The study is designed to collect magnetically-controlled capsule endoscopy (MCE) images, esophagogastroduodenoscopy (EGD) images, CT images, and histopathology results to establish a database for early upper gastrointestinal cancer, precancerous lesions, and precancerous conditions.

Images of MCE for UGIC are scarce, limiting the development of corresponding AI diagnostic model. We aim to create an image database with MCE images, EGD images, CT images, and histopathology results, in patients treated by ESD for early upper gastrointestinal cancer, or biopsy for precancerous lesions, and precancerous conditions, so as to facilitate image-based deep learning studies.

Inclusion Criteria:

1. Male or female participants aged ≥ 18 years
2. Patients with early upper gastrointestinal cancer referred for endoscopic submucosal dissection (ESD), or patients with upper gastrointestinal precancerous lesions or precancerous conditions referred for biopsy by esophagogastroduodenoscopy (EGD)
3. Participants who are willing and able to participate in this study, complete all required examinations and follow-up visits, and who have provided written informed consent

Exclusion Criteria:

1. Patients who are not eligible for surgery or who refuse any abdominal surgery in case of capsule retention and endoscopic retrieval failure
2. Patients with cardiac pacemaker or other implanted electromedical devices which could interfere with magnetic resonance
3. Suspected or known intestinal stenosis or other known risk factors for capsule retention
4. Patients with dysphagia
5. Pregnancy
6. Allergy to polymeric materials, latex, simethicone, or pronase
7. Patients plan to undergo magnetic resonance imaging examination before excretion of the capsule
8. Patients with any condition that precludes compliance with the study

Tertiary Hospital