Ai-Assisted Diagnosis System for Acute Pancreatitis

Project: National Science and Technology CouncilNational Science and Technology Council Academic Grants

Project Details

Abstract

At present, the surgical treatment of acute pancreatitis often uses minimally invasive percutaneous drainage. In the drainage treatment, the physician usually guides the drainage tube into the abdominal cavity using hand-eye coordination and personal experience under the guidance of CT images. The key to the success of CT-guided puncture and drainage is to be able to accurately locate the positioning area and the path planning before puncture. The positioning and puncture path in clinical treatment is affected by the personal experience and ability of the physician.This study will use CT images to develop an AI-assisted diagnostic system for acute pancreatitis, use artificial intelligence to determine residual pancreas and Inflamed areas in CT images, and combine three-dimensional images and related structural positioning to plan intubation paths and provide physicians. The plan will be divided into four parts to complete the plan: (1) automatic segmentation of medical images, spatial positioning and edge detection, and mark 2000 case on the shared platform; (2) research on 3D image reconstruction and calculation CTSI scores; ( 3) research the automatic planning of drainage tube puncture path based on clinical multiple constraints; (4) accelerate the modeling and path planning with GPU and develop auxiliary systems.The pancreatitis sharing platform will contain 2000 cases of diagnosis and treatment data, case reports, test reports, image pictures, and other data, covering the patient's history from the ICU to the general ward, rehabilitation center, and data desensitization. The project uses artificial intelligence model algorithms, verifying model accuracy and comparing CTSI scores through experiments and images marked by doctors, and using the model to calculate the 3D images of the CTSI scores to convert them to the location of the area requiring treatment (free fluid) , plan the drainage tube puncture path, provide the objective reference information for the doctor before the operation in the future.

Project IDs

Project ID:PB10907-2885
External Project ID:MOST109-2221-E182-015
StatusFinished
Effective start/end date01/08/2031/07/21

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.