The small number of patients with follow-up multiparametric MRI findings showing worsening disease supports the role of MRI in patients with early-stage prostate cancer. Multiparametric MRI is useful in monitoring patients on active surveillance and may identify patients with clinically significant cancer amenable to definitive treatment.

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.

Our aim was to evaluate clinical management and outcomes in cancer patients who had an indeterminate Computed Tomographic Pulmonary Angiogram (CTPA) for the assessment of pulmonary embolus. We reviewed 1000 CTPA studies and identified 251 limited (indeterminate) CTPA. We examined follow-up imaging and reviewed clinical management decisions and any positive diagnosis of venous thromboembolic disease (VTE) within the subsequent 90 days. 60 patients (23.9%) had a follow-up imaging study within five days. 8 had a positive study for VTE disease within 5 days. 3 patients (1.2%) were placed on anticoagulation therapy based on the limited CT result.

Post-PFC mpMRI, at Phoenix suspicion of BCR, may help identify a significant number of patients failing post-PFC.

INTRODUCTION AND OBJECTIVES:The Prostate Imaging-Reporting and Data System (PIRADS) score was developed to evaluate the likelihood of malignancy for lesions seen in the peripheral zone and transition zone on multi-parametric Magnetic Resonance Imaging (MRI) of the prostate. We aim to determine if this same scoring system can be used to evaluate central zone lesions on MRI.METHODS: A retrospective review was performed of 73 patients who underwent MR/US fusion-guided biopsy of 139 suspicious lesions between February 2014 and October 2015. All patients underwent a 3 Tesla multi-parametric MRI. Indications for MRI include an abnormal digital rectal exam, PSA velocity >0.75ng/dl/year and patients on active surveillance. Our multi-parametric MRI sequence involved T2, diffusion weighted imaging and dynamic contrast enhancement. Using a 3-dimensional model software [InVivo (Phillips), Gainesville (USA)], at least 3 MR/US fusion-guided biopsies were performed on each prostate lesion seen on MRI regardless of PIRADS score under local anesthesia in the outpatient clinic.RESULTS: There were 80 peripheral zone lesions, 32 transitional zone lesions and 27 central zone lesions that were biopsied.Median PIRADS score for central zone lesions was 3 (range 1-5). Compared to the peripheral and transition zone, central zone lesions graded PIRADS 4 and 5 were more likely to be false positive, p¼0.012. Only two patients (7%) had clinically significant prostate cancer (Gleason >3+3) seen on central zone lesion. Both patients had lesions which were graded as PIRADS 3. Both lesions involved the transition zone as well and encompassed at least 50% of the entire central zone and transition zone. Both patients previously had transrectal ultrasound guided biopsy of the prostates which were negative for cancer. Both patients underwent a robotic assisted laparoscopic prostatectomy which yielded a Gleason score that were similar to MRI fusion biopsy.CONCLUSIONS:Lesions involving only the central zone seen on multi-parametric MRI are less concerning for malignancy and should not be given equal weightage as peripheral zone lesions. In our series, no lesions involving solely the central zone, regardless of PIRADS score was positive for malignancy on MR/US fusion-guided biopsy. A better PIRADS scoring system should be developed to help identify central zone lesions with malignant potential.