Jennifer S. Golia Pernicka

Objectives To evaluate the completeness and discordance of outside rectal MRI initial staging reports compared to second-opinion reviews, and to assess the potential clinical impact of major discordance on treatment decisions in patients with rectal adenocarcinoma. Materials and methods A retrospective analysis of outside rectal MRI reviews submitted for second-opinion interpretation by subspecialized radiologists from June 2014–March 2020 was conducted. Outside and second review reports were compared side-by-side; cases with discordance (and those with major discordance, i.e., may alter treatment, particularly) were identified. Two colorectal surgeons, blinded to report origins, reviewed cases with major discordance to evaluate their theoretical impact on patient management and rated their confidence level of the reports on a five-point Likert scale (1=lowest confidence). Results In 461 patients (median age, 57 years [IQR: 49–67]; 274 male), compared to outside reviews, second reviews demonstrated improved report completeness across tumor characteristics, local extent, and nodal/metastatic disease clinical staging categories. The largest reporting gaps were in tumor morphology (66.4%, 306/461 vs. 98.7%, 455/461) and extramural venous invasion (29.1%, 134/461 vs. 93.9%, 433/461). Overall, 53.8% (248/461) of cases showed discordance, with 56.5% (140/248) classified as major. In this subset, second reviews led to fewer cases with insufficient information, e.g., 18.6% (26/140) to 7.1% (10/140) for Surgeon 1, and changes in surgical planning in 38.1–46.3% (43/113 to 44/95) of patients. Surgeons rated reports from second reviews with higher confidence scores (median, 4 vs. 2–3, p < 0.001). Conclusions Second-opinion reviews may impact patient care and improve surgeons’ confidence for treatment planning. Key Points Question Do subspecialist radiologist second-opinion reviews improve rectal MRI report completeness and affect treatment decisions in rectal adenocarcinoma staging? Findings While outside reviews often under-reported key imaging findings, second reviews improved report completeness, which led to increased confidence by our surgeons in treatment planning. Clinical relevance These findings underscore the value of expert interpretation and highlight the need for optimized MRI protocols and broader adoption of structured reporting in rectal cancer staging. Graphical Abstract

Purpose To identify computed tomography (CT) and clinical characteristics that may aid the radiologist and clinician in differentiating desmoid tumor from peritoneal metastasis in patients with a history of resected colon cancer. Methods This retrospective, single-center study included patients with resected colon cancer with one or more new peritoneal nodules on CT that then yielded pathology-proven desmoid tumor or peritoneal metastasis. Three radiologists, blinded to histopathology results, assessed peritoneal nodule characteristics on CT, including location, multiplicity, shape, contour, margin, density, homogeneity, presence of other metastatic disease, and size. Clinical characteristics assessed included age, biological sex, race, history of familial adenomatosis polyposis, carcinoembryonic antigen (CEA) level within 4 weeks of CT, time between colon cancer resection and diagnosis of either desmoid tumor or peritoneal metastasis, colon primary histology, and pathology TNM stage. Clinical characteristics as well as imaging characteristics (assessed by the most experienced radiologist) were tested for significance using Fisher’s exact test or Wilcoxon rank sum test. Results Of 30 patients (median age at colon cancer resection, 51 years [range, 20–79]; 18/30 (60%) male), 18/30 (60%) had desmoid tumor and 12/30 (40%) had peritoneal metastasis. On CT, location in the root of the mesentery and whorled margin were associated with desmoid tumor (p = 0.02 for both), and multiplicity and presence of other metastatic disease with peritoneal metastasis (p = 0.002 and 0.006). CEA > 5 ng/mL was associated with peritoneal metastasis (p = 0.0005). Conclusion Some CT characteristics and elevated CEA may help differentiate desmoid tumor from peritoneal metastasis, guiding prognosis and disease management.

Anal cancer treatment response assessment can be challenging with both magnetic resonance imaging (MRI) and clinical evaluation considered essential. MRI, in particular, has shown to be useful for the assessment of treatment response, the detection of recurrent disease in follow up and surveillance, and the evaluation of possible post-treatment complications as well as complications from the tumor itself. In this review, we focus on the role of imaging, mainly MRI, in anal cancer treatment response assessment. We also describe the treatment complications that can occur, and the imaging findings associated with those complications.

The role and method of image-based staging of anal cancer has evolved with the rapid development of newer imaging modalities and the need to address the rising incidence of this rare cancer. In 2014, the European Society of Medical Oncology mandated pelvic magnetic resonance imaging (MRI) for anal cancer and subsequently other societies such as the National Comprehensive Cancer Network followed suit with similar recommendations. Nevertheless, great variability exists from center to center and even within individual centers. Notably, this is in stark contrast to the imaging of the anatomically nearby rectal cancer. As participating team members for this malignancy, we embarked on a comprehensive literature review of anal cancer imaging to understand the relative merits of these new technologies which developed after computed tomography (CT), e.g., MRI and positron emission tomography/computed tomography (PET/CT). The results of this literature review helped to inform our next stage: questionnaire development regarding the imaging of anal cancer. Next, we distributed the questionnaire to members of the Society of Abdominal Radiology (SAR) Rectal and Anal Disease-Focused Panel, a group of abdominal radiologists with special interest, experience, and expertise in rectal and anal cancer, to provide expert radiologist opinion on the appropriate anal cancer imaging strategy. In our expert opinion survey, experts advocated the use of MRI in general (65% overall and 91–100% for primary staging clinical scenarios) and acknowledged the superiority of PET/CT for nodal assessment (52–56% agreement for using PET/CT in primary staging clinical scenarios compared to 30% for using MRI). We therefore support the use of MRI and PET and suggest further exploration of PET/MRI as an optimal combined evaluation. Our questionnaire responses emphasized the heterogeneity in imaging practice as performed at numerous academic cancer centers across the United States and underscore the need for further reconciliation and establishment of best imaging practice guidelines for optimized patient care in anal cancer.

This work presents a deep-learning-based denoising technique to accelerate the acquisition of high b-value diffusion-weighted MRI for rectal cancer. A denoising convolutional neural network (DCNN) with a combined L1–L2 loss function was developed to denoise high b-value diffusion-weighted MRI data acquired with fewer repetitions (NEX: number of excitations) using the low b-value image as an anatomical guide. DCNN was trained using 85 datasets acquired on patients with rectal cancer and tested on 20 different datasets with NEX = 1, 2, and 4, corresponding to acceleration factors of 16, 8, and 4, respectively. Image quality was assessed qualitatively by expert body radiologists. Reader 1 scored similar overall image quality between denoised images with NEX = 1 and NEX = 2, which were slightly lower than the reference. Reader 2 scored similar quality between NEX = 1 and the reference, while better quality for NEX = 2. Denoised images with fourfold acceleration (NEX = 4) received even higher scores than the reference, which is due in part to the effect of gas-related motion in the rectum, which affects longer acquisitions. The proposed deep learning denoising technique can enable eightfold acceleration with similar image quality (average image quality = 2.8 ± 0.5) and fourfold acceleration with higher image quality (3.0 ± 0.6) than the clinical standard (2.5 ± 0.8) for improved diagnosis of rectal cancer.

PURPOSE: Natural language processing (NLP) applied to radiology reports can help identify clinically relevant M1 subcategories of patients with colorectal cancer (CRC). The primary purpose was to compare the overall survival (OS) of CRC according to American Joint Committee on Cancer TNM staging and explore an alternative classification. The secondary objective was to estimate the frequency of metastasis for each organ. METHODS: Retrospective study of CRC who underwent computed tomography (CT) chest, abdomen, and pelvis between July 1, 2009, and March 26, 2019, at a tertiary cancer center, previously labeled for the presence or absence of metastasis by an NLP prediction model. Patients were classified in M0, M1a, M1b, and M1c (American Joint Committee on Cancer), or an alternative classification on the basis of the metastasis organ number: M1, single; M2, two; M3, three or more organs. Cox regression models were used to estimate hazard ratios; Kaplan-Meier curves were used to visualize survival curves using the two M1 subclassifications. RESULTS: Nine thousand nine hundred twenty-eight patients with a total of 48,408 CT chest, abdomen, and pelvis reports were included. On the basis of NLP prediction, the median OS of M1a, M1b, and M1c was 4.47, 1.72, and 1.52 years, respectively. The median OS of M1, M2, and M3 was 4.24, 2.05, and 1.04 years, respectively. Metastases occurred most often in liver (35.8%), abdominopelvic lymph nodes (32.9%), lungs (29.3%), peritoneum (22.0%), thoracic nodes (19.9%), bones (9.2%), and pelvic organs (7.5%). Spleen and adrenal metastases occurred in < 5%. CONCLUSION: NLP applied to a large radiology report database can identify clinically relevant metastatic phenotypes and be used to investigate new M1 substaging for CRC. Patients with three or more metastatic disease organs have the worst prognosis, with an OS of 1 year.

Objective To review existing structured MRI reports for primary staging of rectal cancer and create a new, freely available structured report based on multidisciplinary expert opinion and literature review. Methods Twenty abdominal imaging experts from the Society of Abdominal Radiology (SAR)’s Disease Focused Panel (DFP) on Rectal and Anal Cancer completed a questionnaire and participated in a subsequent consensus meeting based on the RAND-UCLA Appropriateness Method. Twenty-two items were classified via a group survey as “appropriate” or “inappropriate” (defined by ≥ 70% consensus), or “needs group discussion” (defined by < 70% consensus). Certain items were also discussed with multidisciplinary team members from colorectal surgery, oncology and pathology. Results After completion of the questionnaire, 16 (72%) items required further discussion (< 70% consensus). Following group discussion, consensus was achieved for 21 (95%) of the items. Based on the consensus meeting, a revised structured report was developed. The most significant modifications included (1) Exclusion of the T2/early T3 category; (2) Replacement of the term “circumferential resection margin (CRM)” with “mesorectal fascia (MRF)”; (3) A revised definition of “mucinous content”; (4) Creation of two distinct categories for suspicious lymph nodes (LNs) and tumor deposits; and (5) Classification of suspicious extra-mesorectal LNs by anatomic location. Conclusion The SAR DFP on Rectal and Anal Cancer recommends using this newly updated reporting template for primary MRI staging of rectal cancer. Graphical abstract

International challenges have become the de facto standard for comparative assessment of image analysis algorithms. Although segmentation is the most widely investigated medical image processing task, the various challenges have been organized to focus only on specific clinical tasks. We organized the Medical Segmentation Decathlon (MSD)—a biomedical image analysis challenge, in which algorithms compete in a multitude of both tasks and modalities to investigate the hypothesis that a method capable of performing well on multiple tasks will generalize well to a previously unseen task and potentially outperform a custom-designed solution. MSD results confirmed this hypothesis, moreover, MSD winner continued generalizing well to a wide range of other clinical problems for the next two years. Three main conclusions can be drawn from this study: (1) state-of-the-art image segmentation algorithms generalize well when retrained on unseen tasks; (2) consistent algorithmic performance across multiple tasks is a strong surrogate of algorithmic generalizability; (3) the training of accurate AI segmentation models is now commoditized to scientists that are not versed in AI model training.