| Title： | Expert consensus on the application of the effect evaluation system of neoadjuvant therapy for rectal cancer based on artificial intelligence platform (2021 edition) |
| Edition： | Original |
| Classification： | Experts consensus |
| Field： | Diagnosis and Treatment |
| Countries and regions： | China |
| Guidelines users： | Gastrointestinal surgeon and radiologist |
| Evidence classification method： | Grade A recommendation: Good scientific evidence suggests that the benefits of this medical practice substantially outweigh the potential risks. The clinician should discuss the medical practice with the applicable patient; Grade B recommendation: at least moderate evidence suggests that the benefits of the medical practice outweigh the potential risks. The clinician should discuss the medical behavior with the appropriate patient; Grade C recommendation: At least reasonable scientific evidence suggests that the medical practice provides benefits, but the benefits are too close to the risks to make a general recommendation. The clinician is not required to provide this medical practice unless there are some individual considerations; Grade D recommendation: at least reasonable scientific evidence suggests that the potential risks of the medical practice outweigh the potential benefits; Clinicians should not routinely perform this medical practice on asymptomatic patients; Level I recommendation: The medical practice lacks scientific evidence, or the evidence is of poor quality, or is conflicting, such as risks and benefits that cannot be measured and assessed. The clinician should help the patient understand the uncertainty of the medical practice. |
| Development unit： | Colorectal Surgery Group of the Surgery Branch in the Chinese Medical Association; Beihang University State Key Laboratory of Virtual Reality Technology and Systems |
| Registration time: | 2021-06-13 |
| Registration number： | IPGRP-2021CN147 |
| Purpose of the guideline: | Neoadjuvant chemoradiotherapy (NAT) is one of the important treatment options for rectal cancer, and its therapeutic effect will determine different outcomes of patients, including waiting for observation treatment, local anal resection and total mesorectal resection. Therefore, a comprehensive and accurate evaluation of the effect of NAT will determine the choice of treatment options for patients. At present, magnetic resonance imaging (MRI) is internationally recognized as the most important means to evaluate the clinical staging and NAT effect of rectal cancer, but the reading of image results requires a large number of experienced radiologists, and the shortage and uneven distribution of personnel will inevitably lead to delay and bias of image results. In order to solve this problem, we combined artificial intelligence (AI) technology with imaging, and independently developed a colorectal cancer staging identification system based on AI platform, aiming to partially replace the work of imaging physicians, and realize accurate identification of colorectal cancer staging and lateral lymph nodes. We performed deep learning on imaging data of rectal cancer T stage, N stage, peripheral margin (CRM), extramural vascular invasion (EMVI) and apparent diffusion coefficient (ADC), and established a faster region-based convolutional neural network (FR-CNN). |
Latest registration
