Optimal Deep CNN–Based Vectorial Variation Filter for Medical Image Denoising

Journal of Digital Imaging - Medical imaging has acquired more attention due to the emerging design of wireless technologies, the internet, and data storage. The reflection of these technologies...     

Medical Image Data and Datasets in the Era of Machine Learning—Whitepaper from the 2016 C-MIMI Meeting Dataset Session

At the first annual Conference on Machine Intelligence in Medical Imaging (C-MIMI), held in September 2016, a conference session on medical image data and datasets for machine learning identified multiple issues. The common theme from attendees was that everyone participating in medical image evaluation with machine learning is data starved. There is an urgent need to find better ways to collect, annotate, and reuse medical imaging data. Unique domain issues with medical image datasets require further study, development, and dissemination of best practices and standards, and a coordinated effort among medical imaging domain experts, medical imaging informaticists, government and industry data scientists, and interested commercial, academic, and government entities. High-level attributes of reusable medical image datasets suitable to train, test, validate, verify, and regulate ML products should be better described. NIH and other government agencies should promote and, where applicable, enforce, access to medical image datasets. We should improve communication among medical imaging domain experts, medical imaging informaticists, academic clinical and basic science researchers, government and industry data scientists, and interested commercial entities.     

French 2010–2011 measles outbreak in adults: report from a Parisian teaching hospital

We reviewed 80 adult cases of measles seen in a Parisian hospital during the French 2010–2011 outbreak. Fifty per cent had at least one complication: pneumonia and hepatitis were the most frequent. Forty per cent of hospitalized cases did not have any complications, suggesting clinically poor tolerance of measles in adults. The outcome was always favourable. Subjects were younger, were more often French nationals and had a higher socio-economic status than the overall population. This report suggests that immunity resulting from natural disease in patients from an area where the disease is endemic is protective in the long term.     

Predictors of Default from Treatment for Tuberculosis: a Single Center Case–Control Study in Korea

Default from tuberculosis (TB) treatment could exacerbate the disease and result in the emergence of drug resistance. This study identified the risk factors for default from TB treatment in Korea. This single-center case–control study analyzed 46 default cases and 100 controls. Default was defined as interrupting treatment for 2 or more consecutive months. The reasons for default were mainly incorrect perception or information about TB (41.3%) and experience of adverse events due to TB drugs (41.3%). In univariate analysis, low income (< 2,000 US dollars/month, 88.1% vs. 68.4%, P = 0.015), absence of TB stigma (4.3% vs. 61.3%, P < 0.001), treatment by a non-pulmonologist (74.1% vs. 25.9%, P < 0.001), history of previous treatment (37.0% vs. 19.0%, P = 0.019), former defaulter (15.2% vs. 2.0%, P = 0.005), and combined extrapulmonary TB (54.3% vs. 34.0%, P = 0.020) were significant risk factors for default. In multivariate analysis, the absence of TB stigma (adjusted odd ratio [aOR]: 46.299, 95% confidence interval [CI]: 8.078–265.365, P < 0.001), treatment by a non-pulmonologist (aOR: 14.567, 95% CI: 3.260–65.089, P < 0.001), former defaulters (aOR: 33.226, 95% CI: 2.658–415.309, P = 0.007), and low income (aOR: 5.246, 95% CI: 1.249–22.029, P = 0.024) were independent predictors of default from TB treatment. In conclusion, patients with absence of disease stigma, treated by a non-pulmonologist, who were former defaulters, and with low income should be carefully monitored during TB treatment in Korea to avoid treatment default.     

Assessment of incidental findings in 232 whole-exome sequences from the Baylor–Hopkins Center for Mendelian Genomics

PurposeIn March 2013 the American College of Medical Genetics and Genomics published a list of 56 genes with the recommendation that pathogenic and likely pathogenic variants detected incidentally by clinical sequencing be reported to patients. As an initial step in determining the practical consequences of this recommendation in the research setting, we searched for variants in these genes in 232 whole-exome sequences from the Baylor-Hopkins Center for Mendelian Genomics.MethodsWe identified rare, nonsynonymous, and splicing single-nucleotide variants and insertions/deletions and assessed variant classification using the Human Gene Mutation, Emory, and ClinVar databases. We analyzed the burden of mutation in each of the 56 genes and determined which variants should be reported to patients.ResultsOur filtering resulted in 249 distinct variants, with a mean of 1.69 variants per individual. Half of these were novel missense mutations not classified by any of the three reference databases. Of 101 variants listed in the Human Gene Mutation Database, 48 were also in ClinVar and 3 were also in Emory; half of these shared variants were classified discordantly between databases. Some genes consistently had greater variation than others. In total, 0.86% of individuals had a reportable incidental variant.ConclusionThese observations demonstrate some current challenges of assessing phenotypic consequences of incidental variants for counseling patients.     

Payer decision making for next-generation sequencing–based genetic tests: insights from cell-free DNA prenatal screening

PurposeCell-free DNA (cfDNA) prenatal screening tests have been rapidly adopted into clinical practice, due in part to positive insurance coverage. We evaluated the framework payers used in making coverage decisions to describe a process that should be informative for other sequencing tests.MethodsWe analyzed coverage policies from the 19 largest US private payers with publicly available policies through February 2016, building from the University of California San Francisco TRANSPERS Payer Coverage Policy Registry.ResultsAll payers studied cover cfDNA screening for detection of trisomies 21, 18, and 13 in high-risk, singleton pregnancies, based on robust clinical validity (CV) studies and modeled evidence of clinical utility (CU). Payers typically evaluated the evidence for each chromosomal abnormality separately, although results are offered as part of a panel. Starting in August 2015, 8 of the 19 payers also began covering cfDNA screening in average-risk pregnancies, citing recent CV studies and updated professional guidelines. Most payers attempted, but were unable, to independently assess analytic validity (AV).ConclusionPayers utilized the standard evidentiary framework (AV/CV/CU) when evaluating cfDNA screening but varied in their interpretation of the sufficiency of the evidence. Professional guidelines, large CV studies, and decision analytic models regarding health outcomes appeared highly influential in coverage decisions.Genet Med advance online publication 22 September 2016     

Quantifying the risk of multiple myeloma from symptoms reported in primary care patients: a large case–control study using electronic records

Background

Patients with myeloma experience the longest diagnostic delays compared with patients with other cancers in the UK; 37% are diagnosed through emergency presentations.

Aim

To identify and quantify the risk of myeloma from specific clinical features reported by primary care patients.

Design and setting

Matched case–control study using General Practice Research Database primary care electronic records.

Method

Putative clinical features of myeloma were identified and analysed using conditional logistic regression. Positive predictive values (PPVs) were calculated for the consulting population.

Results

A total of 2703 patients aged ≥40 years, diagnosed with myeloma between 2000 and 2009, and 12 157 age, sex, and general practice-matched controls were identified. Sixteen features were independently associated with myeloma: hypercalcaemia, odds ratio 11.4 (95% confidence interval [CI] = 7.1 to 18), cytopenia 5.4 (95% CI = 4.6 to 6.4), raised inflammatory markers 4.9 (95% CI = 4.2 to 5.8), fracture 3.1 (95% CI = 2.3 to 4.2), raised mean corpuscular volume 3.1 (95% CI = 2.4 to 4.1), weight loss 3.0 (95% CI = 2.0 to 4.5), nosebleeds 3.0 (95% CI = 1.9 to 4.7), rib pain 2.5 (95% CI = 1.5 to 4.4), back pain 2.2 (95% CI = 2.0 to 2.4), other bone pain 2.1 (95% CI = 1.4 to 3.1), raised creatinine 1.8 (95% CI = 1.5 to 2.2), chest pain 1.6 (95% CI = 1.4 to 1.8), joint pain 1.6 (95% CI = 1.2 to 2.2), nausea 1.5 (95% CI = 1.1 to 2.1), chest infection 1.4 (95% CI = 1.2 to 1.6), and shortness of breath 1.3 (95% CI = 1.1 to 1.5). Individual symptom PPVs were generally <1%, although were >10% for some symptoms when combined with leucopenia or hypercalcaemia.

Conclusion

Individual symptoms of myeloma in primary care are generally low risk, probably explaining diagnostic delays. Once simple primary care blood tests are taken, risk estimates change. Hypercalcaemia and leucopenia are particularly important abnormalities, and coupled with symptoms, strongly suggest myeloma.     

Supporting information retrieval from electronic health records: A report of University of Michigan’s nine-year experience in developing and using the Electronic Medical Record Search Engine (EMERSE)

ObjectiveThis paper describes the University of Michigan’s nine-year experience in developing and using a full-text search engine designed to facilitate information retrieval (IR) from narrative documents stored in electronic health records (EHRs). The system, called the Electronic Medical Record Search Engine (EMERSE), functions similar to Google but is equipped with special functionalities for handling challenges unique to retrieving information from medical text.Materials and methodsKey features that distinguish EMERSE from general-purpose search engines are discussed, with an emphasis on functions crucial to (1) improving medical IR performance and (2) assuring search quality and results consistency regardless of users’ medical background, stage of training, or level of technical expertise.ResultsSince its initial deployment, EMERSE has been enthusiastically embraced by clinicians, administrators, and clinical and translational researchers. To date, the system has been used in supporting more than 750 research projects yielding 80 peer-reviewed publications. In several evaluation studies, EMERSE demonstrated very high levels of sensitivity and specificity in addition to greatly improved chart review efficiency.DiscussionIncreased availability of electronic data in healthcare does not automatically warrant increased availability of information. The success of EMERSE at our institution illustrates that free-text EHR search engines can be a valuable tool to help practitioners and researchers retrieve information from EHRs more effectively and efficiently, enabling critical tasks such as patient case synthesis and research data abstraction.ConclusionEMERSE, available free of charge for academic use, represents a state-of-the-art medical IR tool with proven effectiveness and user acceptance.     

Care Utilization with China’s New Rural Cooperative Medical Scheme: Updated Evidence from the China Health and Retirement Longitudinal Study 2011–2012

Background

China’s New Rural Cooperative Medical Scheme (NRCMS), a healthcare financing system for rural residents in China, underwent significant enhancement since 2008. Studies based on pre-2008 NRCMS data showed an increase in inpatient care utilization after NRCMS coverage. However evidence was mixed for the relationship between outpatient care use and NRCMS coverage.

Purpose

We assessed whether enrollment in the enhanced NRCMS was associated with less delaying or foregoing medical care, as a reduction in foregoing needed care signals about removing liquidity constraint among the enrollees.

Method

Using a national sample of rural residents (N?=?12,740) from the 2011–2012 wave of China Health and Retirement Longitudinal Study, we examined the association between NRCMS coverage and the likelihood of delaying or foregoing medical care (outpatient and inpatient) by survey-weighted regression models controlling for demographics, education, geographic regions, household expenditures, pre-existing chronic diseases, and access to local healthcare facilities. Zero-inflated negative binomial model was used to estimate the association between NRCMS coverage and number of medical visits.

Results

NRCMS coverage was significantly associated with lower odds of delaying or foregoing inpatient care (OR: 0.42, 95 % CI: 0.22–0.81). A negative but insignificant association was found between NRCMS coverage and delaying/foregoing outpatient care when ill. Among those who needed health care, the expected number of outpatient visits for NRCMS enrollees was 1.35 (95 % CI: 1.03–1.77) times of those uninsured, and the expected number of inpatient visits for NRCMS enrollees was 1.83 (95 % CI: 1.16–2.88) times of those uninsured.

Conclusion

This study shows that the enhanced NRCMS coverage was associated with less delaying or foregoing inpatient care deemed as necessary by health professionals, which is likely to result from improved financial reimbursement of the NRCMS.

     

A Cohort of 169 Chronic Granulomatous Disease Patients Exposed to BCG Vaccination: a Retrospective Study from a Single Center in Shanghai,China (2004–2017)

Purpose

Clinical diagnosis and treatment for chronic granulomatous disease (CGD) have advanced greatly in recent years. However, CGD patients in China have unique clinical features and infection spectrums, which are challenging to their caretakers. Here, we summarized the clinical characteristics, genetic features, treatment, and prognosis of CGD in a single center in Shanghai.

Methods

One hundred sixty-nine CGD patients were recruited between January 2004 and May 2017 based on clinical diagnosis. Electronic medical charts were reviewed to collect clinical data.

Results

Among the 169 patients recruited, CYBB mutations were identified in 150 cases, whereas CYBA mutations were identified in 7 cases, NCF1 in 5, and NCF2 in 7. The medium age at onset was 1 month (interquartile range 1–3). The medium age at diagnosis was 8 months (interquartile range 3–19). The most common infection sites were the lung (95.9%), lymph node (58.5%), skin (45.4%), intestinal (43.1%), and perianal (38.5%). Bacillus Calmette-Guérin (BCG) infections were common (59.2%). In addition, other non-infectious complications were also common, including anemia (55.4%) and impaired liver functions (34.6%). Thirty-one patients received stem cell transplantation. By the end of this study, 83/131 patients survived.

Conclusions

Similar to other non-consanguineous populations, X-linked CGD accounted for the majority of the cases in China. However, BCG infections were a clinical challenge unique to China. In addition, severe infections were the major cause of death and the overall mortality was still high in China.

     

Barriers to Allogeneic Hematopoietic Stem Cell Transplantation for Human T Cell Lymphotropic Virus 1–Associated Adult T Cell Lymphoma–Leukemia in the United States: Experience from a Large Cohort in a Major Tertiary Center

In the United States adult T cell lymphoma–leukemia (ATLL) carries a dismal prognosis and mainly affects immigrants from human T cell lymphotropic virus 1 endemic areas. Allogeneic hematopoietic stem cell transplant (alloHSCT) can be effective and is recommended as an upfront treatment in the National Comprehensive Cancer Network guidelines. We studied the barriers to alloHSCT in one of the largest ATLL populations in the United States. Comprehensive chart and donor registry reviews were conducted for 88 ATLL patients treated at Montefiore Medical Center from 2003 to 2018. Among 49 patients with acute and 32 with lymphomatous subtypes, 48 (59.5%) were ineligible for alloHSCT because of early mortality (52%), loss to follow-up (21%), uninsured status (15%), patient declination (10%), and frailty (2%). Among 28 HLA-typed eligible patients (34.6%), matched related donors were identified for 7 (25%). A matched unrelated donor (MUD) search yielded HLA-matched in 2 patients (9.5%), HLA mismatched in 6 (28.5%), and no options in 13 (62%). Haploidentical donors were identified for 6 patients (46%) with no unrelated options. There were no suitable donors for 7 (25%) alloHSCT-eligible patients. The main limitation for alloHSCT after donor identification was death from progressive disease (82%). AlloHSCT was performed in 10 patients (12.3%) and was associated with better relapse-free survival (26 versus 11 months, P = .04) and overall survival (47 versus 10 months, P = .03). Early mortality and progressive disease are the main barriers to alloHSCT, but poor follow-up, uninsured status, and lack of suitable donor, including haploidentical, are also substantial limitations that might disproportionally affect this vulnerable population. AlloHSCT can achieve long-term remissions, and strategies aiming to overcome these barriers are urgently needed to improve outcomes in ATLL.     

Comparison of community-associated meticillin-resistant <Emphasis Type="Italic">Staphylococcus aureus</Emphasis> from teaching hospitals in London and the USA, 2004–2006: where is USA300 in the UK?

Community-associated meticillin-resistant Staphylococcus aureus (CA-MRSA) is increasingly common in the USA, but rare in the UK. We compared CA-MRSA from the UK and USA to examine differences in the molecular epidemiology. We investigated patients presenting with MRSA in the first 72 h of hospital admission or in out-patient settings in a UK and a US hospital from January 2004 to March 2006. Fluoroquinolone susceptibility was used as a screening marker to select presumptive CA-MRSA. One hundred and eighteen and 49 such strains were identified, representing a prevalence of 0.1 and 0.2 isolates per 1,000 patient days in the UK and US respectively. Panton–Valentine leukocidin (PVL)-positive ST8-IVa (USA300)-type strains predominated among 43 surviving US isolates, whereas PVL-negative ST1-IVa predominated among 71 surviving UK isolates. There are striking differences between the molecular epidemiology of CA-MRSA in UK and US hospitals, which may have implications for control.