Faculty Bibliography

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.

Accurate evaluation of tumor response to treatment is critical to allow personalized treatment regimens according to the predicted response and to support clinical trials investigating new therapeutic agents by providing them with an accurate response indicator. Recent advances in medical imaging, computer hardware, and machine-learning algorithms have resulted in the increased use of these tools in the field of medicine as a whole and specifically in cancer imaging for detection and characterization of malignant lesions, prognosis, and assessment of treatment response. Among the currently available imaging techniques, magnetic resonance imaging (MRI) plays an important role in the evaluation of treatment assessment of many cancers, given its superior soft-tissue contrast and its ability to allow multiplanar imaging and functional evaluation. In recent years, deep learning (DL) has become an active area of research, paving the way for computer-assisted clinical and radiological decision support. DL can uncover associations between imaging features that cannot be visually identified by the naked eye and pertinent clinical outcomes. The aim of this review is to highlight the use of DL in the evaluation of tumor response assessed on MRI. In this review, we will first provide an overview of common DL architectures used in medical imaging research in general. Then, we will review the studies to date that have applied DL to magnetic resonance imaging for the task of treatment response assessment. Finally, we will discuss the challenges and opportunities of using DL within the clinical workflow.

PURPOSE/OBJECTIVE:To evaluate computed tomography (CT) findings in patients with ovarian cancer presenting to a comprehensive cancer center's urgent care unit with acute abdominal symptoms. METHODS:This retrospective study included consecutive patients with ovarian cancer who underwent abdominal CT at a comprehensive cancer center's urgent care unit between January 1, 2018, and January 14, 2020, due to acute abdominal symptoms. Two abdominal radiologists reviewed the abdominal CT reports, categorizing imaging findings as follows: (a) no new or acute finding, (b) new or increased bowel or gastric obstruction, (c) new or increased ascites, (d) new or increased peritoneal carcinomatosis, (e) new or increased nonperitoneal metastases, (f) new inflammatory or infectious changes, (g) new or increased hydronephrosis, (h) new or increased biliary dilatation, (i) new vascular complications, or (j) new bowel perforation. RESULTS:A total of 200 patients (mean age, 59 years; range, 22-87) underwent a total of 259 abdominal CT scans, of which 217/259 (83.8%) scans were found to have new or increased findings. A total of 115/259 (44.4%) scans had only one finding while 102/259 (39.4%) scans had 2 or more findings. Altogether, 382 new or increased findings were detected: findings were most commonly related to bowel or gastric obstruction (92/382, 24.1%) with small bowel obstruction being the most common finding (80/382, 20.9%); ascites (78/382, 20.4%); peritoneal carcinomatosis (62/382, 16.2%); and nonperitoneal metastases (62/382, 16.2%). Inflammatory or infectious findings accounted for 30/382 (7.9%) findings. CONCLUSION/CONCLUSIONS:Most patients with ovarian cancer presenting with acute abdominal had relevant positive findings on abdominal CT, with small bowel obstruction being the most common finding.

A T1 sequence on routine baseline staging rectal magnetic resonance imaging (MRI) is thought to help detect bone lesions. Our primary aim was to evaluate the incidence of bone lesions encountered on baseline staging rectal MRI, particularly the prevalence of bone metastases. This retrospective study included patients with rectal adenocarcinoma who underwent baseline rectal MRI at our institution between January 2010 and December 2017. The MRI report was reviewed for presence of bone lesions. When found, lesion type, presence of axial T1 non-fat-suppressed sequence, primary tumor T-stage, and presence of other organ metastases were recorded. In the absence of bone biopsy, the reference standard was follow-up imaging via computed tomography (CT), MRI, and/or positron emission tomography/CT (PET/CT) ≥ 1 year after the baseline MRI. The Wilcoxon rank-sum test and Fisher's exact test were used to compare clinicopathologic data of patients with malignant or benign bone lesions. A total of 1197 patients were included. 62/1197 patients (mean age 56.8 years (SD: 13.8), with 39 men) had bone lesions on baseline imaging, with 6 being bone metastases (0.5%, 95% CI 0.2%-1.1%). Of the 6 patients with bone metastases, 5/6 had other metastases (i.e., liver, lung) at baseline. Bone metastases on baseline rectal MRI performed for rectal adenocarcinoma are extremely rare. Furthermore, bone metastases without other organ (i.e., liver, lung) involvement is extremely rare.

Artificial intelligence (AI) has been successful at solving numerous problems in machine perception. In radiology, AI systems are rapidly evolving and show progress in guiding treatment decisions, diagnosing, localizing disease on medical images, and improving radiologists' efficiency. A critical component to deploying AI in radiology is to gain confidence in a developed system's efficacy and safety. The current gold standard approach is to conduct an analytical validation of performance on a generalization dataset from one or more institutions, followed by a clinical validation study of the system's efficacy during deployment. Clinical validation studies are time-consuming, and best practices dictate limited re-use of analytical validation data, so it is ideal to know ahead of time if a system is likely to fail analytical or clinical validation. In this paper, we describe a series of sanity tests to identify when a system performs well on development data for the wrong reasons. We illustrate the sanity tests' value by designing a deep learning system to classify pancreatic cancer seen in computed tomography scans.

OBJECTIVES/OBJECTIVE:To investigate what findings are new on contrast-enhanced abdominopelvic CT in patients infected with SARS-CoV-2. METHODS:Contrast-enhanced CT of the abdomen and pelvis of patients with COVID-19 at a tertiary oncologic center acquired over a 2-month period were reviewed independently by two readers and scored for new imaging abnormalities compared with a prior scan. CT scans were included if the study was performed between - 3 and 45 days from the time of COVID-19 diagnosis. Clinical information was gathered from the medical records. RESULTS:A total of 63 patients (34 male, 29 female; mean age 60.6 years, range 24.4-85.0 years) were included in this retrospective cross-sectional study. Aside from new ground glass opacities seen at the lung bases (29/63, 46.0%), the most common findings were new thickening of the stomach, small bowel or colon or fluid-filled colon (14/63, 22.2%), new small volume ascites (7/63, 14.3%), gallbladder distention in those without prior cholecystectomy (3/43, 7.0%), and single cases each of acute pancreatitis (1/63, 1.6%) as well as new portal vein thrombosis (1/63, 1.6%). CONCLUSION/CONCLUSIONS:Aside from lung base ground glass opacities, the most common new imaging abnormality on abdominopelvic CT in patients with COVID-19 finding in our cohort was abnormalities of the gastrointestinal tract, followed by small volume ascites, gallbladder distention, and isolated cases of pancreatitis and portal vein thrombosis. These findings overlap with those previously reported that did not have a prior scan for comparison, and provide supportive evidence that some of these findings may be related to SARS-CoV-2 infection.

PURPOSE/OBJECTIVE:This study explored the clinicopathologic outcomes of rectal tumor morphological descriptors used in a synoptic rectal MRI reporting template and determined that prognostic differences were observed. METHODS:This retrospective study was conducted at a comprehensive cancer center. Fifty patients with rectal tumors for whom the synoptic descriptor "polypoid" was chosen by three experienced radiologists were compared with ninety comparator patients with "partially circumferential" and "circumferential" rectal tumors. Two radiologists re-evaluated all cases. The outcome measures were agreement among two re-interpreting radiologists, clinical T staging with MRI (mrT) and descriptive nodal features, and degrees of wall attachment of tumors (on MRI) compared with pathological (p) T and N stage when available. RESULTS:Re-evaluation by two radiologists showed moderate to excellent agreement in tumor morphology, presence of a pedicle, and degree of wall attachment (k = 0.41-0.76) and excellent agreement on lymph node presence and size (ICC = 0.83-0.91). Statistically significant lower mrT stage was noted for polypoid morphology, wherein 98% were mrT1/2, while only 7% and 2% of partially circumferential and circumferential tumors respectively were mrT1/2. Pathologic T and N stages among the three morphologies also differed significantly, with only 14% of polypoid cases higher than stage pT2 compared to 48% of partially circumferential cases and 60% of circumferential cases. CONCLUSION/CONCLUSIONS:Using a "polypoid" morphology in rectal cancer MRI synoptic reports revealed a seemingly distinct phenotype with lower clinical and pathologic T and N stages when compared with alternative available descriptors. PRECIS/CONCLUSIONS:"Polypoid" morphology in rectal cancer confers a lower clinical and pathologic T and N stage and may be useful in determining whether to proceed with surgery versus neoadjuvant treatment.

BACKGROUND:To evaluate quantitative iodine parameters from the arterial phase dual-energy computed tomography (DECT) scans as an imaging biomarker for tumor grade (TG), mitotic index (MI), and Ki-67 proliferation index of hepatic metastases from neuroendocrine tumors (NETs) of the gastrointestinal (GI) tract. Imaging biomarkers have the potential to provide relevant clinical information about pathologic processes beyond lesion morphology. NETs are a group of rare, heterogeneous neoplasms classified by World Health Organization (WHO) TG, which is derived from MI and Ki-67 proliferation index. Imaging biomarkers for these pathologic features and TG may be useful. METHODS:Between January 2014 and April 2019, 73 unique patients with hepatic metastases from NET of the GI tract underwent DECT of the abdomen with an arterial phase were analyzed after exclusions. Using GSIViewer software (GE Healthcare, Madison, Wisconsin), elliptical regions of interest (ROIs) were placed over selected hepatic metastases by a fellowship trained abdominal radiologist. Quantitative iodine concentration (IC) data was extracted from the lesion ROIs, and the normalized IC (lesion IC/aorta IC) and relative IC (lesion IC/liver IC) for each liver were calculated. Spearman correlation was calculated for lesion mean IC, normalized IC, and relative IC to both Ki-67 proliferation and mitotic indices. Student's t-test was performed to compare lesion mean IC, normalized IC and relative IC between WHO TGs. RESULTS:There was very weak correlation between both normalized IC and relative IC for both Ki-67 proliferation and mitotic indices. A significant difference was not observed between normalized IC and relative IC to distinguish metastases from G1 and G2/3 tumors. CONCLUSIONS:Our study finds limited potential for quantitative parameters from DECT to distinguish neuroendocrine hepatic metastases by WHO TG, as well as limited potential as an imaging biomarker for Ki-67 proliferation and mitotic indices in this setting. Our findings of a lack of correlation between Ki-67 and quantitative iodine parameters stands in contrast to existing literature that reports positive correlations for these parameters in the rectum and stomach.

Complete mesocolic excision (CME) with central vascular ligation (CVL) has gained traction in recent years as an effective advanced surgical technique for surgical treatment of colon cancer. In this review, the concept of CME is described, along with relevant variant vascular anatomy and local staging implications for radiologists. Future directions in research are also explored, with a focus on correlation between radiology and pathology.