Health Condition 1: K319- Disease of stomach and duodenum, unspecified
Conditions
Interventions
Intervention1: AI-assisted endoscopy: In AI-assisted examinations, lesions will be identified by stable red-box markings generated by AI model
Control Intervention1: Standard high-definition white-lig
Sponsors
Department of Gastroenterology
Eligibility
Inclusion criteria
Inclusion criteria: 1 Age more than or equal to 18 years 2 Individuals undergoing upper gastrointestinal endoscopy who are either symptomatic with upper gastrointestinal complaints and or apparently healthy individuals presenting with high risk alarm features including a Loss of appetite b Unintentional weight loss less than or equal to 5 percent of body weight in the past 3 months c Anemia as per WHO criteria hemoglobin less than or equal to 13 g per dL in men and less than 12 g per dL in women d Family history of UGI malignancy e Diets high in salted smoked and pickled foods f Prior history of Helicobacter pylori infection or related atrophic gastritis and Barretts esophagus 3 Willingness to provide informed consent
Exclusion criteria
Exclusion criteria: 1 Prior diagnosis of UGI malignancy 2 History of previous major surgery altering anatomy 3 Absolute contraindications to upper GI endoscopy 4 Contraindications to biopsy 5 Pregnancy and Lactating mothers 6 Unwilling to provide written informed consent
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| 1 To compare the missing rate of upper gastrointestinal malignancies between AI assisted HD WLE and standard HD WLE 2 To compare the diagnostic yield per biopsy between AI assisted HD WLE and standard HD WLE Timepoint: Baseline | — |
Secondary
| Measure | Time frame |
|---|---|
| 1 To compare the UGI malignancies detection rate between AI assisted HD WLE and standard HD WLE. 2 To compare biopsy rates between AI assisted HD WLE and standard HD WLE. 3 To compare the mean number of neoplastic lesions detected per patient between AI-assisted HD WLE and standard HD WLE. 4 To compare the patient-level missing rate of UGI malignancies between AI-assisted HD WLE and standard HD WLE. 5 To determine and compare the positive predictive value of neoplastic lesion identification between AI assisted HD WLE and standard HD WLE. 6 To compare the total diagnostic observation time between AI assisted HD WLE and standard HD WLE. Timepoint: Baseline | — |
Countries
India
Contacts
Public ContactVineet Ahuja
All India Institute of Medical Sciences, New Delhi
Outcome results
None listed