No abstract

The development of digital cancer twins relies on the capture of high-resolution representations of individual cancer patients throughout the course of their treatment. Our research aims to improve the detection of metastatic disease over time from structured radiology reports by exposing prediction models to historical information. We demonstrate that Natural language processing (NLP) can generate better weak labels for semi-supervised classification of computed tomography (CT) reports when it is exposed to consecutive reports through a patient's treatment history. Around 714,454 structured radiology reports from Memorial Sloan Kettering Cancer Center adhering to a standardized departmental structured template were used for model development with a subset of the reports included for validation. To develop the models, a subset of the reports was curated for ground-truth: 7,732 total reports in the lung metastases dataset from 867 individual patients; 2,777 reports in the liver metastases dataset from 315 patients; and 4,107 reports in the adrenal metastases dataset from 404 patients. We use NLP to extract and encode important features from the structured text reports, which are then used to develop, train, and validate models. Three models—a simple convolutional neural network (CNN), a CNN augmented with an attention layer, and a recurrent neural network (RNN)—were developed to classify the type of metastatic disease and validated against the ground truth labels. The models use features from consecutive structured text radiology reports of a patient to predict the presence of metastatic disease in the reports. A single-report model, previously developed to analyze one report instead of multiple past reports, is included and the results from all four models are compared based on accuracy, precision, recall, and F1-score. The best model is used to label all 714,454 reports to generate metastases maps. Our results suggest that NLP models can extract cancer progression patterns from multiple consecutive reports and predict the presence of metastatic disease in multiple organs with higher performance when compared with a single-report-based prediction. It demonstrates a promising automated approach to label large numbers of radiology reports without involving human experts in a time- and cost-effective manner and enables tracking of cancer progression over time.

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.

PURPOSE To assess the accuracy of a natural language processing (NLP) model in extracting splenomegaly described in patients with cancer in structured computed tomography radiology reports. METHODS In this retrospective study between July 2009 and April 2019, 3,87,359 consecutive structured radiology reports for computed tomography scans of the chest, abdomen, and pelvis from 91,665 patients spanning 30 types of cancer were included. A randomized sample of 2,022 reports from patients with colorectal cancer, hepatobiliary cancer (HB), leukemia, Hodgkin lymphoma (HL), and non-HL patients was manually annotated as positive or negative for splenomegaly. NLP model training/testing was performed on 1,617/405 reports, and a new validation set of 400 reports from all cancer subtypes was used to test NLP model accuracy, precision, and recall. Overall survival was compared between the patient groups (with and without splenomegaly) using Kaplan-Meier curves. RESULTS The final cohort included 3,87,359 reports from 91,665 patients (mean age 60.8 years; 51.2% women). In the testing set, the model achieved accuracy of 92.1%, precision of 92.2%, and recall of 92.1% for splenomegaly. In the validation set, accuracy, precision, and recall were 93.8%, 92.9%, and 86.7%, respectively. In the entire cohort, splenomegaly was most frequent in patients with leukemia (32.5%), HB (17.4%), non-HL (9.1%), colorectal cancer (8.5%), and HL (5.6%). A splenomegaly label was associated with an increased risk of mortality in the entire cohort (hazard ratio 2.10; 95% CI, 1.98 to 2.22; P < .001). CONCLUSION Automated splenomegaly labeling by NLP of radiology report demonstrates good accuracy, precision, and recall. Splenomegaly is most frequently reported in patients with leukemia, followed by patients with HB.