A multimodal deep learning model for cardiac resynchronisation therapy response prediction

We present a novel multimodal deep learning framework for cardiac resynchronisation therapy (CRT) response prediction from 2D echocardiography and cardiac magnetic resonance (CMR) data. The proposed method first uses the ‘nnU-Net’ segmentation model to extract segmentations of the heart over the full cardiac cycle from the two modalities. Next, a multimodal deep learning classifier is used for CRT response prediction, which combines the latent spaces of the segmentation models of the two modalities. At test time, this framework can be used with 2D echocardiography data only, whilst taking advantage of the implicit relationship between CMR and echocardiography features learnt from the model. We evaluate our pipeline on a cohort of 50 CRT patients for whom paired echocardiography/CMR data were available, and results show that the proposed multimodal classifier results in a statistically significant improvement in accuracy compared to the baseline approach that uses only 2D echocardiography data. The combination of multimodal data enables CRT response to be predicted with 77.38% accuracy (83.33% sensitivity and 71.43% specificity), which is comparable with the current state-of-the-art in machine learning-based CRT response prediction. Our work represents the first multimodal deep learning approach for CRT response prediction.     

Deep weakly-supervised breast tumor segmentation in ultrasound images with explicit anatomical constraints

Breast tumor segmentation is an important step in the diagnostic procedure of physicians and computer-aided diagnosis systems. We propose a two-step deep learning framework for breast tumor segmentation in breast ultrasound (BUS) images which requires only a few manual labels. The first step is breast anatomy decomposition handled by a semi-supervised semantic segmentation technique. The input BUS image is decomposed into four breast anatomical structures, namely fat, mammary gland, muscle and thorax layers. Fat and mammary gland layers are used as constrained region to reduce the search space for breast tumor segmentation. The second step is breast tumor segmentation performed in a weakly-supervised learning scenario where only image-level labels are available. Breast tumors are first recognized by a classification network and then segmented by the proposed class activation mapping and deep level set (CAM-DLS) method. For breast anatomy decomposition, the proposed framework achieves Dice similarity coefficient (DSC) of 83.0 ± 11.8%, 84.3 ± 10.0%, 80.7 ± 15.4% and 91.0 ± 11.4% for fat, mammary gland, muscle and thorax layers, respectively. For breast tumor recognition, the proposed framework achieves sensitivity of 95.8%, precision of 92.4%, specificity of 93.9%, accuracy of 94.8% and F1-score of 0.941. For breast tumor segmentation, the proposed framework achieves DSC of 77.3% and intersection-over-union (IoU) of 66.0%. In conclusion, the proposed framework could efficiently perform breast tumor recognition and segmentation simultaneously in a weakly-supervised setting with anatomical constraints.     

An automated near-real time computational method for induction and treatment of scar-related ventricular tachycardias

Catheter ablation is currently the only curative treatment for scar-related ventricular tachycardias (VTs). However, not only are ablation procedures long, with relatively high risk, but success rates are punitively low, with frequent VT recurrence. Personalized in-silico approaches have the opportunity to address these limitations. However, state-of-the-art reaction diffusion (R-D) simulations of VT induction and subsequent circuits used for in-silico ablation target identification require long execution times, along with vast computational resources, which are incompatible with the clinical workflow. Here, we present the Virtual Induction and Treatment of Arrhythmias (VITA), a novel, rapid and fully automated computational approach that uses reaction-Eikonal methodology to induce VT and identify subsequent ablation targets. The rationale for VITA is based on finding isosurfaces associated with an activation wavefront that splits in the ventricles due to the presence of an isolated isthmus of conduction within the scar; once identified, each isthmus may be assessed for their vulnerability to sustain a reentrant circuit, and the corresponding exit site automatically identified for potential ablation targeting. VITA was tested on a virtual cohort of 7 post-infarcted porcine hearts and the results compared to R-D simulations. Using only a standard desktop machine, VITA could detect all scar-related VTs, simulating activation time maps and ECGs (for clinical comparison) as well as computing ablation targets in 48 minutes. The comparable VTs probed by the R-D simulations took 68.5 hours on 256 cores of high-performance computing infrastructure. The set of lesions computed by VITA was shown to render the ventricular model VT-free. VITA could be used in near real-time as a complementary modality aiding in clinical decision-making in the treatment of post-infarction VTs.     

Knee impairments: Comparison between new clinical classification by cluster analysis and movement system impairment model

BackgroundAlthough Movement System Impairment (MSI) model classifies patients with knee impairments, it has some limitations. This study compares MSI model with a new Partitioning Around Medoids (PAM) model in knee pain patients.MethodsIn this cross-sectional study, knee movement impairments (signs) and symptoms. (pain) of 200 patients were studied in eight different functional positions. After modification of impairments, the examinations were repeated to record the changes in symptoms. The patients were then classified based on the signs and symptoms firstly by MSI model and secondly using PAM clustering by R software.ResultsPAM model has a similar acceptable grouping compared to MSI because most patients are in 4 similar categories in both methods: valgus, varus, hyperextension and hypomobility. However, due to low discriminative power of the tests used for finding hypermobility and patellar dysfunctions, these sub-clusters are absent in PAM model. Instead, two new sub-clusters of self-management and valgus with hypomobility were found. Most importantly, the PAM model sorted the signs and symptoms based on their discriminative power and eliminated trivial tests so that the therapist can classify patients more quickly by performing clinically relevant tests.ConclusionThe new PAM method can be advantageous for therapists since it defines the importance of signs over symptoms in examination, prioritizes examination tests, and outlines tests with lower discriminative power. In PAM model, patients in the hypermobility and patellar subgroups of MSI model merged into other sub-clusters due to low discriminating power of their characteristics.     

DeepSMILE: Contrastive self-supervised pre-training benefits MSI and HRD classification directly from H&E whole-slide images in colorectal and breast cancer

We propose a Deep learning-based weak label learning method for analyzing whole slide images (WSIs) of Hematoxylin and Eosin (H&E) stained tumor tissue not requiring pixel-level or tile-level annotations using Self-supervised pre-training and heterogeneity-aware deep Multiple Instance LEarning (DeepSMILE). We apply DeepSMILE to the task of Homologous recombination deficiency (HRD) and microsatellite instability (MSI) prediction. We utilize contrastive self-supervised learning to pre-train a feature extractor on histopathology tiles of cancer tissue. Additionally, we use variability-aware deep multiple instance learning to learn the tile feature aggregation function while modeling tumor heterogeneity. For MSI prediction in a tumor-annotated and color normalized subset of TCGA-CRC (n=360 patients), contrastive self-supervised learning improves the tile supervision baseline from 0.77 to 0.87 AUROC, on par with our proposed DeepSMILE method. On TCGA-BC (n=1041 patients) without any manual annotations, DeepSMILE improves HRD classification performance from 0.77 to 0.81 AUROC compared to tile supervision with either a self-supervised or ImageNet pre-trained feature extractor. Our proposed methods reach the baseline performance using only 40% of the labeled data on both datasets. These improvements suggest we can use standard self-supervised learning techniques combined with multiple instance learning in the histopathology domain to improve genomic label classification performance with fewer labeled data.     

Skin Cancer in Solid Organ Transplant Recipients: A Review for the Nondermatologist

Solid organ transplant recipients (SOTRs) are at increased risk for the development of skin cancer compared with the general population, which requires consistent monitoring and management from a multidisciplinary team. The aim of this review is to provide a comprehensive overview for nondermatologist clinicians, outlining skin cancer diagnosis, treatment pearls, and skin cancer prevention strategies as they relate to SOTRs. A comprehensive search of the literature was conducted through the MEDLINE database with search terms including organ transplantation, transplant recipient, skin cancer, cutaneous neoplasms, management, and therapies. The search was limited to the English language and dates ranging from January 1, 2011, to December 28, 2021. All studies were reviewed for inclusion. Skin cancer will develop in more than half of SOTRs at some point in their life, most often nonmelanoma skin cancer such as basal cell carcinoma or squamous cell carcinoma. Melanoma and rarer cutaneous malignant neoplasms, such as Merkel cell carcinoma and Kaposi sarcoma, are also more frequent among SOTRs. A multidisciplinary effort at skin cancer screening and patient education is invaluable to prevent skin cancer–related morbidity and mortality in this population of patients. Reduction in immunosuppressive medications and surgical intervention are effective therapeutic approaches, and more novel systemic therapies including G protein–coupled receptor inhibitors and immune checkpoint inhibitors are possible options when traditional treatment approaches are not feasible. Checkpoint inhibitor therapy, however, comes with the risk of allograft rejection. With a growing and aging SOTR population, it is essential that SOTRs have support from dermatologists and nondermatologists alike in skin cancer prevention and treatment.     

Non-pharmacological interventions for spasticity in adults: An overview of systematic reviews

ObjectivesSpasticity causes significant long-term disability-burden, requiring comprehensive management. This review evaluates evidence from published systematic reviews of clinical trials for effectiveness of non-pharmacological interventions for improved spasticity outcomes.MethodsData sources: a literature search was conducted using medical and health science electronic (MEDLINE, EMBASE, CINAHL, PubMed, and the Cochrane Library) databases for published systematic reviews up to 15th June 2017. Data extraction and synthesis: two reviewers applied inclusion criteria to select potential systematic reviews, independently extracted data for methodological quality using Assessment of Multiple Systematic Reviews (AMSTAR). Quality of evidence was critically appraised with Grades of Recommendation, Assessment, Development and Evaluation (GRADE).ResultsOverall 18 systematic reviews were evaluated for evidence for a range of non-pharmacological interventions currently used in managing spasticity in various neurological conditions. There is “moderate” evidence for electro-neuromuscular stimulation and acupuncture as an adjunct therapy to conventional routine care (pharmacological and rehabilitation) in persons following stroke. “Low” quality evidence for rehabilitation programs targeting spasticity (such as induced movement therapy, stretching, dynamic elbow-splinting, occupational therapy) in stroke and other neurological conditions; extracorporeal shock-wave therapy in brain injury; transcranial direct current stimulation in stroke; transcranial magnetic stimulation and transcutaneous electrical nerve stimulation for other neurological conditions; physical activity programs and repetitive magnetic stimulation in persons with MS, vibration therapy for SCI and stretching for other neurological condition. For other interventions, evidence was inconclusive.ConclusionsDespite the available range of non-pharmacological interventions for spasticity, there is lack of high-quality evidence for many modalities. Further research is needed to judge the effect with appropriate study designs, timing and intensity of modalities, and associate costs of these interventions.     

Bleeding in Patients With Gastrointestinal Cancer Compared With Nongastrointestinal Cancer Treated With Apixaban,Rivaroxaban, or Enoxaparin for Acute Venous Thromboembolism

ObjectiveTo compare the bleeding risk in patients with gastrointestinal (GI) cancer with that in patients with non-GI cancer treated with anticoagulation for acute cancer-associated venous thromboembolism (Ca-VTE).Patients and MethodsConsecutive patients with Ca-VTE seen at the Mayo Thrombophilia Clinic between March 1, 2013, and April 20, 2020, were observed prospectively to assess major bleeding and clinically relevant nonmajor bleeding (CRNMB).ResultsIn the group of 1392 patients with Ca-VTE, 499 (35.8%) had GI cancer including 272 with luminal GI cancer (lower GI, 208; upper GI, 64), 176 with pancreatic cancer, and 51 with hepatobiliary cancer. The rate of major bleeding and CRNMB in patients with GI cancer was similar to that in 893 (64.2%) patients with non-GI cancer treated with apixaban, rivaroxaban, or enoxaparin. Apixaban had a higher rate of major bleeding in luminal GI cancer compared with the non-GI cancer group (15.59 vs 3.26 per 100 person-years; P=.004) and compared with enoxaparin in patients with luminal GI cancer (15.59 vs 3.17; P=.04). Apixaban had a lower rate of CRNMB compared with rivaroxaban in patients with GI cancer (3.83 vs 9.40 per 100 person-years; P=.03). Patients treated with rivaroxaban in the luminal GI cancer group had a major bleeding rate similar to that of patients with non-GI cancer (2.04 vs 4.91 per 100 person-years; P=.37).ConclusionApixaban has a higher rate of major bleeding in patients with luminal GI cancer compared with patients with non-GI cancer and compared with enoxaparin in patients with luminal GI cancer. Rivaroxaban shows no increased risk of major bleeding in patients with GI cancer or luminal GI cancer compared with patients with non-GI cancer.Trial RegistrationClinicalTrials.gov identifier: NCT03504007.     

Real-Life Evaluation of an Algorithm for the Diagnosis of Cardiac Amyloidosis

ObjectiveTo evaluate the real-life use of a modified Gillmore algorithm with a "one-stop-shop” approach, bone scintigraphy (BS), a monoclonal gammopathy test (GT), a salivary gland biopsy (SGB), and genetic testing performed at the same time for the diagnosis of cardiac amyloidosis at the French National Reference Centre for Cardiac Amyloidosis (Henri Mondor Hospital, Créteil, France).MethodsThis retrospective cohort study included a total of 1222 patients with suspected amyloidosis who underwent BS and GT between June 2008 and May 2019.ResultsOf 1222 patients, 349 had no cardiac uptake on BS and negative GT (BS?/GT?), 276 were BS?/GT positive (GT+), 420 patients were BS+/GT?, and 177 were BS+/GT+. Our one-stop-shop check-up enabled us to diagnose 892 (72.9%) patients; 330 (27.0%) patients required additional examinations, such as mass spectrometry and/or a cardiac biopsy. This subset notably included 112 patients with amyloid light chain amyloidosis. More than 64% of the patients with transthyretin amyloidosis or another type of amyloidosis were diagnosed during the one-stop shop visit. Sensitivity and specificity of BS for transthyretin amyloidosis diagnosis was 99% and 96%, respectively. For amyloid light chain diagnosis, sensitivity and specificity were 100% and 76%, respectively, for GT and 54% and 100%, respectively, for SGB. Of 910 transthyretin genetic tests, 205 (17%) detected mutations.ConclusionThe results of our real-life cohort study confirmed the ability of a one-stop-shop approach with a modified Gillmore algorithm to diagnose cardiac amyloidosis and the interest of simultaneous testing for earlier diagnosis. The SGB has diagnostic value because it is easy, quick, and less invasive than a cardiac biopsy.     

Indirect Comparison of Glucocorticoid-Sparing Agents for Remission Maintenance in Giant Cell Arteritis: A Network Meta-analysis

ObjectiveTo compare and rank the effect of glucocorticoid-sparing agents in giant cell arteritis (GCA), for which several drugs have been evaluated but with a benefit-risk balance that remains uncertain.MethodsThe MEDLINE and Clinical Trials databases were searched up to November 2021; all randomized controlled trials investigating glucocorticoids in GCA were included. The glucocorticoid regimen was dichotomized into short (≤6 months) or prolonged (>6 months) use. Risk of relapse and safety were estimated using network meta-analysis with frequentist random effects models.ResultsOf the 96 records screened, 8 trials were included (572 patients). The trials compared glucocorticoids and a sparing agent: tocilizumab (2 trials), oral methotrexate (3 trials), infliximab (1 trial), etanercept (1 trial), and adalimumab (1 trial). The pooled prevalence of GCA relapse was 52.6% (95% CI, 38.1 to 66.9). The risk of relapse was significantly lower with tocilizumab compared with methotrexate (relative risk [RR], 0.41; 95% CI, 0.17 to 0.97) and prolonged (RR, 0.41; 95% CI, 0.20 to 0.83) and short (RR, 0.32; 95% CI, 0.16 to 0.66) glucocorticoid use. The risk of relapse was not significantly different with methotrexate compared with short (RR, 0.79; 95% CI, 0.48 to 1.31) and prolonged (RR, 0.95; 95% CI, 0.31 to 2.89) glucocorticoid use. The frequency of serious adverse events and serious infection was comparable between the different drugs. The certainty of the evidence was low to very low.ConclusionThis meta-analysis suggests that tocilizumab may be superior to other sparing agents to prevent GCA relapse, but with a low to very low certainty of evidence, and that safety is comparable to the other drugs.RegistrationThe protocol of the meta-analysis is registered in the international prospective register of systematic reviews PROSPERO (https://www.crd.york.ac.uk/prospero/; registration CRD42020112387).     

Integrating informatics into undergraduate nursing education: A case study using a spiral learning approach

A gap in informatics expertise amongst nursing students, practising staff and faculty has been noted globally, which reduces the potential for nurses to utilise technology to enhance patient care. National nursing education strategies and recommendations from professional associations have identified digital health as an area that needs investment. This case study describes how health informatics is being integrated into a Bachelor of Nursing programme in the United Kingdom. An international collaboration with a US-UK Fulbright Specialist Scholar enabled individual learning units corresponding to key health informatics competencies to be designed and incorporated into a pedagogic framework grounded in the spiral learning approach. This approach is proposed as one way to integrate informatics into nursing education, so students can become competent clinicians that are able to deliver technology enabled care in the health service.     

Construction and Application of Chain Management Information System for Cancer Pain

BackgroundWith the rapid development of information technology, hospital information systems (HISs) have been deeply applied in the medical field and have shown broad application prospects. There are still some noninteroperable clinical information systems that pose an obstacle to the effective coordination of care, such as cancer pain management.AimTo construct a chain management information system for cancer pain and explore its clinical application effect.MethodsA quasiexperimental study was conducted in the inpatient department of Sir Run Run Shaw Hospital, Zhejiang University School of Medicine. A total of 259 patients were nonrandomly divided into two groups: the experimental group (after the system was applied, n = 123) and the control group (before the system was applied, n = 136). The cancer pain management evaluation form score, patient satisfaction with pain control, pain score at admission and discharge, and the worst pain score during hospitalization were compared between the two groups.ResultsCompared with the control group, the score of the cancer pain management evaluation form was significantly higher (p < .05). There were no statistically significant differences in worst pain intensity, pain score at admission and discharge, and patients’ satisfaction with pain control between the two groups.ConclusionsThe cancer pain chain management information system can enable nurses to evaluate and record pain in a more standardized way, but it has no significant effect on the pain intensity of cancer patients.     

Population-Based Trends in Amputations and Revascularizations for Peripheral Artery Disease From 1990 to 2009

ObjectiveTo determine trends in amputations and revascularizations for peripheral artery disease (PAD) in a well-defined population.MethodsA population-based cohort study of Olmsted County, Minnesota, residents with PAD undergoing amputation or revascularization was conducted between January 1, 1990, and December 31, 2009. Population-level 5-year incidence trends for endovascular, open surgical, and hybrid revascularizations and major and minor amputations were determined. Limb-specific outcomes after revascularization, including major adverse limb events and amputation-free survival, were compared between initial surgical and endovascular or hybrid revascularization groups using Kaplan-Meier analysis.ResultsWe identified 773 residents who underwent 1906 limb-procedures, including 689 open revascularizations, 685 endovascular or hybrid revascularizations, and 220 major amputations. During the 20-year study period, the incidence of endovascular and hybrid revascularizations increased, whereas the incidence of open surgical revascularizations and major amputations decreased. Incidence of revascularizations for chronic limb-threatening ischemia (CLTI) did not change. Among residents with CLTI undergoing their first revascularization on a limb, endovascular revascularization was associated with more major adverse limb events and major amputations compared with surgical revascularization during the ensuing 15 years.ConclusionThe rising incidence of endovascular and hybrid revascularizations and the decreasing incidence of open surgical revascularizations for PAD were associated with a decreasing incidence of major amputations in this population between 1990 and 2009, despite a stable incidence of revascularizations for CLTI. With more major adverse limb events and major amputations after endovascular revascularization, these trends suggest that additional emphasis should be placed on improving limb salvage efforts beyond just mode of revascularization.     

Incidence and Risk of Various Types of Arterial Thromboembolism in Patients With Cancer

ObjectiveTo describe the temporal association and identify risk factors between cancer diagnosis and various types of arterial thromboembolism (ATE).Patients and MethodsWe inquired an aggregated electronic health record database (Explorys, IBM Corp., Armonk, New York) and identified patients with cancer from January 1999 to October 2019, with various types of ATE, including myocardial infarction, acute ischemic stroke, acute limb ischemia, acute mesenteric ischemia, acute renal infarction, and retinal artery occlusion. We investigated the temporal relationship between cancer diagnosis and ATE events by examining the incidence ratio (IR) of ATE before and after diagnosis of cancer.ResultsWe identified 305,384 patients with cancer and ATE. The 30-day interval IR of total ATE was elevated shortly before and after cancer diagnosis, which was consistent among different ATE and cancer types. The incidence was highest within a 330-day window (90 days before and 240 days after cancer diagnosis), and IR peaked at 13.9 (95% confidence interval [CI], 13.6 to 14.2) in the first 30 days following diagnosis of cancer. Compared with patients with cancer who never developed ATE, patients with ATE had more cardiovascular risk factors at baseline. Patients with brain cancer, lung cancer, colorectal cancer, and pancreatic cancer had the highest risk of developing ATE, whereas ATE type was anatomically associated with cancer type.ConclusionIn this observational study of an aggregated US patient population, those with newly diagnosed cancer had increased risk of ATE events. This risk was most elevated in a 330-day window around cancer diagnosis and was consistent across different types of ATE and cancer.