Artificial Intelligence, Endoscopy, Intestinal Metaplasia of Gastric Mucosa
Conditions
Brief summary
Gastric intestinal metaplasia(GIM) is an important stage in the gastric cancer(GC). With technical advance of image-enhanced endoscopy (IEE), studies have demonstrated IEE has high accuracy for diagnosis of GIM. The endoscopic grading system (EGGIM), a new endoscopic risk scoring system for GC, have been shown to accurately identify a wide range of patients with GIM. However, the high diagnostic accuracy of GIM using IEE and EGGIM assessments performed all require much experience, which limits the application of EGGIM. The investigators aim to design a computer-aided diagnosis program using deep neural network to automatically evaluate the extent of IM and calculate the EGGIM scores.
Detailed description
Globally, gastric cancer is the fifth most prevalent malignancy and the third leading cause of cancer mortality. Gastric intestinal metaplasia (GIM) is an intermediate precancerous gastric lesion in the gastric cancer cascade. Studies have shown that the 5-year cumulative incidence of gastric cancer in IM patients ranges from 5.3% to 9.8% . With technical advance of image-enhanced endoscopy (IEE), studies have demonstrated IEE has high accuracy for diagnosis of GIM. The endoscopic grading system (EGGIM), a new endoscopic risk scoring system for GC, have been shown to accurately identify a wide range of patients with GIM. However, The high diagnostic accuracy of GIM using IEE and EGGIM assessments performed all require much experience, which limits the application of EGGIM. The investigators aim to design a computer-aided diagnosis program using deep neural network to automatically evaluate the extent of IM and calculate the EGGIM scores.
Interventions
None listed
Sponsors
Shandong University
Study design
Observational model
OTHER
Time perspective
RETROSPECTIVE
Eligibility
Sex/Gender
ALL
Age
18 Years to 80 Years
Healthy volunteers
No
Inclusion criteria
* patients aged 18-80 years who undergo the IEE examination
Exclusion criteria
* patients with severe cardiac, cerebral, pulmonary or renal dysfunction or psychiatric disorders who cannot participate in gastroscopy * patients with previous surgical procedures on the stomach * patients who refuse to sign the informed consent form
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| The specificity of AI model to assess the degree of intestinal metaplasia in an endoscopic picture | 2 years | The specificity of AI model to assess the degree of intestinal metaplasia in an endoscopic picture |
| The accuracy of AI model to assess the degree of intestinal metaplasia in an endoscopic picture | 2 years | The accuracy of AI model to assess the degree of intestinal metaplasia in an endoscopic picture |
| The sensitivity of AI model to assess the degree of intestinal metaplasia in an endoscopic picture | 2 years | The sensitivity of AI model to assess the degree of intestinal metaplasia in an |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Accuracy of the experienced endoscopists to assess the degree of intestinal metaplasia | 2 years | Accuracy of the experienced endoscopists to assess the degree of intestinal metaplasia in an endoscopic picture |
| Inter-observer agreement among inexperienced endoscopists in identifying degree of intestinal metaplasia | 2 years | Inter-observer agreement among inexperienced endoscopists in identifying degree of intestinal metaplasia in an endoscopic picture |
| Accuracy of the inexperienced endoscopists to assess the degree of intestinal metaplasia | 2 years | Accuracy of the inexperienced endoscopists to assess the degree of intestinal metaplasia in an endoscopic picture |
| Inter-observer agreement among experienced endoscopists in identifying the degree of intestinal metaplasia | 2 years | Inter-observer agreement among experienced endoscopists in identifying the degree of intestinal metaplasia in an endoscopic picture |
Countries
China
Contacts
Primary Contactyanqing Li, MD, PHD
Outcome results
None listed