity of ulcer disease which will predetermine medical and rating tactics concerning these patients. In the author's opinion it is necessary to include clinical variants of this disease into classification chart to improve the early diagnosis of recidivations.

[Classification and formulation of clinical diagnosis in peptic ulcer]. / Klassifikatsiia i formulirovanie klinicheskogo diagnoza iazvennoi bolezni.

An ulcer disease classification, basing on the disease clinical course criteria and the local changes characteristics, is proposed. An adequate estimation of the patient state and optimal tactics of treatment are guaranteed by the diagnosis formulation using the classification.