Gachon University Gil Hospital, 'AI-based gastroscopy Smart Endo' clinical application
Nov 25, 2024
Gachon University Gil Hospital (Hospital Director Kim Woo-kyung) has set a new milestone in the field of gastroscopy with artificial intelligence (AI)-based gastroscopy diagnostic assistance software.
Gachon University Gil Hospital, which introduced the AI endoscope system 'Smart Endo' for the first time among domestic medical institutions, will apply the artificial intelligence-based gastroscopy diagnostic assistance software Smart Endo' produced as part of the 'AI Precision Medical Solution (Dr. Answer 2.0) Development' project organized by the Ministry of Science and ICT and the Information and Communication Industry Promotion Agency on the 25th.
According to the National Cancer Information Center, as of 2021, gastric cancer surpassed lung cancer, liver cancer, and colon cancer as the number one cancer prevalence in Korea (excluding thyroid cancer). About 70% of all gastric cancer patients are early gastric cancer, and this frequency is also increasing every year. However, in the case of advanced gastric cancer, the 5-year survival rate is less than 40%, and the endoscope is used as the most powerful tool to detect and diagnose gastritis and ulcers. In the case of early gastric cancer, the survival rate can be improved by up to 90%.
Smart Endo, developed as part of the Doctor Answer 2.0 project, is characterized by the occurrence of endoscopic time and effectiveness depending on the skill level and condition of the medical staff. Through this, medical staff receive help from AI for the detection, detection, and diagnosis of gastric lesion diseases, and early detection can be performed smoothly.
This Smart Endo is a system that automatically detects, diagnoses, and predicts stomach cancer and gastric lesions in real time and warns users. Here, automatic diagnosis technology for lesions and gastric cancer using deep learning classification architecture is utilized. It also provides probabilities for chronic atrophic gastritis, gastric mucosal dysplasia, and gastric cancer, as well as tight diagnosis of lesions. At this time, the patient's history and clinical information are considered together.
Smart Endo also uses deep learning-based AI technology to help detect gastric lesions. Smart Endo has learned more than 30,000 gastroscopy and genetic big data, which helps medical staff detect gastric lesions. At this time, 'Transfer Leaning' technology that detects the movement of medical staff is applied to improve accuracy.
As a result, Smart Endo has advantages such as ▲ lowering the liver rate of gastroscopy, improving the diagnosis rate, ▲ improving the possibility of early diagnosis of gastric cancer ▲ improving the diagnosis rate of dysplasia micro-gastric cancer ▲ reducing unnecessary biopsy costs.
In fact, the use of Smart Endo in self-developed clinical trials had the effect of improving the lesion detection performance of unskilled clinicians. This is to kill two birds with one stone to prevent and detect gastric cancer early through predicting the risk of gastric cancer and high-performance diagnostic technology.
Professor Kim Kyung-oh of the Department of Gastroenterology, who participated in the Doctor Answer 2.0 task and led the development of the artificial intelligence (AI)-based gastroscopy diagnostic assistance software Smart Endo, said, "Discovering gastric ulcers, gastritis, and stomach cancer requires a lot of experience and concentration"There is a large difference between experienced professionals and beginners in the detection rate (ADR) of lesions that are likely to develop into cancer, and this gap can be very narrowed if Smart Endo is used."
He then added that the reason why AI should be applied to gastric cancer research is that AI can provide experience and concentration.Currently, in the case of gastroscopy, if there is an undiscovered gastrointestinal disease depending on the examiner's skill and condition, it takes up to five years to wait until the next test. However, if Smart Endo is used, it automatically recognizes gastrointestinal diseases in real time during the test and warns the examiner to shorten the detection of gastrointestinal diseases to less than 20 minutes.
Smart Endo consists of two 'Smart Endo Gastro' monitors with video AI SW, two 'Endoscopic Original Monitor' monitors, and a body computer system jointly produced by Infinite Healthcare, a global medical imaging company. Through a total of four monitors, medical staff can make precise diagnosis by comparing the screen to which AI SW is applied and the original screen in real time during the endoscope.
Gachon University Gil Hospital has obtained medical device certification from the Ministry of Food and Drug Safety for AI-Doctor for Gastric Cancer through medical records (EMR) and interviews of patients prior to the clinical launch of Smart Endo, an artificial intelligence (AI)-based gastroscopy diagnostic assistance software developed by Doctor Answer 2.0 task. The software uses the patient's blood test results, health checkups such as Helicobacter tests, electronic medical record data, and survey results on lifestyle to derive the risk rate of gastric cancer. Through this, it is expected that patients' endoscopic compliance can be increased by deriving a quantified risk rate for gastric cancer.
"Gastric cancer is the most common type of cancer in Korea, and early detection and treatment are very important because early detection can increase survival rates," said Kim Woo-kyung, director of Gachon University Gil Hospital. "It is expected to dramatically improve the detection rate of gastric lesions by incorporating artificial intelligence technology, reducing risk factors for gastric cancer treatment and contributing to the improvement of people's health."
Gachon University Gil Hospital, which introduced the AI endoscope system 'Smart Endo' for the first time among domestic medical institutions, will apply the artificial intelligence-based gastroscopy diagnostic assistance software Smart Endo' produced as part of the 'AI Precision Medical Solution (Dr. Answer 2.0) Development' project organized by the Ministry of Science and ICT and the Information and Communication Industry Promotion Agency on the 25th.
According to the National Cancer Information Center, as of 2021, gastric cancer surpassed lung cancer, liver cancer, and colon cancer as the number one cancer prevalence in Korea (excluding thyroid cancer). About 70% of all gastric cancer patients are early gastric cancer, and this frequency is also increasing every year. However, in the case of advanced gastric cancer, the 5-year survival rate is less than 40%, and the endoscope is used as the most powerful tool to detect and diagnose gastritis and ulcers. In the case of early gastric cancer, the survival rate can be improved by up to 90%.
Smart Endo, developed as part of the Doctor Answer 2.0 project, is characterized by the occurrence of endoscopic time and effectiveness depending on the skill level and condition of the medical staff. Through this, medical staff receive help from AI for the detection, detection, and diagnosis of gastric lesion diseases, and early detection can be performed smoothly.
This Smart Endo is a system that automatically detects, diagnoses, and predicts stomach cancer and gastric lesions in real time and warns users. Here, automatic diagnosis technology for lesions and gastric cancer using deep learning classification architecture is utilized. It also provides probabilities for chronic atrophic gastritis, gastric mucosal dysplasia, and gastric cancer, as well as tight diagnosis of lesions. At this time, the patient's history and clinical information are considered together.
Smart Endo also uses deep learning-based AI technology to help detect gastric lesions. Smart Endo has learned more than 30,000 gastroscopy and genetic big data, which helps medical staff detect gastric lesions. At this time, 'Transfer Leaning' technology that detects the movement of medical staff is applied to improve accuracy.
As a result, Smart Endo has advantages such as ▲ lowering the liver rate of gastroscopy, improving the diagnosis rate, ▲ improving the possibility of early diagnosis of gastric cancer ▲ improving the diagnosis rate of dysplasia micro-gastric cancer ▲ reducing unnecessary biopsy costs.
In fact, the use of Smart Endo in self-developed clinical trials had the effect of improving the lesion detection performance of unskilled clinicians. This is to kill two birds with one stone to prevent and detect gastric cancer early through predicting the risk of gastric cancer and high-performance diagnostic technology.
Professor Kim Kyung-oh of the Department of Gastroenterology, who participated in the Doctor Answer 2.0 task and led the development of the artificial intelligence (AI)-based gastroscopy diagnostic assistance software Smart Endo, said, "Discovering gastric ulcers, gastritis, and stomach cancer requires a lot of experience and concentration"There is a large difference between experienced professionals and beginners in the detection rate (ADR) of lesions that are likely to develop into cancer, and this gap can be very narrowed if Smart Endo is used."
He then added that the reason why AI should be applied to gastric cancer research is that AI can provide experience and concentration.Currently, in the case of gastroscopy, if there is an undiscovered gastrointestinal disease depending on the examiner's skill and condition, it takes up to five years to wait until the next test. However, if Smart Endo is used, it automatically recognizes gastrointestinal diseases in real time during the test and warns the examiner to shorten the detection of gastrointestinal diseases to less than 20 minutes.
Smart Endo consists of two 'Smart Endo Gastro' monitors with video AI SW, two 'Endoscopic Original Monitor' monitors, and a body computer system jointly produced by Infinite Healthcare, a global medical imaging company. Through a total of four monitors, medical staff can make precise diagnosis by comparing the screen to which AI SW is applied and the original screen in real time during the endoscope.
Gachon University Gil Hospital has obtained medical device certification from the Ministry of Food and Drug Safety for AI-Doctor for Gastric Cancer through medical records (EMR) and interviews of patients prior to the clinical launch of Smart Endo, an artificial intelligence (AI)-based gastroscopy diagnostic assistance software developed by Doctor Answer 2.0 task. The software uses the patient's blood test results, health checkups such as Helicobacter tests, electronic medical record data, and survey results on lifestyle to derive the risk rate of gastric cancer. Through this, it is expected that patients' endoscopic compliance can be increased by deriving a quantified risk rate for gastric cancer.
"Gastric cancer is the most common type of cancer in Korea, and early detection and treatment are very important because early detection can increase survival rates," said Kim Woo-kyung, director of Gachon University Gil Hospital. "It is expected to dramatically improve the detection rate of gastric lesions by incorporating artificial intelligence technology, reducing risk factors for gastric cancer treatment and contributing to the improvement of people's health."
|
bellho@sportschosun.com
