Two radiologists with 3 years and 8 years of experience in rectal cancer MRI interpretation who were blinded to the histopathology results evaluated the MR images.Ĭonventional semantic evaluation indicators included MRI-reported LN status, which were performed using the qualitative criteria of the LNs according to the updated recommendations from the 2016 European Society of Gastrointestinal and Abdominal Radiology (ESGAR) consensus meeting ( 18). The process of patient selection is summarized in Figure 1. A total of 186 patients met the criteria and were included in this study they were divided randomly into a training cohort (n = 123) and a testing cohort (n = 63) at a ratio of 2:1. The exclusion criteria were as follows: (i) distant metastases (ii) not undergoing surgery at our hospital or lack of diffusion-weighted imaging (DWI) or high-resolution T2-weighted imaging (T2WI) data (iii) insufficient MRI quality to obtain measurements ( e.g., owing to motion artifacts) and (iv) lack of presurgical carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) data. The inclusion criteria were as follows: (i) rectal MRI examination was performed within the 2 weeks before surgery (ii) the distal border of the tumor was ≤15 cm above the anal verge based on colonoscopy (iii) subsequent radical surgical resection was performed (iv) postoperative histopathological examination confirmed rectal adenocarcinoma and (v) all LNs were assessed. The data of 238 consecutive patients with rectal cancer from January 2016 to December 2018 were initially retrieved from the institutional database. This would allow clinicians to make personalized treatment plans. This study aimed to develop and validate a multiregional radiomics prediction model based on MRI and combine it with clinical-semantic data for the individualized preoperative prediction of LN metastasis in rectal cancer patients. To the best of our knowledge, the topic has not been previously studied. MRI can provide multiparameter images different from those obtained by CT, so it is of interest whether there exists an association between LN status and multiregional radiomics features of multiparametric MR images in rectal cancer patients. In previous studies, a CT radiomics signature-based nomogram ( 16) and T2-weighted histogram of the primary tumor ( 17) have been applied and shown to successfully discriminate LN metastasis in colorectal- and rectal cancer patients. From standard-of-care medical images, data can be extracted via high-throughput mining of quantitative image features, which are undetectable by the naked eye, and applied within clinical-decision support systems ( 9– 13) radiomics plays an important role in early diagnosis, treatment evaluation, and tumor prognosis prediction, ultimately aiding in the achievement of precision medicine ( 11, 14, 15).
Unlike traditional image evaluation methods, radiomics is an emerging and effective method for quantitatively analyzing the classification and prognosis of diseases using medical imaging ( 10). All diagnostic clues rely heavily on the size, shape, and margins of LNs, but these semantic characteristics alone are insufficient to reliably distinguish malignant from benign LNs in rectal cancer patients ( 2, 4, 5, 9). However, MRI, computed tomography (CT) and endorectal ultrasound cannot reliably evaluate LN metastasis ( 2, 4, 8). Magnetic resonance imaging (MRI) is considered the most accurate method to assess the primary staging of rectal cancer ( 2). However, preoperative LN staging in rectal cancer patients remains a challenge for radiologists ( 4). Therefore, accurate preoperative assessment of LN status or assessment of the N stages of regional LNs in rectal cancer patients via medical imaging is essential for precise individualized decision making and patient prognosis ( 2, 6, 7). Lymph node (LN) status plays a vital role in determining whether to perform adjuvant therapy or additional surgical resection ( 2– 6). Nearly one-third of colorectal tumors are located in the rectum ( 2). Colorectal cancer was the third most common type of malignant tumor and the second leading cause of cancer death in the world in 2018 ( 1).
0 kommentar(er)
