A clinically applicable approach to continuous prediction of future acute kidney injury.

The early prediction of deterioration could have an important role in supporting healthcare professionals, as an estimated 11% of deaths in hospital follow a failure to promptly recognize and treat deteriorating patients1. To achieve this goal requires predictions of patient risk that are continuously updated and accurate, and delivered at an individual level with sufficient context and enough time to act. Here we develop a deep learning approach for the continuous risk prediction of future deterioration in patients, building on recent work that models adverse events from electronic health records2-17 and using acute kidney injury-a common and potentially life-threatening condition18-as an exemplar. Our model was developed on a large, longitudinal dataset of electronic health records that cover diverse clinical environments, comprising 703,782 adult patients across 172 inpatient and 1,062 outpatient sites. Our model predicts 55.8% of all inpatient episodes of acute kidney injury, and 90.2% of all acute kidney injuries that required subsequent administration of dialysis, with a lead time of up to 48 h and a ratio of 2 false alerts for every true alert. In addition to predicting future acute kidney injury, our model provides confidence assessments and a list of the clinical features that are most salient to each prediction, alongside predicted future trajectories for clinically relevant blood tests9. Although the recognition and prompt treatment of acute kidney injury is known to be challenging, our approach may offer opportunities for identifying patients at risk within a time window that enables early treatment.

BalanCI: Head-Referenced Cochlear Implant Stimulation Improves Balance in Children with Bilateral Cochleovestibular Loss.

INTRODUCTION: To determine the impact of a head-referenced cochlear implant (CI) stimulation system, BalanCI, on balance and postural control in children with bilateral cochleovestibular loss (BCVL) who use bilateral CI. METHODS: Prospective, blinded case-control study. Balance and postural control testing occurred in two settings: (1) quiet clinical setting and (2) immersive realistic virtual environment (Challenging Environment Assessment Laboratory [CEAL], Toronto Rehabilitation Institute). Postural control was assessed in 16 and balance in 10 children with BCVL who use bilateral CI, along with 10 typically developing children. Children with neuromotor, cognitive, or visual deficits that would prevent them from performing the tests were excluded. Children wore the BalanCI, which is a head-mounted device that couples with their CIs through the audio port and provides head-referenced spatial information delivered via the intracochlear electrode array. Postural control was measured by center of pressure (COP) and time to fall using the WiiTM (Nintendo, WA, USA) Balance Board for feet and the BalanCI for head, during the administration of the Modified Clinical Test of Sensory Interaction in Balance (CTSIB-M). The COP of the head and feet were assessed for change by deviation, measured as root mean square around the COP (COP-RMS), rate of deviation (COP-RMS/duration), and rate of path length change from center (COP-velocity). Balance was assessed by the Bruininks-Oseretsky Test of Motor Proficiency 2, balance subtest (BOT-2), specifically, BOT-2 score as well as time to fall/fault. RESULTS: In the virtual environment, children demonstrated more stable balance when using BalanCI as measured by an improvement in BOT-2 scores. In a quiet clinical setting, the use of BalanCI led to improved postural control as demonstrated by significant reductions in COP-RMS and COP-velocity. With the use of BalanCI, the number of falls/faults was significantly reduced and time to fall increased. CONCLUSIONS: BalanCI is a simple and effective means of improving postural control and balance in children with BCVL who use bilateral CI. BalanCI could potentially improve the safety of these children, reduce the effort they expend maintaining balance and allow them to take part in more complex balance tasks where sensory information may be limited and/or noisy.

Management of a type Ia endoleak with the Nellix endovascular aneurysm sealing system.

PURPOSE: To report the application of the Nellix endovascular aneurysm sealing (EVAS) device, including two chimney grafts, to successfully treat a type Ia endoleak. CASE REPORT: An 87-year-old man had an asymptomatic 7.6-cm infrarenal abdominal aortic aneurysm (AAA) and a 4.5-cm right internal iliac artery aneurysm treated using an aortouni-iliac stent-graft. Two years after the index endovascular repair, an asymptomatic type Ia endoleak was detected on duplex ultrasound; the computed tomographic angiogram (CTA) demonstrated significant sac enlargement and stent-graft migration. Initial attempts to treat the leak with 2 aortic cuffs only reduced the size of the endoleak. Another procedure was undertaken using the Nellix device with chimney grafts to increase the proximal sealing zone above the existing stent-graft. Imaging postoperatively demonstrated successful resolution of the endoleak and continuing patency of both renal artery chimney stent-grafts. CTA at 6 months demonstrated persistent sealing of the endoleak. CONCLUSION: The use of the EVAS system may represent another endovascular solution that can be added to the clinician's repertoire for treating type Ia endoleak after conventional endovascular repair of infrarenal AAA.

A Systematic Review and Meta-analysis of Endovascular Popliteal Aneurysm Repair Using the Hemobahn/Viabahn Stent-Graft.

PURPOSE: To perform an evidence synthesis study to assess outcomes of endovascular repair of popliteal artery aneurysms (PAAs) using the Hemobahn or Viabahn stent-graft. METHODS: A systematic literature review was conducted conforming to established standards to identify articles published between 1996 (the date of introduction of the Hemobahn stent-graft) and 2013 reporting stent-graft repair of PAAs in at least 10 patients. The data were pooled for Kaplan-Meier analysis of primary and secondary patency rates [presented with 95% confidence intervals (CIs)] as the primary outcomes. Random effects meta-analysis was performed for secondary outcomes that included rates of reintervention, endoleak, stent-graft fracture, and limb salvage. RESULTS: Fourteen studies reported outcomes for 514 PAAs. There was considerable heterogeneity in reporting standards among studies. Pooled primary and secondary patency rates were 69.4% (95% CI 63.3% to 76.2%) and 77.4% (95% CI 70.1% to 85.3%), respectively, at 5 years. Five studies (including only one randomized controlled trial) compared surgical to endovascular repair; no difference was found in primary patency on evidence synthesis (hazard ratio 1.30, 95% CI 0.79 to 12.14, p=0.189). CONCLUSION: Stent-graft repair provides a feasible treatment option for anatomically suitable PAAs. Further studies are required to optimize both patient selection and follow-up protocols.

Influence of foot ulceration on cause-specific mortality in patients with diabetes mellitus.

OBJECTIVE: The purpose of this study was to assess the odds of all-cause mortality in individuals with diabetic foot ulceration (DFU) compared with those with diabetes and no history of DFU. In addition, we sought to determine the strength of association of DFU with cardiovascular and nonvascular mortality. METHODS: We obtained data for a cohort of patients who attended a secondary care diabetic foot clinic or a general diabetes clinic between 2009 and 2010. A clinic cohort of patients with diabetes and no history of DFU provided a control group. Cause-specific mortality was recorded during a median follow-up duration of 3.6 years (interquartile range, 3.3-4.2 years). The association between DFU and all-cause mortality was evaluated by Cox regression. The association between DFU and cardiovascular mortality was determined by competing risk modeling. RESULTS: We recorded 145 events of all-cause mortality and 27 events of cardiovascular mortality among 869 patients with diabetes. After adjustment for potential confounders, DFU was associated with both cardiovascular disease (hazard ratio, 2.53; 95% confidence interval, 0.98-6.49; P = .05) and all-cause mortality (hazard ratio, 3.98; 95% confidence interval, 2.55-6.21; P < .001). The proportion of deaths attributable to cardiovascular disease was similar between the groups (18% with diabetes only and 19% with DFU; P = .91). CONCLUSIONS: DFU is associated with premature death from vascular and nonvascular causes.

Deep learning based linear energy transfer calculation for proton therapy.

Objective.This study aims to address the limitations of traditional methods for calculating linear energy transfer (LET), a critical component in assessing relative biological effectiveness (RBE). Currently, Monte Carlo (MC) simulation, the gold-standard for accuracy, is resource-intensive and slow for dose optimization, while the speedier analytical approximation has compromised accuracy. Our objective was to prototype a deep-learning-based model for calculating dose-averaged LET (LETd) using patient anatomy and dose-to-water (DW) data, facilitating real-time biological dose evaluation and LET optimization within proton treatment planning systems.Approach. 275 4-field prostate proton Stereotactic Body Radiotherapy plans were analyzed, rendering a total of 1100 fields. Those were randomly split into 880, 110, and 110 fields for training, validation, and testing. A 3D Cascaded UNet model, along with data processing and inference pipelines, was developed to generate patient-specific LETddistributions from CT images and DW. The accuracy of the LETdof the test dataset was evaluated against MC-generated ground truth through voxel-based mean absolute error (MAE) and gamma analysis.Main results.The proposed model accurately inferred LETddistributions for each proton field in the test dataset. A single-field LETdcalculation took around 100 ms with trained models running on a NVidia A100 GPU. The selected model yielded an average MAE of 0.94 ± 0.14 MeV cm-1and a gamma passing rate of 97.4% ± 1.3% when applied to the test dataset, with the largest discrepancy at the edge of fields where the dose gradient was the largest and counting statistics was the lowest.Significance.This study demonstrates that deep-learning-based models can efficiently calculate LETdwith high accuracy as a fast-forward approach. The model shows great potential to be utilized for optimizing the RBE of proton treatment plans. Future efforts will focus on enhancing the model's performance and evaluating its adaptability to different clinical scenarios.

Validation of clinical acceptability of deep-learning-based automated segmentation of organs-at-risk for head-and-neck radiotherapy treatment planning.

Introduction: Organ-at-risk segmentation for head and neck cancer radiation therapy is a complex and time-consuming process (requiring up to 42 individual structure, and may delay start of treatment or even limit access to function-preserving care. Feasibility of using a deep learning (DL) based autosegmentation model to reduce contouring time without compromising contour accuracy is assessed through a blinded randomized trial of radiation oncologists (ROs) using retrospective, de-identified patient data. Methods: Two head and neck expert ROs used dedicated time to create gold standard (GS) contours on computed tomography (CT) images. 445 CTs were used to train a custom 3D U-Net DL model covering 42 organs-at-risk, with an additional 20 CTs were held out for the randomized trial. For each held-out patient dataset, one of the eight participant ROs was randomly allocated to review and revise the contours produced by the DL model, while another reviewed contours produced by a medical dosimetry assistant (MDA), both blinded to their origin. Time required for MDAs and ROs to contour was recorded, and the unrevised DL contours, as well as the RO-revised contours by the MDAs and DL model were compared to the GS for that patient. Results: Mean time for initial MDA contouring was 2.3 hours (range 1.6-3.8 hours) and RO-revision took 1.1 hours (range, 0.4-4.4 hours), compared to 0.7 hours (range 0.1-2.0 hours) for the RO-revisions to DL contours. Total time reduced by 76% (95%-Confidence Interval: 65%-88%) and RO-revision time reduced by 35% (95%-CI,-39%-91%). All geometric and dosimetric metrics computed, agreement with GS was equivalent or significantly greater (p<0.05) for RO-revised DL contours compared to the RO-revised MDA contours, including volumetric Dice similarity coefficient (VDSC), surface DSC, added path length, and the 95%-Hausdorff distance. 32 OARs (76%) had mean VDSC greater than 0.8 for the RO-revised DL contours, compared to 20 (48%) for RO-revised MDA contours, and 34 (81%) for the unrevised DL OARs. Conclusion: DL autosegmentation demonstrated significant time-savings for organ-at-risk contouring while improving agreement with the institutional GS, indicating comparable accuracy of DL model. Integration into the clinical practice with a prospective evaluation is currently underway.

Validation and Clinical Applicability of Whole-Volume Automated Segmentation of Optical Coherence Tomography in Retinal Disease Using Deep Learning.

IMPORTANCE: Quantitative volumetric measures of retinal disease in optical coherence tomography (OCT) scans are infeasible to perform owing to the time required for manual grading. Expert-level deep learning systems for automatic OCT segmentation have recently been developed. However, the potential clinical applicability of these systems is largely unknown. OBJECTIVE: To evaluate a deep learning model for whole-volume segmentation of 4 clinically important pathological features and assess clinical applicability. DESIGN, SETTING, PARTICIPANTS: This diagnostic study used OCT data from 173 patients with a total of 15 558 B-scans, treated at Moorfields Eye Hospital. The data set included 2 common OCT devices and 2 macular conditions: wet age-related macular degeneration (107 scans) and diabetic macular edema (66 scans), covering the full range of severity, and from 3 points during treatment. Two expert graders performed pixel-level segmentations of intraretinal fluid, subretinal fluid, subretinal hyperreflective material, and pigment epithelial detachment, including all B-scans in each OCT volume, taking as long as 50 hours per scan. Quantitative evaluation of whole-volume model segmentations was performed. Qualitative evaluation of clinical applicability by 3 retinal experts was also conducted. Data were collected from June 1, 2012, to January 31, 2017, for set 1 and from January 1 to December 31, 2017, for set 2; graded between November 2018 and January 2020; and analyzed from February 2020 to November 2020. MAIN OUTCOMES AND MEASURES: Rating and stack ranking for clinical applicability by retinal specialists, model-grader agreement for voxelwise segmentations, and total volume evaluated using Dice similarity coefficients, Bland-Altman plots, and intraclass correlation coefficients. RESULTS: Among the 173 patients included in the analysis (92 [53%] women), qualitative assessment found that automated whole-volume segmentation ranked better than or comparable to at least 1 expert grader in 127 scans (73%; 95% CI, 66%-79%). A neutral or positive rating was given to 135 model segmentations (78%; 95% CI, 71%-84%) and 309 expert gradings (2 per scan) (89%; 95% CI, 86%-92%). The model was rated neutrally or positively in 86% to 92% of diabetic macular edema scans and 53% to 87% of age-related macular degeneration scans. Intraclass correlations ranged from 0.33 (95% CI, 0.08-0.96) to 0.96 (95% CI, 0.90-0.99). Dice similarity coefficients ranged from 0.43 (95% CI, 0.29-0.66) to 0.78 (95% CI, 0.57-0.85). CONCLUSIONS AND RELEVANCE: This deep learning-based segmentation tool provided clinically useful measures of retinal disease that would otherwise be infeasible to obtain. Qualitative evaluation was additionally important to reveal clinical applicability for both care management and research.

Use of deep learning to develop continuous-risk models for adverse event prediction from electronic health records.

Early prediction of patient outcomes is important for targeting preventive care. This protocol describes a practical workflow for developing deep-learning risk models that can predict various clinical and operational outcomes from structured electronic health record (EHR) data. The protocol comprises five main stages: formal problem definition, data pre-processing, architecture selection, calibration and uncertainty, and generalizability evaluation. We have applied the workflow to four endpoints (acute kidney injury, mortality, length of stay and 30-day hospital readmission). The workflow can enable continuous (e.g., triggered every 6 h) and static (e.g., triggered at 24 h after admission) predictions. We also provide an open-source codebase that illustrates some key principles in EHR modeling. This protocol can be used by interdisciplinary teams with programming and clinical expertise to build deep-learning prediction models with alternate data sources and prediction tasks.

Clinically applicable deep learning for diagnosis and referral in retinal disease.

The volume and complexity of diagnostic imaging is increasing at a pace faster than the availability of human expertise to interpret it. Artificial intelligence has shown great promise in classifying two-dimensional photographs of some common diseases and typically relies on databases of millions of annotated images. Until now, the challenge of reaching the performance of expert clinicians in a real-world clinical pathway with three-dimensional diagnostic scans has remained unsolved. Here, we apply a novel deep learning architecture to a clinically heterogeneous set of three-dimensional optical coherence tomography scans from patients referred to a major eye hospital. We demonstrate performance in making a referral recommendation that reaches or exceeds that of experts on a range of sight-threatening retinal diseases after training on only 14,884 scans. Moreover, we demonstrate that the tissue segmentations produced by our architecture act as a device-independent representation; referral accuracy is maintained when using tissue segmentations from a different type of device. Our work removes previous barriers to wider clinical use without prohibitive training data requirements across multiple pathologies in a real-world setting.

Microvascular disease and risk of cardiovascular events among individuals with type 2 diabetes: a population-level cohort study.

BACKGROUND: Diabetes confers a two times excess risk of cardiovascular disease, yet predicting individual risk remains challenging. The effect of total microvascular disease burden on cardiovascular disease risk among individuals with diabetes is unknown. METHODS: A population-based cohort of patients with type 2 diabetes from the UK Clinical Practice Research Datalink was studied (n=49 027). We used multivariable Cox models to estimate hazard ratios (HRs) for the primary outcome (the time to first major cardiovascular event, which was a composite of cardiovascular death, non-fatal myocardial infarction, or non-fatal ischaemic stroke) associated with cumulative burden of retinopathy, nephropathy, and peripheral neuropathy among individuals with no history of cardiovascular disease at baseline. FINDINGS: During a median follow-up of 5·5 years, 2822 (5·8%) individuals experienced a primary outcome. After adjustment for established risk factors, significant associations were observed for the primary outcome individually for retinopathy (HR 1·39, 95% CI 1·09-1·76), peripheral neuropathy (1·40, 1·19-1·66), and nephropathy (1·35, 1·15-1·58). For individuals with one, two, or three microvascular disease states versus none, the multivariable-adjusted HRs for the primary outcome were 1·32 (95% CI 1·16-1·50), 1·62 (1·42-1·85), and 1·99 (1·70-2·34), respectively. For the primary outcome, measures of risk discrimination showed significant improvement when microvascular disease burden was added to models. In the overall cohort, the net reclassification index for USA and UK guideline risk strata were 0·036 (95% CI 0·017-0·055, p<0·0001) and 0·038 (0·013-0·060, p<0·0001), respectively. INTERPRETATION: The cumulative burden of microvascular disease significantly affects the risk of future cardiovascular disease among individuals with type 2 diabetes. Given the prevalence of diabetes globally, further work to understand the mechanisms behind this association and strategies to mitigate this excess risk are warranted. FUNDING: Circulation Foundation.

Automated analysis of retinal imaging using machine learning techniques for computer vision.

There are almost two million people in the United Kingdom living with sight loss, including around 360,000 people who are registered as blind or partially sighted. Sight threatening diseases, such as diabetic retinopathy and age related macular degeneration have contributed to the 40% increase in outpatient attendances in the last decade but are amenable to early detection and monitoring. With early and appropriate intervention, blindness may be prevented in many cases. Ophthalmic imaging provides a way to diagnose and objectively assess the progression of a number of pathologies including neovascular ("wet") age-related macular degeneration (wet AMD) and diabetic retinopathy. Two methods of imaging are commonly used: digital photographs of the fundus (the 'back' of the eye) and Optical Coherence Tomography (OCT, a modality that uses light waves in a similar way to how ultrasound uses sound waves). Changes in population demographics and expectations and the changing pattern of chronic diseases creates a rising demand for such imaging. Meanwhile, interrogation of such images is time consuming, costly, and prone to human error. The application of novel analysis methods may provide a solution to these challenges. This research will focus on applying novel machine learning algorithms to automatic analysis of both digital fundus photographs and OCT in Moorfields Eye Hospital NHS Foundation Trust patients. Through analysis of the images used in ophthalmology, along with relevant clinical and demographic information, DeepMind Health will investigate the feasibility of automated grading of digital fundus photographs and OCT and provide novel quantitative measures for specific disease features and for monitoring the therapeutic success.