Search

The coronavirus disease 2019 (COVID-19) pandemic has revealed that even the best-resourced hospitals may lack sufficient ventilators to support patients under surge conditions. During a pandemic or mass trauma, an affordable, low-maintenance, off-the-shelf device that would allow health care teams to rapidly expand their ventilator capacity could prove lifesaving, but only if it can be safely integrated into a complex and rapidly changing clinical environment. Here, we define an approach to safe ventilator sharing that prioritizes predictable and independent care of patients sharing a ventilator. Subsequently, we detail the design and testing of a ventilator-splitting circuit that follows this approach and describe our clinical experience with this circuit during the COVID-19 pandemic. This circuit was able to provide individualized and titratable ventilatory support with individualized positive end-expiratory pressure (PEEP) to 2 critically ill patients at the same time, while insulating each patient from changes in the other's condition. We share insights from our experience using this technology in the intensive care unit and outline recommendations for future clinical applications.

Recent advances in artificial intelligence (AI) are providing an opportunity to enhance existing clinical decision support (CDS) tools to improve patient safety and drive value-based imaging. We discuss the advantages and potential applications that may be realized with the synergy between AI and CDS systems. From the perspective of both radiologist and ordering provider, CDS could be significantly empowered using AI. CDS enhanced by AI could reduce friction in radiology workflows and can aid AI developers to identify relevant imaging features their tools should be seeking to extract from images. Furthermore, these systems can generate structured data to be used as input to develop machine learning algorithms, which can drive downstream care pathways. For referring providers, an AI-enabled CDS solution could enable an evolution from existing imaging-centric CDS toward decision support that takes into account a holistic patient perspective. More intelligent CDS could suggest imaging examinations in highly complex clinical scenarios, assist on the identification of appropriate imaging opportunities at the health system level, suggest appropriate individualized screening, or aid health care providers to ensure continuity of care. AI has the potential to enable the next generation of CDS, improving patient care and enhancing providers' and radiologists' experience.

The growing interest in machine learning (ML) in healthcare is driven by the promise of improved patient care. However, how many ML algorithms are currently being used in clinical practice? While the technology is present, as demonstrated in a variety of commercial products, clinical integration is hampered by a lack of infrastructure, processes, and tools. In particular, automating the selection of relevant series for a particular algorithm remains challenging. In this work, we propose a methodology to automate the identification of brain MRI sequences so that we can automatically route the relevant inputs for further image-related algorithms. The method relies on metadata required by the Digital Imaging and Communications in Medicine (DICOM) standard, resulting in generalizability and high efficiency (less than 0.4 ms/series). To support our claims, we test our approach on two large brain MRI datasets (40,000 studies in total) from two different institutions on two different continents. We demonstrate high levels of accuracy (ranging from 97.4 to 99.96%) and generalizability across the institutions. Given the complexity and variability of brain MRI protocols, we are confident that similar techniques could be applied to other forms of radiological imaging.

Recent advances and future perspectives of machine learning techniques offer promising applications in medical imaging. Machine learning has the potential to improve different steps of the radiology workflow including order scheduling and triage, clinical decision support systems, detection and interpretation of findings, postprocessing and dose estimation, examination quality control, and radiology reporting. In this article, the authors review examples of current applications of machine learning and artificial intelligence techniques in diagnostic radiology. In addition, the future impact and natural extension of these techniques in radiology practice are discussed. © RSNA, 2018

Three-dimensional (3D) printing has recently erupted into the medical arena due to decreased costs and increased availability of printers and software tools. Due to lack of detailed information in the medical literature on the methods for 3D printing, we have reviewed the medical and engineering literature on the various methods for 3D printing and compiled them into a practical "how to" format, thereby enabling the novice to start 3D printing with very limited funds. We describe (1) background knowledge, (2) imaging parameters, (3) software, (4) hardware, (5) post-processing, and (6) financial aspects required to cost-effectively reproduce a patient's disease ex vivo so that the patient, engineer and surgeon may hold the anatomy and associated pathology in their hands.

Purpose To evaluate the effect of tomosynthesis in diagnostic mammography on the Breast Imaging Reporting and Data System (BI-RADS) final assessment categories over time. Materials and Methods This retrospective study was approved by the institutional review board. The authors reviewed all diagnostic mammograms obtained during a 12-month interval before (two-dimensional [2D] mammography [June 2, 2010, to June 1, 2011]) and for 3 consecutive years after (tomosynthesis year 1 [2012], tomosynthesis year 2 [2013], and tomosynthesis year 3 [2014]) the implementation of tomosynthesis. The requirement to obtain informed consent was waived. The rates of BI-RADS final assessment categories 1-5 were compared between the 2D and tomosynthesis groups. The positive predictive values after biopsy (PPV3) for BI-RADS category 4 and 5 cases were compared. The mammographic features (masses, architectural distortions, calcifications, focal asymmetries) of lesions categorized as probably benign (BI-RADS category 3) and those for which biopsy was recommended (BI-RADS category 4 or 5) were reviewed. The χ(2) test was used to compare the rates of BI-RADS final assessment categories 1-5 between the two groups, and multivariate logistic regression analysis was performed to compare all diagnostic studies categorized as BI-RADS 3-5. Results There was an increase in the percentage of cases reported as negative or benign (BI-RADS category 1 or 2) with tomosynthesis (58.7% with 2D mammography vs 75.8% with tomosynthesis at year 3, P < .0001). A reduction in the percentage of probably benign (BI-RADS category 3) final assessments also occurred (33.3% with 2D mammography vs 16.4% with tomosynthesis at year 3, P < .0001). Although the rates of BI-RADS 4 or 5 assessments did not change significantly with tomosynthesis (8.0% with 2D mammography vs 7.8% with tomosynthesis at year 3, P = .2), there was a significant increase in the PPV3 (29.6% vs 50%, respectively; P < .0001). These trends increased during the 3 years of tomosynthesis use. Conclusion Tomosynthesis in the diagnostic setting resulted in progressive shifts in the BI-RADS final assessment categories over time, with a significant increase in the proportion of studies classified as normal, a continued decrease in the rate of studies categorized as probably benign, and improved diagnostic confidence in biopsy recommendations. (©) RSNA, 2016.

The ability of humans to distinguish the delicate differences in food flavors depends mostly on retronasal smell, in which food volatiles entrained into the airway at the back of the oral cavity are transported by exhaled air through the nasal cavity to stimulate the olfactory receptor neurons. Little is known whether food volatiles are preferentially carried by retronasal flow toward the nasal cavity rather than by orthonasal flow into the lung. To study the differences between retronasal and orthonasal flow, we obtained computed tomography (CT) images of the orthonasal airway from a healthy human subject, printed an experimental model using a 3D printer, and analyzed the flow field inside the airway. The results show that, during inhalation, the anatomical structure of the oropharynx creates an air curtain outside a virtual cavity connecting the oropharynx and the back of the mouth, which prevents food volatiles from being transported into the main stream toward the lung. In contrast, during exhalation, the flow preferentially sweeps through this virtual cavity and effectively enhances the entrainment of food volatiles into the main retronasal flow. This asymmetrical transport efficiency is also found to have a nonmonotonic Reynolds number dependence: The asymmetry peaks at a range of an intermediate Reynolds number close to 800, because the air curtain effect during inhalation becomes strongest in this range. This study provides the first experimental evidence, to our knowledge, for adaptations of the geometry of the human oropharynx for efficient transport of food volatiles toward the olfactory receptors in the nasal cavity.

The success of cancer therapy can be difficult to predict, as its efficacy is often predicated upon characteristics of the cancer, treatment, and individual that are not fully understood or are difficult to ascertain. Monitoring the response of disease to treatment is therefore essential and has traditionally been characterized by changes in tumor volume. However, in many instances, this singular measure is insufficient for predicting treatment effects on patient survival. Molecular imaging allows repeated in vivo measurement of many critical molecular features of neoplasm, such as metabolism, proliferation, angiogenesis, hypoxia, and apoptosis, which can be employed for monitoring therapeutic response. In this review, we examine the current methods for evaluating response to treatment and provide an overview of emerging PET molecular imaging methods that will help guide future cancer therapies.

Infectious diseases need a globally coordinated genomic-based movement linking sequencing efforts to development of response tools to mitigate the impact of existing and emerging threats., Author Summary The world of genomics is transforming medicine, and is likely to influence the future development of new drugs, diagnostics, and vaccines. To date, the greater focus of genomics and medicine has been on conditions affecting resourcewealthy settings, primarily involving scientists and companies in those settings. However, we believe that it is possible to expand genomics into a more global technology that can also focus on diseases of resource-limited settings. This goal can be achieved if genomics is made a global priority. We feel one way to move in this direction is through a comprehensive approach to infectious diseases—i.e., an Infectious Disease Genomics Project—that would mirror the Human Genome Project. Without an active, unified effort specifically focused on allowing actors at any level to participate in the genomics revolution, infectious diseases that primarily affect the poor will likely not achieve the same level of scientifici advancement as diseases affecting the wealthy.