Artificial Intelligence, Gastric Cancer, Gastric Intestinal Metaplasia, Pathology
Conditions
Keywords
gastric cancer, Gastric Intestinal Metaplasia, Digital Pathology, Whole Slide Image
Brief summary
The OLGIM staging system is highly recommended for a comprehensive assessment of GIM severity to evaluate patients' gastric cancer risk. However, its need to take at least 4 biopsies is not clinically feasible due to a serious shortage of pathologists compared with the large number of gastric cancer screening population. We plan to develop a Digital Pathology artificial intelligence diagnosis system (DPAIDS), to automatically identify tumor areas in whole slide images(WSI) and quickly and accurately quantify the severity of intestinal metaplasia according to the proportion of intestinal metaplasia areas.
Detailed description
Gastric cancer is the fifth most prevalent malignancy and the third most deadly worldwide, and intestinal metaplasia (IM) is a common precancerous state that is closely associated with gastric carcinogenesis .The OLGIM staging system is highly recommended for a comprehensive assessment of GIM severity to evaluate patients' gastric cancer risk. However, its need to take at least four biopsies is not clinically feasible due to a serious shortage of pathologists compared with the large number of gastric cancer screening population. Developing automated screening methods can reduce the heavy diagnostic workload. With advances in digital pathology scanning devices and deep learning technologies, whole-slide images (WSI) have been used to develop automated cancer diagnostic systems. We plan to develop a Digital Pathology artificial intelligence diagnosis system (DPAIDS), to automatically identify tumor areas in whole slide images(WSI) and quickly and accurately quantify the severity of intestinal metaplasia according to the proportion of intestinal metaplasia areas. Then biopsies will be prospectively collected and prepared as WSI for model validation.
Interventions
DIAGNOSTIC_TESTThe diagnosis of Artificial Intelligence and pathologists
Pathologists and AI will assess the severity of intestinal metaplasia and judge the tumor area of whole slide images of gastric biopsy specimens independently. In addition, the pathologists can not see the diagnosis of AI.
Sponsors
Shandong University
Study design
Observational model
OTHER
Time perspective
PROSPECTIVE
Eligibility
Sex/Gender
ALL
Age
40 Years to 75 Years
Healthy volunteers
No
Inclusion criteria
* patients aged 40-75 years who undergo the gastroscopy examination and biopsy
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 with contraindications to biopsy * patients who refuse to sign the informed consent form
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| The diagnostic performance of AI model to assess the severity of intestinal metaplasia | 2 years | The diagnostic performance of AI model to assess the severity of intestinal metaplasia in a single biopsy tissue slide: Accuracy, sensitivity, and specificity |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Accuracy of the digital pathological AI model to identify tumor regions | 2 years | Accuracy of the digital pathological AI model in identifying tumor regions in the whole slide images |
| Accuracy of digital pathological AI models to identify glands, mucosal epithelium, and intestinal metaplasia in non-neoplastic areas | 2 years | Accuracy of digital pathological AI models to identify glands, mucosal epithelium, and intestinal metaplasia in non-neoplastic areas |
Countries
China
Contacts
Primary ContactYanqing Li, MD, PhD
Outcome results
None listed