Gastrointestinal Diagnosis Research Articles

Semi-Supervised Segmentation Framework for Gastrointestinal Lesion Diagnosis in Endoscopic Images.

Accurate gastrointestinal (GI) lesion segmentation is crucial in diagnosing digestive tract diseases. An automatic lesion segmentation in endoscopic images is vital to relieving physicians' burden and improving the survival rate of patients. However, pixel-wise annotations are highly intensive, especially in clinical settings, while numerous unlabeled image datasets could be available, although the significant results of deep learning approaches in several tasks heavily depend on large labeled datasets. Limited labeled data also hinder trained models' generalizability under fully supervised learning for computer-aided diagnosis (CAD) systems. This work proposes a generative adversarial learning-based semi-supervised segmentation framework for GI lesion diagnosis in endoscopic images to tackle the challenge of limited annotations. The proposed approach leverages limited annotated and large unlabeled datasets in the training networks. We extensively tested the proposed method on 4880 endoscopic images. Compared with current related works, the proposed method validates better results (Dice similarity coefficient = 89.42 ± 3.92, Intersection over union = 80.04 ± 5.75, Precision = 91.72 ± 4.05, Recall = 90.11 ± 5.64, and Hausdorff distance = 23.28 ± 14.36) on the challenging multi-sited datasets, confirming the effectiveness of the proposed framework. We explore a semi-supervised lesion segmentation method to employ the full use of multiple unlabeled endoscopic images to improve lesion segmentation accuracy. Experimental results confirmed the potential of our method and outperformed promising results compared with the current related works. The proposed CAD system can minimize diagnostic errors.

Association Between Primary Decision-Maker and Care Intensity Among Patients With Advanced Cancer in Mainland China.

Purpose: In China, decisions regarding the treatment of seriously ill patients are usually made by family caregivers. This study aimed to explore the association between the primary decision-makers and the intensity of care given to patients with advanced cancer in China. Methods: We conducted a survey of family members and other caregivers representing 828 cancer patients who died between July 2013 and July 2016. The survey asked: "After the physician conveyed that the disease is incurable, what treatment did the patient and caregiver prefer?" and "Who was the primary decision-maker?" We compared the treatment intensity with locus of decision-making using multivariable logistic regression, adjusting for socio-demographic and clinical covariates informed. Results: Of the 792 patients in our sample, the majority were male (67·2%), 60 years or older (64·0%), married (82·2%), lived with family (98·2%), had medical insurance (94·8%), earned below-average income (53·5%), lived rurally (61·5%), had a gastrointestinal cancer diagnosis (50·8%), experienced moderate or severe pain (86·3%), never received palliative care (80·4%) and had caregivers as primary decision-makers (70·6%). We found that patients were more likely to receive intensive disease-modifying treatments when the primary decision-maker were their children (adjusted odds ratio [AOR] = 1·86, 95% CI:1·26-2·74), spouse (AOR = 2·04, 95% CI:1·26-3·30), or other caregivers (AOR = 3·46, 95% CI:1·24-9·69). Conclusions: When patients with advanced cancer in China did not make their own medical decisions, they were more likely to receive intensive disease-modifying treatments at the end-of-life. Actions should be taken to better understand and ensure that caregivers' decisions reflect the values and presence of patients.

Endoscopic Image Analysis for Gastrointestinal Tract Disease Diagnosis Using Nature Inspired Algorithm With Deep Learning Approach

Endoscopic Image Analysis for Gastrointestinal Tract Disease Diagnosis Using Nature Inspired Algorithm With Deep Learning Approach

Helicobacter pylori status and associated upper gastrointestinal endoscopic diagnosis in a tertiary hospital: A retrospective study

Helicobacter pylori status and associated upper gastrointestinal endoscopic diagnosis in a tertiary hospital: A retrospective study

A Follow-up Report on the Diagnosis of Gastrointestinal Cancer during the COVID-19 Pandemic in Akita Prefecture, Japan in 2021.

We recently reported the decrease in the number of gastrointestinal (GI) cancer diagnoses in 2020 due to disturbance of the healthcare system by the coronavirus disease 2019 (COVID-19) pandemic, using a hospital-based cancer registration system in Akita prefecture, Japan. In this study, we extended the research by showing the latest data (2021) on the number of cancers and examinations. Information on the occurrence and stage of esophageal, gastric, and colorectal cancers was collected from the same database. The number of GI examinations (cancer screening procedures and endoscopic examinations) was also investigated. Following the immediate decrease in the numbers of both GI examinations and GI cancer diagnoses in 2020, a rebound increase in the numbers of GI cancer diagnoses-especially colorectal cancers-was observed in 2021, resulting from an increased number of GI examinations i.e., the total number of colorectal cancers in 2021 increased by 9.0% and 6.8% in comparison to 2020 and pre-pandemic era, respectively. However, the rebound increase in 2021 was largely due to an increase in early-stage cancers, and there was no apparent trend toward the increased predominance of more advanced cancers. It therefore seems that we managed to escape from the worst-case scenario of disturbance of the healthcare system due to pandemic (i.e., an increase in the number of more advanced cancers due to delayed diagnoses). We need to continue to watch the trends in Akita prefecture, which has the highest rate of mortality from the 3 major GI cancers in Japan.

A hybrid approach of missing data imputation for upper gastrointestinal diagnosis

Gastrointestinal and liver diseases (GILDs) are the major causes of death and disability in Middle East. The investigation of upper gastrointestinal (GI) symptoms of a medically limited area resource is a challenge. The analysis of real-world clinical data of upper gastrointestinal (GI) using data mining techniques often is facing observations that the data contains missing values. In this paper, the proposed approach to missing data imputation is accomplished for categorical data onto upper GI diseases to apply the feature selection and classification algorithms with accurate and effective results for diagnosing upper GI diseases. This approach is evaluated by implementing experimental framework to apply five phases. These phases aim at partitioning the dataset to eight different datasets; with various ratio of missing data, performing the feature selection, imputing the missing data, classifying the imputed data, and finally, evaluating the outcome using k-fold cross validation for nine evaluation measures.

A hybrid approach of missing data imputation for upper gastrointestinal diagnosis

A hybrid approach of missing data imputation for upper gastrointestinal diagnosis

Extending TF-Test to the diagnosis of gastrintestinal parasites of humans and animals.

Parasitic species inhabit the intestinal tract of humans and animals, and they can cause damages whose severity depends on several factors related to the parasite and the host. Immunocompromised individuals are more likely to develop severe forms of parasitic infestation. The gastrointestinal parasites diagnosis is performed mainly by the parasitological examination of feces, which consists of the visualization and identification of the parasitic structures eliminated along with the feces. These exams have low sensitivity and a large number of impurities, which can impair the result of the diagnosis. In order to improve the diagnosis quality, a new parasitological technique called Three Fecal Test (TF‑Test) was developed. Thus, we have approached in this study new techniques to diagnose gastrointestinal parasites from human and animal fecal samples. These techniques originate from adaptations and improvements of the TF‑Test technique protocol. From these operational changes it has been possible to carry out the diagnosis of several human and animal parasites, with satisfactory results.

Deep learning based radiomics for gastrointestinal cancer diagnosis and treatment: A minireview.

Gastrointestinal (GI) cancers are the major cause of cancer-related mortality globally. Medical imaging is an important auxiliary means for the diagnosis, assessment and prognostic prediction of GI cancers. Radiomics is an emerging and effective technology to decipher the encoded information within medical images, and traditional machine learning is the most commonly used tool. Recent advances in deep learning technology have further promoted the development of radiomics. In the field of GI cancer, although there are several surveys on radiomics, there is no specific review on the application of deep-learning-based radiomics (DLR). In this review, a search was conducted on Web of Science, PubMed, and Google Scholar with an emphasis on the application of DLR for GI cancers, including esophageal, gastric, liver, pancreatic, and colorectal cancers. Besides, the challenges and recommendations based on the findings of the review are comprehensively analyzed to advance DLR.

The association between upper gastrointestinal endoscopic findings and internal radiation exposure in residents living in areas affected by the Chernobyl nuclear accident.

Many people living around the Chernobyl Nuclear Power Plant (CNPP) have been exposed to 137Cs for several decades after the CNPP accident. Although half-life of 137Cs is about 30 years, some wild forest foodstuffs are contaminated by 137Cs even now. We pointed out in a previous report that low-dose internal radiation has been occasionally detected in people's body. Moreover, some doctors in local hospitals have claimed that internal exposure from contaminated foodstuffs may affect the digestive organs and possibly cause gastrointestinal (GI) diseases. Thus, we attempt to assess whether internal radiation exposure affects digestive organs or not, and the possible factors that influence digestive organs. Overall, 1,612 residents were assessed for internal 137Cs concentration using Whole-Body Counter and their digestive organs were screened with upper GI endoscopy from 2016-2018 in the Zhytomyr region, Ukraine. All participants answered to the questionnaire including their background, intake of wild forest foodstuff, intake frequency, smoking habits, and alcohol consumption. We checked the number of upper GI endoscopic diagnosis per person to assess the extent of damage to the upper digestive organs. Next, we statistically analyzed associations between this number and age, sex, level of internal exposure dose, alcohol consumption, wild forest foodstuff intake, and smoking. Consequently, we revealed that the number of GI diagnosis is significantly increased by factors such as sex, intake of wild forest foodstuff, and alcohol consumption. However, the average level of internal exposure of 137Cs and smoking did not relate to the number of GI diagnosis. Thus, the results of multiple regression revealed that alcohol consumption is independently related to the number of GI diagnosis that is most likely accompanied by the intake of wild forest foodstuff. In conclusion, the low-dose internal exposure may not affect the digestive organs of residents living around CNPP.

Psychiatric comorbidities among adult patients with disorders of gut-brain interaction: Prevalence and relationships to treatment outcomes.

Little is known about the impact of psychiatric comorbidity on pharmacologic treatment outcomes, including neuromodulators (medications targeting the gut-brain axis), among adult patients with disorders of gut-brain interaction (DGBI). Accordingly, we aimed to examine associations between psychiatric comorbidity and DGBI pharmacologic treatment outcomes. In a retrospective study of consecutively referred new patients (N=410; ages 18-90; 73% female) to a tertiary neurogastroenterology clinic in 2016 with follow-up through 2018, relationships between psychiatric illness (any psychiatric illness, anxiety disorders, depressive disorders) and pharmacologic treatment selection (any medication, neuromodulating medication) and treatment outcomes, respectively, were examined using multivariable logistic regression, adjusting for demographics, gastrointestinal (GI) diagnoses, and pre-existing neuromodulator use. Anxiety disorders (35%) were the most common psychiatric comorbidity, followed by depressive disorders (29%). Patients with anxiety disorders were more likely to be prescribed a neuromodulator by their gastroenterologist (OR=1.72 [95% CI 1.10-2.75]) yet less likely to respond to neuromodulators (OR=0.43 [0.21-0.90]) or any GI medication (OR=0.24 [0.12-0.50]) in fully adjusted analyses. In contrast, depressive disorders were not associated with neuromodulator prescription or response. Anxiety disorders are common among patients with DGBI and significantly reduce the likelihood of GI pharmacologic treatment response to any medication prescribed, including neuromodulators.

Abstract 10222: Impact of Bleeding Event for New Cancer Diagnosis in Patients With Antiplatelet Therapy After Percutaneous Coronary Intervention

Background: It has been reported that the bleeding event represents unmasking of an occult cancer in anticoagulated patients with atrial fibrillation. However, the impact of bleeding event in patients with antiplatelet therapy after percutaneous coronary intervention (PCI) for new cancer diagnosis remains unclear. Methods and Results: Consecutive 1796 patients with coronary artery disease who underwent PCI and taking antiplatelet therapy were enrolled. Patients taking oral anticoagulants were excluded, and relationships between bleeding event and new cancer diagnosis were examined. Among 1565 patients, 189 (12.0%) experienced any bleeding events defined as Bleeding Academic Research Consortium (BARC) type-2/3/5 bleeding, 78 (4.9%) experienced minor bleeding events defined as BARC type-2 bleeding, and 116 (7.4%) were diagnosed with new cancer during a mean follow-up period of 1528 days. Patients with any bleeding were older (71.3 vs. 68.8 years, P = 0.005) and had higher prevalence of being at Academic Research Consortium criteria for high bleeding risk (62.4% vs. 43.8%, P < 0.001) than those without bleeding event. Of 189 patients with any bleeding and 78 patients with minor bleeding events, 31 (16.4%) and 16 (20.5%) were diagnosed with new cancer, respectively. In the univariate Cox proportional hazard analysis, the incidence of new cancer diagnosis was higher in patients with any bleeding and minor bleeding events than those without bleeding event (hazard ratio [HR] 3.19, P < 0.001 and HR 5.70, P < 0.001, respectively). In the multivariate analysis after adjusting for confounding factors, any bleeding and minor bleeding events were the independent predictors of new cancer diagnosis (HR 2.27, P = 0.003 and HR 3.93, P < 0.001, respectively). In addition, any gastrointestinal bleeding and minor bleeding events were associated with higher rates of new gastrointestinal cancer diagnosis (HR 4.46, P = 0.041 and HR 6.73, P = 0.009, respectively). Conclusion: In patients with antiplatelet therapy after PCI, any bleeding and minor bleeding events were associated with subsequent new cancer diagnosis. Even minor bleeding should be taken care as the first manifestation of underlying cancer during antiplatelet therapy.

Clinical and Demographic Characteristics of Young Patients Requiring Surgery for Diverticulitis

Introduction: Diverticulitis is a common gastrointestinal diagnosis. Diverticulitis is as increasing for all demographics, and the rise has been greatest for the younger subgroup. This study aimed to characterize the clinical and demographic features associated with patients ≤40 years needing surgery after their index episode of diverticulitis. Methods: Retrospective review of patients undergoing surgical intervention for index episode of diverticulitis from 2015 to 2021 at a single tertiary care center. Patients ≤40 years were compared with a control group consisting of patients >40 years old. Results: The analysis included 62 patients; 21 patients ≤40 years old (mean age 32.7 years) and 41 controls >40 years old (mean age 57.8 years); p = 0.001. Patients ≤ 40 years were more likely to be male (71.4% vs 34%, p = 0.005), Hispanic (66.7% vs 34.1%, p = 0.019), and obese (71.4% vs 29.3%, p = 0.017) with a greater mean BMI (32.9 kg/m2 vs 28.2 kg/m2, p = 0.002). The most likely indication for surgery in the ≤40 group was recurrent disease (38.1%), while the most common indication in the >40 group was development of a fistula (31.7%); p = 0.734. There was no difference in overall postoperative complication or readmission. Conclusion: We see an increasing proportion of advanced diverticulitis in the young population and our study demonstrates that the characteristics of this group differs from older patients. Further work to elucidate potential social, environmental, and biological etiologies of these findings is warranted.

Progress with Metabolomic Blood Tests for Gastrointestinal Cancer Diagnosis-An Assessment of Biomarker Translation.

There is an urgent need for cost-effective, non-invasive tools to detect early stages of gastrointestinal cancer (colorectal, gastric, and esophageal cancers). Despite many publications suggesting circulating metabolites acting as accurate cancer biomarkers, few have reached the clinic. In upper gastrointestinal cancer this is critically important, as there is no test to complement gold-standard endoscopic evaluation in patients with mild symptoms that do not meet referral criteria. Therefore, this study aimed to describe and solve this translational gap. Studies reporting diagnostic accuracy of metabolomic blood-based gastrointestinal cancer biomarkers from 2007 to 2020 were systematically reviewed and progress of each biomarker along the discovery-validation-adoption pathway was mapped. Successful biomarker translation was defined as a composite endpoint, including patent protection/FDA approval/recommendation in national guidelines. The review found 77 biomarker panels of gastrointestinal cancer, including 25 with an AUROC >0.9. All but one was stalled at the discovery phase, 9.09% were patented and none were clinically approved, confirming the extent of biomarker translational gap. In addition, there were numerous "re-discoveries," including histidine, discovered in 7 colorectal studies. Finally, this study quantitatively supports the presence of a translational gap between discovery and clinical adoption, despite clear evidence of highly performing biomarkers with significant potential clinical value.

S519 Impact of COVID-19 on Motility Practices: New Onset Gastrointestinal and Autonomic Dysfunction After COVID-19 and the Effects of Vaccination in a Nationwide Cohort Study

S519 Impact of COVID-19 on Motility Practices: New Onset Gastrointestinal and Autonomic Dysfunction After COVID-19 and the Effects of Vaccination in a Nationwide Cohort Study

S1488 Characteristics and Healthcare Patterns of Super-Utilizers Among Patients Hospitalized With a GI Diagnosis

Introduction: In the United States, a small number of patients, labeled as super-utilizers accounts for a disproportionately large fraction of healthcare expenditure. The knowledge about the characteristics of super-utilizers and their impact on healthcare utilization would be helpful to various stakeholders to target interventions aimed at reducing the disproportionate use of the healthcare system. Methods: We extracted data from the Nationwide Readmission Database (NRD) 2019, on patients admitted primarily with a gastrointestinal (GI) pathology. We used Agency for Healthcare Research and Quality’s definition for super-utilizers (2 standard deviations above the average number of hospital stays for patients in each payer category which were ≥4 for Medicaid/Medicare and ≥3 for privately /uninsured in 2019). We collected data on demographics, readmission rates, and hospitalization costs. Results: Of 1,704,913 patients who were discharged alive, 6.3% were categorized as super-utilizers. The highest proportion of super-utilizers was observed among patients with hepatic failure (19.6%), non-infectious hepatitis (13.6%), and inflammatory bowel disease (10.4%) (Figure 1). On multivariate analysis, several characteristics were independently associated with super-utilizers, especially the comorbidity burden and disease severity (Table 1). The mean number of admissions for an average super-utilizer was ∼4 times compared to other patients (mean admissions: 5.5 vs 1.4; P < 0.001). Super-utilizers were mostly admitted for medical diagnosis (83.4% vs 67.3%) as compared to surgical (16.5% vs 31.4%) and for non-elective reasons (92.1% vs 88.9%). The 30-day readmission rate was ∼6 times higher among super-utilizers compared to other patients (36.6% vs 6.4% P < 0.001). Compared to other patients, super-utilizers had longer hospital stays (mean days: 5.8 vs 4.4; P < 0.001) and higher average hospitalization costs (mean US$: 15060 vs 13023; P < 0.001). Super-utilizers although only 6.3%, accounted for 31.5% of hospitalization days, 28.2% of costs, 19.6% of hospital stays, and 27.3% of all-cause 30-day readmissions. These findings were consistent across the different payer types on subgroup analysis. Conclusion: 6.3% of patients admitted with a GI diagnosis were classified as super-utilizers. Comorbidity burden and disease severity were strongly associated with super-utilizers. Super-utilizers accounted for a disproportionate share of hospital days, costs, stays, and 30-day readmissions compared to other patients.Figure 1.: Percentage of super-utilizers among various GI pathologies resulting in hospitalization in 2019. Table 1. - Multivariate analysis showing association between super-utilizer status and various patient- and hospital-level characteristics Characteristics Odds Ratio [95% CI] ‘P’ value Charlson Comorbidity index 1.70 (1.68-1.73) < 0.001 Severity 1.37 (1.35-1.38) < 0.001 Teaching hospital 1.21 (1.15-1.28) < 0.001 Hospital Bed size 1.05 (1.03-1.07) < 0.001 Hospital control 1.04 (1.01-1.07) 0.006 Age 0.98 (0.98-0.98) < 0.001 Income 0.98 (0.97-0.99) < 0.001 Female gender 0.97 (0.95-0.98) < 0.001 Urban location 0.88 (0.83-0.92) < 0.001

Nationwide implementation of a multifaceted, tailored strategy to improve uptake of standardized structured reporting in oncological pathology: An effect and process evaluation.

336 Background: Implementation strategies improve guideline adherence. Effect and process evaluations are conducted to provide insights into success or failure of implementation strategies. We conducted both evaluations to determine nationwide implementation of standardized structured reporting (SSR) in pathology, to facilitate optimal treatment decisions in oncology care. Methods: We conducted an interrupted time series analysis to evaluate the effect of a fully digital nationwide implementation strategy on SSR use in pathology laboratories. We further conducted a process evaluation to evaluate the exposure to the strategy elements, the experiences of the users with the implementation strategy and remaining barriers of SSR implementation. We also tested whether being exposed to a specific element of the strategy resulted in more increase of average SSR use. Results: There was a significant increase in average SSR use directly after the strategy introduction for reporting of gastrointestinal ( p =.018) and urological ( p =.003) oncological diagnoses. A significant increase was present for all oncological resections as a group ( p =.007). The “Feedback button”, an option within the templates used for standardized structured reporting to provide feedback to the provider and one of the elements of the implementation strategy, was most frequently used by the SSR users and effectiveness results showed that it increased average SSR use after strategy introduction. Remaining barriers most related to rigidity, time consumption and content of SSR. Conclusions: Nationwide SSR implementation improved for specific tumor types and retrieval methods. Next step will be to further improve the use of SSR, and simultaneously, to further develop potential benefits of high SSR use, focusing on re-using discrete pathology data. In this way, we can eventually facilitate proper treatment decisions in oncology.

Future of Endoscopy in Inflammatory Bowel Diseases (IBDs)

Gastrointestinal (GI) endoscopy has transformed over the years in scope, safety, accuracy, acceptability, and cost effectiveness of the clinical practice. There has been a reduction in the superiority of the endoscopic devices as innovations have taken place and increased the diagnostic values with certain limitations. There are particular difficulties in striking a balance between the development of new technology and the device's acceptance. The wide use of endoscopy for investigating GI lesions and diagnosis has led to an increase in more advanced methods and their broad application. It can simultaneously diagnose pre-malignant and malignant lesions, and newer interventions have made the biopsy specimen uptake possible. In this review article, we focus on the more recent roles, indications, applications, and usage of the innovative methods of endoscopy.

A comparative cohort study of gastrointestinal oncology patients: Impact of a shift to telehealth on delivery of interprofessional cancer care

Early studies of oncology visits performed via telehealth demonstrate patient and provider satisfaction; however, understanding of the impact of telehealth on clinic workflows is limited. The incorporation of telehealth visits into an interprofessional model of oncology care was evaluated to assess for changes in care delivery and patient engagement. New patients with a gastrointestinal cancer diagnosis who were actively undergoing treatment and followed for at least three months were divided into two cohorts based on telehealth utilization. Individual patient charts were reviewed by touchpoint, consisting of in-person visits, telehealth visits, phone calls, and patient portal messages. A total of 28 patient charts were analyzed, 11 pre-telehealth conventional care patients, and 17 telehealth patients. Telehealth cohort patients demonstrated an increased average number of total touchpoints when compared to the pre-telehealth cohort (p-value = 0.008) and had an increased number of patient portal and phone call touchpoints (p-value = 0.00 and 0.002). Telehealth provided more interactions between patients and providers demonstrating increased connectivity between a patient and their care team throughout their complex cancer journey. Clinic workflows may need to adjust to account for the increased demand of unscheduled patient interactions.

Design and Control of a Magnetically-Actuated Capsule Robot With Biopsy Function.

Wireless capsule endoscopy has been well used for gastrointestinal (GI) tract diagnosis. However, it can only obtain images and cannot take samples of GI tract tissues. In this study, we designed a magnetically-actuated biopsy capsule (MABC) robot for GI tract diagnosis. The proposed robot can achieve locomotion and biopsy functions under the control of external electromagnetic actuation (EMA) system. Two types of active locomotion can be achieved, plane motion refers to the robot rolling on the surface of the GI tract with a rotating uniform magnetic field. 3D motion refers to the robot moving in 3D space under the control of the EMA system. After reaching the target position, the biopsy needle can be sprung out for sampling and then retracted under a gradient magnetic field. A pill-shaped robot prototype ( ϕ15 mm×32 mm) has been fabricated and tested with phantom experiments. The average motion control error is 0.32 mm in vertical direction, 3.3 mm in horizontal direction, and the maximum sampling error is about 5.0 mm. The average volume of the sampled tissue is about 0.35 mm 3. We designed a MABC robot and proposed a control framework which enables planar and 3D spatial locomotion and biopsy sampling. The untethered MABC robot can be remotely controlled to achieve accurate sampling in multiple directions without internal power sources, paving the way towards precision sampling techniques for GI diseases in clinical procedures.