Investigating the impact of structured reporting on the linguistic standardization of radiology reports through natural language processing over a 10-year period.

OBJECTIVES: To investigate how a transition from free text to structured reporting affects reporting language with regard to standardization and distinguishability. METHODS: A total of 747,393 radiology reports dictated between January 2011 and June 2020 were retrospectively analyzed. The body and cardiothoracic imaging divisions introduced a reporting concept using standardized language and structured reporting templates in January 2016. Reports were segmented by a natural language processing algorithm and converted into a 20-dimension document vector. For analysis, dimensionality was reduced to a 2D visualization with t-distributed stochastic neighbor embedding and matched with metadata. Linguistic standardization was assessed by comparing distinct report types' vector spreads (e.g., run-off MR angiography) between reporting standards. Changes in report type distinguishability (e.g., CT abdomen/pelvis vs. MR abdomen) were measured by comparing the distance between their centroids. RESULTS: Structured reports showed lower document vector spread (thus higher linguistic similarity) compared with free-text reports overall (21.9 [free-text] vs. 15.9 [structured]; - 27.4%; p < 0.001) and for most report types, e.g., run-off MR angiography (15.2 vs. 1.8; - 88.2%; p < 0.001) or double-rule-out CT (26.8 vs. 10.0; - 62.7%; p < 0.001). No changes were observed for reports continued to be written in free text, e.g., CT head reports (33.2 vs. 33.1; - 0.3%; p = 1). Distances between the report types' centroids increased with structured reporting (thus better linguistic distinguishability) overall (27.3 vs. 54.4; + 99.3 ± 98.4%) and for specific report types, e.g., CT abdomen/pelvis vs. MR abdomen (13.7 vs. 37.2; + 171.5%). CONCLUSION: Structured reporting and the use of factual language yield more homogenous and standardized radiology reports on a linguistic level, tailored to specific reporting scenarios and imaging studies. CLINICAL RELEVANCE: Information transmission to referring physicians, as well as automated report assessment and content extraction in big data analyses, may benefit from standardized reporting, due to consistent report organization and terminology used for pathologies and normal findings. KEY POINTS: • Natural language processing and t-distributed stochastic neighbor embedding can transform radiology reports into numeric vectors, allowing the quantification of their linguistic standardization. • Structured reporting substantially increases reports' linguistic standardization (mean: - 27.4% in vector spread) and distinguishability (mean: + 99.3 ± 98.4% increase in vector distance) compared with free-text reports. • Higher standardization and homogeneity outline potential benefits of structured reporting for information transmission and big data analyses.

An Analysis of Early Results after Valve Replacement in Isolated Aortic Valve Stenosis by Using Sutureless vs. Stented Bioprostheses: A Single-Center Middle-Income Country Experience.

Background and Objectives: There is a lack of data about the survival of patients after the implantation of sutureless relative to stented bioprostheses in middle-income settings. The objective of this study was to compare the survival of people with isolated severe aortic stenosis after the implantation of sutureless and stented bioprostheses in a tertiary referral center in Serbia. Materials and Methods: This retrospective cohort study included all people treated for isolated severe aortic stenosis with sutureless and stented bioprostheses from 1 January 2018 to 1 July 2021 at the Institute for Cardiovascular Diseases "Dedinje". Demographic, clinical, perioperative and postoperative data were extracted from the medical records. The follow-up lasted for a median of 2 years. Results: The study sample comprised a total of 238 people with a stented (conventional) bioprosthesis and 101 people with a sutureless bioprosthesis (Perceval). Over the follow-up, 13.9% of people who received the conventional and 10.9% of people who received the Perceval valve died (p = 0.400). No difference in the overall survival was observed (p = 0.797). The multivariate Cox proportional hazard model suggested that being older, having a higher preoperative EuroScore II, having a stroke over the follow-up period and having valve-related complications were independently associated with all-cause mortality over a median of 2 years after the bioprosthesis implantation. Conclusions: This research conducted in a middle-income country supports previous findings in high-income countries regarding the survival of people with sutureless and stented valves. Survival after bioprosthesis implantation should be monitored long-term to ensure optimum postoperative outcomes.

Natural Language Processing and Graph Theory: Making Sense of Imaging Records in a Novel Representation Frame.

BACKGROUND: A concise visualization framework of related reports would increase readability and improve patient management. To this end, temporal referrals to prior comparative exams are an essential connection to previous exams in written reports. Due to unstructured narrative texts' variable structure and content, their extraction is hampered by poor computer readability. Natural language processing (NLP) permits the extraction of structured information from unstructured texts automatically and can serve as an essential input for such a novel visualization framework. OBJECTIVE: This study proposes and evaluates an NLP-based algorithm capable of extracting the temporal referrals in written radiology reports, applies it to all the radiology reports generated for 10 years, introduces a graphical representation of imaging reports, and investigates its benefits for clinical and research purposes. METHODS: In this single-center, university hospital, retrospective study, we developed a convolutional neural network capable of extracting the date of referrals from imaging reports. The model's performance was assessed by calculating precision, recall, and F1-score using an independent test set of 149 reports. Next, the algorithm was applied to our department's radiology reports generated from 2011 to 2021. Finally, the reports and their metadata were represented in a modulable graph. RESULTS: For extracting the date of referrals, the named-entity recognition (NER) model had a high precision of 0.93, a recall of 0.95, and an F1-score of 0.94. A total of 1,684,635 reports were included in the analysis. Temporal reference was mentioned in 53.3% (656,852/1,684,635), explicitly stated as not available in 21.0% (258,386/1,684,635), and omitted in 25.7% (317,059/1,684,635) of the reports. Imaging records can be visualized in a directed and modulable graph, in which the referring links represent the connecting arrows. CONCLUSIONS: Automatically extracting the date of referrals from unstructured radiology reports using deep learning NLP algorithms is feasible. Graphs refined the selection of distinct pathology pathways, facilitated the revelation of missing comparisons, and enabled the query of specific referring exam sequences. Further work is needed to evaluate its benefits in clinics, research, and resource planning.

Automated Detection of Pancreatic Cystic Lesions on CT Using Deep Learning.

Pancreatic cystic lesions (PCL) are a frequent and underreported incidental finding on CT scans and can transform into neoplasms with devastating consequences. We developed and evaluated an algorithm based on a two-step nnU-Net architecture for automated detection of PCL on CTs. A total of 543 cysts on 221 abdominal CTs were manually segmented in 3D by a radiology resident in consensus with a board-certified radiologist specialized in abdominal radiology. This information was used to train a two-step nnU-Net for detection with the performance assessed depending on lesions' volume and location in comparison to three human readers of varying experience. Mean sensitivity was 78.8 ± 0.1%. The sensitivity was highest for large lesions with 87.8% for cysts ≥220 mm3 and for lesions in the distal pancreas with up to 96.2%. The number of false-positive detections for cysts ≥220 mm3 was 0.1 per case. The algorithm's performance was comparable to human readers. To conclude, automated detection of PCL on CTs is feasible. The proposed model could serve radiologists as a second reading tool. All imaging data and code used in this study are freely available online.

Revealing the most common reporting errors through data mining of the report proofreading process.

OBJECTIVES: To investigate the most common errors in residents' preliminary reports, if structured reporting impacts error types and frequencies, and to identify possible implications for resident education and patient safety. MATERIAL AND METHODS: Changes in report content were tracked by a report comparison tool on a word level and extracted for 78,625 radiology reports dictated from September 2017 to December 2018 in our department. Following data aggregation according to word stems and stratification by subspecialty (e.g., neuroradiology) and imaging modality, frequencies of additions/deletions were analyzed for findings and impression report section separately and compared between subgroups. RESULTS: Overall modifications per report averaged 4.1 words, with demonstrably higher amounts of changes for cross-sectional imaging (CT: 6.4; MRI: 6.7) than non-cross-sectional imaging (radiographs: 0.2; ultrasound: 2.8). The four most frequently changed words (right, left, one, and none) remained almost similar among all subgroups (range: 0.072-0.117 per report; once every 9-14 reports). Albeit representing only 0.02% of analyzed words, they accounted for up to 9.7% of all observed changes. Subspecialties solely using structured reporting had substantially lower change ratios in the findings report section (mean: 0.2 per report) compared with prose-style reporting subspecialties (mean: 2.0). Relative frequencies of the most changed words remained unchanged. CONCLUSION: Residents' most common reporting errors in all subspecialties and modalities are laterality discriminator confusions (left/right) and unnoticed descriptor misregistration by speech recognition (one/none). Structured reporting reduces overall error rates, but does not affect occurrence of the most common errors. Increased error awareness and measures improving report correctness and ensuring patient safety are required. KEY POINTS: • The two most common reporting errors in residents' preliminary reports are laterality discriminator confusions (left/right) and unnoticed descriptor misregistration by speech recognition (one/none). • Structured reporting reduces the overall the error frequency in the findings report section by a factor of 10 (structured reporting: mean 0.2 per report; prose-style reporting: 2.0) but does not affect the occurrence of the two major errors. • Staff radiologist review behavior noticeably differs between radiology subspecialties.

Unusual Presentation of Patent Ductus Arteriosus in Elderly Patient.

We presented a case of a 56-year-old man with giant pulmonary artery aneurysm caused by a misdiagnosed patent ductus arteriosus, severe multivalvular disease and active aortic valve endocarditis successfully treated by surgery. The correct diagnosis was missed despite preoperative diagnostics because the small patent ductus arteriosus was located at the distal part of common pulmonary trunk and a huge regurgitant signal overlapped its Doppler signal. Thorough evaluation of every patient, regardless of age, is necessary to recognize and treat this congenital anomaly.

Unusual presentation of patent ductus arteriosus in elderly patient

Abstract We presented a case of a 56-year-old man with giant pulmonary artery aneurysm caused by a misdiagnosed patent ductus arteriosus, severe multivalvular disease and active aortic valve endocarditis successfully treated by surgery. The correct diagnosis was missed despite preoperative diagnostics because the small patent ductus arteriosus was located at the distal part of common pulmonary trunk and a huge regurgitant signal overlapped its Doppler signal. Thorough evaluation of every patient, regardless of age, is necessary to recognize and treat this congenital anomaly.

Development and clinical implementation of tailored image analysis tools for COVID-19 in the midst of the pandemic: The synergetic effect of an open, clinically embedded software development platform and machine learning.

PURPOSE: During the emerging COVID-19 pandemic, radiology departments faced a substantial increase in chest CT admissions coupled with the novel demand for quantification of pulmonary opacities. This article describes how our clinic implemented an automated software solution for this purpose into an established software platform in 10 days. The underlying hypothesis was that modern academic centers in radiology are capable of developing and implementing such tools by their own efforts and fast enough to meet the rapidly increasing clinical needs in the wake of a pandemic. METHOD: Deep convolutional neural network algorithms for lung segmentation and opacity quantification on chest CTs were trained using semi-automatically and manually created ground-truth (Ntotal = 172). The performance of the in-house method was compared to an externally developed algorithm on a separate test subset (N = 66). RESULTS: The final algorithm was available at day 10 and achieved human-like performance (Dice coefficient = 0.97). For opacity quantification, a slight underestimation was seen both for the in-house (1.8 %) and for the external algorithm (0.9 %). In contrast to the external reference, the underestimation for the in-house algorithm showed no dependency on total opacity load, making it more suitable for follow-up. CONCLUSIONS: The combination of machine learning and a clinically embedded software development platform enabled time-efficient development, instant deployment, and rapid adoption in clinical routine. The algorithm for fully automated lung segmentation and opacity quantification that we developed in the midst of the COVID-19 pandemic was ready for clinical use within just 10 days and achieved human-level performance even in complex cases.

Risk factors for neurological dysfunctions after surgical repair of acute aortic dissection type A.

INTRODUCTION: Technical improvement and new operative strategies significantly influence survival and outcomes after the treatment of acute aortic dissection type A (AADA). However, postoperative complications and particularly neurological dysfunctions (ND) are still very common. AIM: To identify preoperative and intraoperative factors as well as immediate postoperative conditions with an influence on the occurrence of neurological complications of surgical treatment of AADA and accordingly take action to reduce them. MATERIAL AND METHODS: Between January 2013 and December 2018, 240 patients with AADA were emergently surgically treated. All patients were divided into two groups: group I - patients with postoperative ND (subgroup Ia - patients with mild, transient ND and Ib - patients with severe ND) and group II - patients without ND. RESULTS: Neurological damage after the operation was registered in 87 (39.5%) patients. Thirty (13.6%) patients had mild ND and 57 (25.9%) severe. Presence of preoperative neurological deficit, reduced level of consciousness, supra-aortic vessel dissection, hemodynamic instability, and excessive postoperative bleeding with hypotension are factors with a highly statistically significant association with the occurrence of severe ND. Neurological complications were not identified in 66.7% of patients who were axillary cannulated versus 55.9% of patients cannulated in the other way but the difference did not reach statistical significance (p = 0.1099). CONCLUSIONS: Advanced neuroprotective strategies during surgical treatment of AADA are associated with favorable neurological outcomes, especially in a group of patients with identified risk factors for ND.

Towards automated generation of curated datasets in radiology: Application of natural language processing to unstructured reports exemplified on CT for pulmonary embolism.

PURPOSE: To design and evaluate a self-trainable natural language processing (NLP)-based procedure to classify unstructured radiology reports. The method enabling the generation of curated datasets is exemplified on CT pulmonary angiogram (CTPA) reports. METHOD: We extracted the impressions of CTPA reports created at our institution from 2016 to 2018 (n = 4397; language: German). The status (pulmonary embolism: yes/no) was manually labelled for all exams. Data from 2016/2017 (n = 2801) served as a ground truth to train three NLP architectures that only require a subset of reference datasets for training to be operative. The three architectures were as follows: a convolutional neural network (CNN), a support vector machine (SVM) and a random forest (RF) classifier. Impressions of 2018 (n = 1377) were kept aside and used for general performance measurements. Furthermore, we investigated the dependence of classification performance on the amount of training data with multiple simulations. RESULTS: The classification performance of all three models was excellent (accuracies: 97 %-99 %; F1 scores 0.88-0.97; AUCs: 0.993-0.997). Highest accuracy was reached by the CNN with 99.1 % (95 % CI 98.5-99.6 %). Training with 470 labelled impressions was sufficient to reach an accuracy of > 93 % with all three NLP architectures. CONCLUSION: Our NLP-based approaches allow for an automated and highly accurate retrospective classification of CTPA reports with manageable effort solely using unstructured impression sections. We demonstrated that this approach is useful for the classification of radiology reports not written in English. Moreover, excellent classification performance is achieved at relatively small training set sizes.

A Practical Guide to Artificial Intelligence-Based Image Analysis in Radiology.

The use of artificial intelligence (AI) is a powerful tool for image analysis that is increasingly being evaluated by radiology professionals. However, due to the fact that these methods have been developed for the analysis of nonmedical image data and data structure in radiology departments is not "AI ready", implementing AI in radiology is not straightforward. The purpose of this review is to guide the reader through the pipeline of an AI project for automated image analysis in radiology and thereby encourage its implementation in radiology departments. At the same time, this review aims to enable readers to critically appraise articles on AI-based software in radiology.