Publications by authors named "R W Tobin"

229 Publications

The reporting quality of natural language processing studies: systematic review of studies of radiology reports.

BMC Med Imaging 2021 10 2;21(1):142. Epub 2021 Oct 2.

Centre for Clinical Brain Sciences, University of Edinburgh, Chancellor's Building, Little France, Edinburgh, EH16 4TJ, Scotland, UK.

Background: Automated language analysis of radiology reports using natural language processing (NLP) can provide valuable information on patients' health and disease. With its rapid development, NLP studies should have transparent methodology to allow comparison of approaches and reproducibility. This systematic review aims to summarise the characteristics and reporting quality of studies applying NLP to radiology reports.

Methods: We searched Google Scholar for studies published in English that applied NLP to radiology reports of any imaging modality between January 2015 and October 2019. At least two reviewers independently performed screening and completed data extraction. We specified 15 criteria relating to data source, datasets, ground truth, outcomes, and reproducibility for quality assessment. The primary NLP performance measures were precision, recall and F1 score.

Results: Of the 4,836 records retrieved, we included 164 studies that used NLP on radiology reports. The commonest clinical applications of NLP were disease information or classification (28%) and diagnostic surveillance (27.4%). Most studies used English radiology reports (86%). Reports from mixed imaging modalities were used in 28% of the studies. Oncology (24%) was the most frequent disease area. Most studies had dataset size > 200 (85.4%) but the proportion of studies that described their annotated, training, validation, and test set were 67.1%, 63.4%, 45.7%, and 67.7% respectively. About half of the studies reported precision (48.8%) and recall (53.7%). Few studies reported external validation performed (10.8%), data availability (8.5%) and code availability (9.1%). There was no pattern of performance associated with the overall reporting quality.

Conclusions: There is a range of potential clinical applications for NLP of radiology reports in health services and research. However, we found suboptimal reporting quality that precludes comparison, reproducibility, and replication. Our results support the need for development of reporting standards specific to clinical NLP studies.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1186/s12880-021-00671-8DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8487512PMC
October 2021

Clinical outcomes in patients co-infected with COVID-19 and Staphylococcus aureus: a scoping review.

BMC Infect Dis 2021 Sep 21;21(1):985. Epub 2021 Sep 21.

Morristown Medical Center of Atlantic Health System, Morristown, New Jersey, USA.

Background: Endemic to the hospital environment, Staphylococcus aureus (S. aureus) is a leading bacterial pathogen that causes deadly infections such as bacteremia and endocarditis. In past viral pandemics, it has been the principal cause of secondary bacterial infections, significantly increasing patient mortality rates. Our world now combats the rapid spread of COVID-19, leading to a pandemic with a death toll greatly surpassing those of many past pandemics. However, the impact of co-infection with S. aureus remains unclear. Therefore, we aimed to perform a high-quality scoping review of the literature to synthesize the existing evidence on the clinical outcomes of COVID-19 and S. aureus co-infection.

Methods: A scoping review of the literature was conducted in PubMed, Scopus, Ovid MEDLINE, CINAHL, ScienceDirect, medRxiv, and the WHO COVID-19 database using a combination of terms. Articles that were in English, included patients infected with both COVID-19 and S. aureus, and provided a description of clinical outcomes for patients were eligible. From these articles, the following data were extracted: type of staphylococcal species, onset of co-infection, patient sex, age, symptoms, hospital interventions, and clinical outcomes. Quality assessments of final studies were also conducted using the Joanna Briggs Institute's critical appraisal tools.

Results: Searches generated a total of 1922 publications, and 28 articles were eligible for the final analysis. Of the 115 co-infected patients, there were a total of 71 deaths (61.7%) and 41 discharges (35.7%), with 62 patients (53.9%) requiring ICU admission. Patients were infected with methicillin-sensitive and methicillin-resistant strains of S. aureus, with the majority (76.5%) acquiring co-infection with S. aureus following hospital admission for COVID-19. Aside from antibiotics, the most commonly reported hospital interventions were intubation with mechanical ventilation (74.8 %), central venous catheter (19.1 %), and corticosteroids (13.0 %).

Conclusions: Given the mortality rates reported thus far for patients co-infected with S. aureus and COVID-19, COVID-19 vaccination and outpatient treatment may be key initiatives for reducing hospital admission and S. aureus co-infection risk. Physician vigilance is recommended during COVID-19 interventions that may increase the risk of bacterial co-infection with pathogens, such as S. aureus, as the medical community's understanding of these infection processes continues to evolve.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1186/s12879-021-06616-4DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8453255PMC
September 2021

Melanoma Metastases to the Adrenal Gland Are Highly Resistant to Immune Checkpoint Inhibitors.

J Natl Compr Canc Netw 2021 Aug 4. Epub 2021 Aug 4.

3Center for Rare Melanomas.

Background: Adrenal gland metastases (AGMs) are common in advanced-stage melanoma, occurring in up to 50% of patients. The introduction of immune checkpoint inhibitors (ICIs) has markedly altered the outcome of patients with melanoma. However, despite significant successes, anecdotal evidence has suggested that treatment responses in AGMs are significantly lower than in other metastatic sites. We sought to investigate whether having an AGM is associated with altered outcomes and whether ICI responses are dampened in the adrenal glands.

Patients And Methods: We retrospectively compared ICI responses and overall survival (OS) in 68 patients with melanoma who were diagnosed with an AGM and a control group of 100 patients without AGMs at a single institution. Response was determined using RECIST 1.1. OS was calculated from time of ICI initiation, anti-PD-1 initiation, initial melanoma diagnosis, and stage IV disease diagnosis. Tumor-infiltrating immune cells were characterized in 9 resected AGMs using immunohistochemical analysis.

Results: Response rates of AGMs were significantly lower compared with other metastatic sites in patients with AGMs (16% vs 22%) and compared with those without AGMs (55%). Patients with AGMs also had significantly lower median OS compared with those without AGMs (3.1 years vs not reached, respectively). We further observed that despite this, AGMs exhibited high levels of tumor-infiltrating immune cells.

Conclusions: In this cohort of patients with melanoma, those diagnosed with an AGM had lower ICI response rates and OS. These results suggest that tissue-specific microenvironments of AGMs present unique challenges that may require novel, adrenal gland-directed therapies or surgical resection.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.6004/jnccn.2020.7800DOI Listing
August 2021

A systematic review of natural language processing applied to radiology reports.

BMC Med Inform Decis Mak 2021 06 3;21(1):179. Epub 2021 Jun 3.

School of Literatures, Languages and Cultures (LLC), University of Edinburgh, Edinburgh, Scotland.

Background: Natural language processing (NLP) has a significant role in advancing healthcare and has been found to be key in extracting structured information from radiology reports. Understanding recent developments in NLP application to radiology is of significance but recent reviews on this are limited. This study systematically assesses and quantifies recent literature in NLP applied to radiology reports.

Methods: We conduct an automated literature search yielding 4836 results using automated filtering, metadata enriching steps and citation search combined with manual review. Our analysis is based on 21 variables including radiology characteristics, NLP methodology, performance, study, and clinical application characteristics.

Results: We present a comprehensive analysis of the 164 publications retrieved with publications in 2019 almost triple those in 2015. Each publication is categorised into one of 6 clinical application categories. Deep learning use increases in the period but conventional machine learning approaches are still prevalent. Deep learning remains challenged when data is scarce and there is little evidence of adoption into clinical practice. Despite 17% of studies reporting greater than 0.85 F1 scores, it is hard to comparatively evaluate these approaches given that most of them use different datasets. Only 14 studies made their data and 15 their code available with 10 externally validating results.

Conclusions: Automated understanding of clinical narratives of the radiology reports has the potential to enhance the healthcare process and we show that research in this field continues to grow. Reproducibility and explainability of models are important if the domain is to move applications into clinical use. More could be done to share code enabling validation of methods on different institutional data and to reduce heterogeneity in reporting of study properties allowing inter-study comparisons. Our results have significance for researchers in the field providing a systematic synthesis of existing work to build on, identify gaps, opportunities for collaboration and avoid duplication.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1186/s12911-021-01533-7DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8176715PMC
June 2021

Robust real-time 3D imaging of moving scenes through atmospheric obscurant using single-photon LiDAR.

Sci Rep 2021 May 27;11(1):11236. Epub 2021 May 27.

School of Engineering and Physical Sciences, Heriot-Watt University, Edinburgh, EH14 4AS, UK.

Recently, time-of-flight LiDAR using the single-photon detection approach has emerged as a potential solution for three-dimensional imaging in challenging measurement scenarios, such as over distances of many kilometres. The high sensitivity and picosecond timing resolution afforded by single-photon detection offers high-resolution depth profiling of remote, complex scenes while maintaining low power optical illumination. These properties are ideal for imaging in highly scattering environments such as through atmospheric obscurants, for example fog and smoke. In this paper we present the reconstruction of depth profiles of moving objects through high levels of obscurant equivalent to five attenuation lengths between transceiver and target at stand-off distances up to 150 m. We used a robust statistically based processing algorithm designed for the real time reconstruction of single-photon data obtained in the presence of atmospheric obscurant, including providing uncertainty estimates in the depth reconstruction. This demonstration of real-time 3D reconstruction of moving scenes points a way forward for high-resolution imaging from mobile platforms in degraded visual environments.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1038/s41598-021-90587-8DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8159934PMC
May 2021
-->