Objectives The purpose of this study was to build a deep learning model to derive labels from neuroradiology reports and assign these to the corresponding examinations, overcoming a bottleneck to computer vision model development. Methods Reference-standard labels were generated by a team of neuroradiologists for model training and evaluation. Three thousand examinations were labelled for the presence or absence of any abnormality by manually scrutinising the corresponding radiology reports (‘reference-standard report labels’); a subset of these examinations (n = 250) were assigned ‘reference-standard image labels’ by interrogating the actual images. Separately, 2000 reports were labelled for the presence or absence of 7 specialised categories of abnormality (acute stroke, mass, atrophy, vascular abnormality, small vessel disease, white matter inflammation, encephalomalacia), with a subset of these examinations (n = 700) also assigned reference-standard image labels. A deep learning model was trained using labelled reports and validated in two ways: comparing predicted labels to (i) reference-standard report labels and (ii) reference-standard image labels. The area under the receiver operating characteristic curve (AUC-ROC) was used to quantify model performance. Accuracy, sensitivity, specificity, and F1 score were also calculated. Results Accurate classification (AUC-ROC > 0.95) was achieved for all categories when tested against reference-standard report labels. A drop in performance (ΔAUC-ROC > 0.02) was seen for three categories (atrophy, encephalomalacia, vascular) when tested against reference-standard image labels, highlighting discrepancies in the original reports. Once trained, the model assigned labels to 121,556 examinations in under 30 min. Conclusions Our model accurately classifies head MRI examinations, enabling automated dataset labelling for downstream computer vision applications. Key Points • Deep learning is poised to revolutionise image recognition tasks in radiology; however, a barrier to clinical adoption is the difficulty of obtaining large labelled datasets for model training. • We demonstrate a deep learning model which can derive labels from neuroradiology reports and assign these to the corresponding examinations at scale, facilitating the development of downstream computer vision models. • We rigorously tested our model by comparing labels predicted on the basis of neuroradiology reports with two sets of reference-standard labels: (1) labels derived by manually scrutinising each radiology report and (2) labels derived by interrogating the actual images.
Purpose There is some evidence suggesting a different nature of response to selective laser trabeculoplasty (SLT) among different races. Therefore, we aimed to assess the short-term efficacy, safety and nature of outcome of SLT in Omani eyes. Patients and Methods A retrospective review was performed of patients with open-angle glaucoma (OAG) or ocular hypertension (OHTN) who underwent a single session of 360-degree SLT between January 1, 2017 and December 31, 2018. The main outcome was mean IOP reduction and attainment of treatment success at 5 weeks and 12 weeks post treatment defined as at least 20% IOP reduction from baseline without further medications or interventions. Secondary outcomes were frequency of adverse events and factors predicting success. Results A total of 33 eyes of 33 Omani patients who underwent treatment with SLT were analyzed. The nature of response to laser followed a gradual pattern as the mean IOP reduction from baseline was 20.2% (5.21 mm Hg, P <0.001) at 5 weeks and further enhanced to 27.2% (6.95 mm Hg, P <0.001) at 12 weeks. Short-term success was achieved in 51.5% and 72.2% of eyes at 5 and 12 weeks, respectively. SLT was most effective in OHTN subgroup and those with higher baseline IOP (both P <0.001). Side effects were an infrequent occurrence, minor and transient. Conclusion The short-term success of SLT in Omani eyes was clinically relevant and comparable to the gradual pattern seen in patients of Indian ancestry. It is a safe therapeutic option in selective Omani eyes.
PURPOSE: To investigate the impact of coronavirus infection disease-19 (COVID-19) pandemic on ophthalmic referrals within an academic tertiary center in Oman. METHODS: Retrospective chart review of internal referrals received and evaluated by the ophthalmology department between March 1and August 31, 2020 (COVID-19 period) compared to a corresponding period in 2019 (pre COVID-19). Data included patient demographics, referral details, ocular diagnosis, intervention, and discharge plan. RESULTS: Referral volume significantly decreased by 58.2%; from 2019 prepandemic to 510 ( P = 0.001), with the lowest in April and May 2020. Patient demographics did not differ significantly, but “urgent” referrals reduced by 96.2% ( P < 0.001). Main reasons for referrals were reduced vision and screening in both periods. During pandemic, referrals for screening purposes increased from 30.3% to 37.9% ( P = 0.013) and for reduced vision decreased from 30% to 23.3% ( P = 0.021). Dry eye syndrome increased in frequency during 2020 (from 2.9% to 7.3%, P = 0.002) but cataracts and conjunctivitis both decreased (from 4.7% to 2.1%, P = 0.046 and from 2.3% to 0.3%, P = 0.013, respectively). Ocular trauma remained stable (from 0.8% to 0.3%, P = 0.456), but the proportion of chemical injuries increased by 13.7% ( P = 0.025). There was a drastic decrease in interventions from 37% to 26.1% ( P < 0.001) and an increase in discharge rate from 61.2% to 75.8% ( P < 0.001). CONCLUSION: The impact of COVID-19 pandemic on ophthalmic referrals within a tertiary academic centre in oman referral reductions and changes in pattern and characteristics as an epiphenomenon of COVID-19 reflect the extent of impact specifically in an Omani context. This information is vital for planning proper resource utilization, the adoption of innovative care delivery, and improving referral system pathways.
Labelling large datasets for training high-capacity neural networks is a major obstacle to the development of deep learning-based medical imaging applications. Here we present a transformer-based network for magnetic resonance imaging (MRI) radiology report classification which automates this task by assigning image labels on the basis of free-text expert radiology reports. Our model's performance is comparable to that of an expert radiologist, and better than that of an expert physician, demonstrating the feasibility of this approach. We make code available online for researchers to label their own MRI datasets for medical imaging applications.
Natural language processing (NLP) shows promise as a means to automate the labelling of hospital-scale neuroradiology magnetic resonance imaging (MRI) datasets for computer vision applications. To date, however, there has been no thorough investigation into the validity of this approach, including determining the accuracy of report labels compared to image labels as well as examining the performance of non-specialist labellers. In this work, we draw on the experience of a team of neuroradiologists who labelled over 5000 MRI neuroradiology reports as part of a project to build a dedicated deep learning-based neuroradiology report classifier. We show that, in our experience, assigning binary labels (i.e. normal vs abnormal) to images from reports alone is highly accurate. In contrast to the binary labels, however, the accuracy of more granular labelling is dependent on the category, and we highlight reasons for this discrepancy. We also show that downstream model performance is reduced when labelling of training reports is performed by a non-specialist. To allow other researchers to accelerate their research, we make our refined abnormality definitions and labelling rules available, as well as our easy-to-use radiology report labelling app which helps streamline this process.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.