Machine learning methods offer great promise for fast and accurate detection and prognostication of coronavirus disease 2019 (COVID-19) from standard-of-care chest radiographs (CXR) and chest computed tomography (CT) images. Many articles have been published in 2020 describing new machine learning-based models for both of these tasks, but it is unclear which are of potential clinical utility. In this systematic review, we consider all published papers and preprints, for the period from 1 January 2020 to 3 October 2020, which describe new machine learning models for the diagnosis or prognosis of COVID-19 from CXR or CT images. All manuscripts uploaded to bioRxiv, medRxiv and arXiv along with all entries in EMBASE and MEDLINE in this timeframe are considered. Our search identified 2,212 studies, of which 415 were included after initial screening and, after quality screening, 62 studies were included in this systematic review. Our review finds that none of the models identified are of potential clinical use due to methodological flaws and/or underlying biases. This is a major weakness, given the urgency with which validated COVID-19 models are needed. To address this, we give many recommendations which, if followed, will solve these issues and lead to higher-quality model development and well-documented manuscripts.
The Canadian Multiple Sclerosis Working Group (CMSWG) developed practical recommendations in 2004 to assist clinicians in optimizing the use of disease-modifying therapies (DMT) in patients with relapsing multiple sclerosis. The CMSWG convened to review how disease activity is assessed, propose a more current approach for assessing suboptimal response, and to suggest a scheme for switching or escalating treatment. Practical criteria for relapses, Expanded Disability Status Scale (EDSS) progression and MRI were developed to classify the clinical level of concern as Low, Medium and High. The group concluded that a change in treatment may be considered in any RRMS patient if there is a high level of concern in any one domain (relapses, progression or MRI), a medium level of concern in any two domains, or a low level of concern in all three domains. These recommendations for assessing treatment response should assist clinicians in making more rational choices in their management of relapsing MS patients.
The risk of conversion from a clinically isolated syndrome to multiple sclerosis was significantly lower with minocycline than with placebo over 6 months but not over 24 months. (Funded by the Multiple Sclerosis Society of Canada; ClinicalTrials.gov number, NCT00666887 .).
Automatic segmentation methods are an important advancement in medical image analysis. Machine learning techniques, and deep neural networks in particular, are the state-of-the-art for most medical image segmentation tasks. Issues with class imbalance pose a significant challenge in medical datasets, with lesions often occupying a considerably smaller volume relative to the background. Loss functions used in the training of deep learning algorithms differ in their robustness to class imbalance, with direct consequences for model convergence. The most commonly used loss functions for segmentation are based on either the cross entropy loss, Dice loss or a combination of the two. We propose the Unified Focal loss, a new hierarchical framework that generalises Dice and cross entropy-based losses for handling class imbalance. We evaluate our proposed loss function on five publicly available, class imbalanced medical imaging datasets: CVC-ClinicDB, Digital Retinal Images for Vessel Extraction (DRIVE), Breast Ultrasound 2017 (BUS2017), Brain Tumour Segmentation 2020 (BraTS20) and Kidney Tumour Segmentation 2019 (KiTS19). We compare our loss function performance against six Dice or cross entropy-based loss functions, across 2D binary, 3D binary and 3D multiclass segmentation tasks, demonstrating that our proposed loss function is robust to class imbalance and consistently outperforms the other loss functions. Source code is available at: https://github.com/mlyg/unified-focal-loss .
We report a trial of minocycline in people with relapsingremitting multiple sclerosis (RRMS) that evaluates safety and estimates its effect on magnetic resonance imaging (MRI). Ten subjects with active RRMS received oral minocycline 100mg twice daily for 6 months after a 3-month run-in period. A 30-month treatment extension is ongoing. Clinical and laboratory assessments were completed at enrollment and then at 3-month intervals. MRI was performed at enrolment and then every 4 weeks. Patients without MRI activity during the run-in phase continued in the study, including completion of all MRI scans, to confirm lack of MRI worsening. The primary outcome was change in the mean number of gadolinium-enhancing lesions per scan during the first 6 months of treatment compared with the run-in period (Wilcoxon signed rank test, two-sided alpha of 0.05).Eighty percent of participants were women. Mean age was 42.8 years (SD 4.0). Mean MS duration was 11.8 years (SD 6.3). Median baseline extended disability status score (EDSS) was 2.5 (range 1.5-5.5). Mean relapse number in the two prior years was 2.6 (range 2-4). During the trial, there were no serious adverse events or laboratory abnormalities and no change in EDSS. Three relapses occurred during the run-in phase, five during the first 6-month treatment phase, and none during the following 6 months. On-treatment relapses included one associated with MRI enhancement (during month 1), two without enhancement (one scan was a postrelapse scan, and one scan was missed because the patient was taking steroids), and two mild truncal sensory attacks unassociated with MRI enhancement (both at 5 months).Mean total enhancing lesion number decreased from 1.38 lesions per scan during the run-in phase to 0.22 during the treatment phase (z ϭ 2.204, p ϭ 0.0276), representing a relative reduction of greater than 84%. During the run-in phase, 47.5% of MRI scans (19/40) were active, whereas 9.3% (5/54) were active during the minocycline phase. There were no active scans after month 2 (Fig) and no new active lesions after month 1. Although five patients accounted for all MRI activity before and after treatment, all patient data were included in all analyses.This study provides preliminary evidence that minocycline may be useful in MS and supports its safety. The MRI results are consistent with the ability of minocycline to inhibit matrix metalloproteinases, 1,2 thus reducing lymphocyte access to the central nervous system. In addition, minocycline may have other beneficial properties including neuroprotection.3 Small sample size and short trial duration limit conclusions, but reduced MRI activity is encouraging and calls for definitive studies to establish minocycline efficacy in MS. Departments of
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.