J. Michael DiMaio

Contemporary extracorporeal membrane oxygenation therapy in adults: Fundamental principles and systematic review of the evidence.

Extracorporeal membrane oxygenation (ECMO) provides days to weeks of support for patients with respiratory, cardiac, or combined cardiopulmonary failure. Since ECMO was first reported in 1974, nearly 70,000 runs of ECMO have been implemented, and the use of ECMO in adults increased by more than 400% from 2006 to 2011 in the United States. A variety of factors, including the 2009 influenza A epidemic, results from recent clinical trials, and improvements in ECMO technology, have motivated this increased use in adults. Because ECMO is increasingly becoming available to a diverse population of critically ill patients, we provide an overview of its fundamental principles and a systematic review of the evidence basis of this treatment modality for a variety of indications in adults.

Outlier detection and removal improves accuracy of machine learning approach to multispectral burn diagnostic imaging.

Abstract. Multispectral imaging (MSI) was implemented to develop a burn tissue classification device to assist burn surgeons in planning and performing debridement surgery. To build a classification model via machine learning, training data accurately representing the burn tissue was needed, but assigning raw MSI data to appropriate tissue classes is prone to error. We hypothesized that removing outliers from the training dataset would improve classification accuracy. A swine burn model was developed to build an MSI training database and study an algorithm’s burn tissue classification abilities. After the ground-truth database was generated, we developed a multistage method based on Z-test and univariate analysis to detect and remove outliers from the training dataset. Using 10-fold cross validation, we compared the algorithm’s accuracy when trained with and without the presence of outliers. The outlier detection and removal method reduced the variance of the training data. Test accuracy was improved from 63% to 76%, matching the accuracy of clinical judgment of expert burn surgeons, the current gold standard in burn injury assessment. Given that there are few surgeons and facilities specializing in burn care, this technology may improve the standard of burn care for patients without access to specialized facilities.

Burn injury diagnostic imaging device's accuracy improved by outlier detection and removal

Multispectral imaging (MSI) was implemented to develop a burn diagnostic device that will assist burn surgeons in planning and performing burn debridement surgery by classifying burn tissue. In order to build a burn classification model, training data that accurately represents the burn tissue is needed. Acquiring accurate training data is difficult, in part because the labeling of raw MSI data to the appropriate tissue classes is prone to errors. We hypothesized that these difficulties could be surmounted by removing outliers from the training dataset, leading to an improvement in the classification accuracy. A swine burn model was developed to build an initial MSI training database and study an algorithm’s ability to classify clinically important tissues present in a burn injury. Once the ground-truth database was generated from the swine images, we then developed a multi-stage method based on Z-test and univariate analysis to detect and remove outliers from the training dataset. Using 10-fold cross validation, we compared the algorithm’s accuracy when trained with and without the presence of outliers. The outlier detection and removal method reduced the variance of the training data from wavelength space, and test accuracy was improved from 63% to 76%. Establishing this simple method of conditioning for the training data improved the accuracy of the algorithm to match the current standard of care in burn injury assessment. Given that there are few burn surgeons and burn care facilities in the United States, this technology is expected to improve the standard of burn care for burn patients with less access to specialized facilities.

Human Factors and Human Nature in Cardiothoracic Surgery

t 7:34 A.M. on September 11, 1974, Eastern Air Lines AFlight 212 from Charleston, SC, crashed in an open field 3.3 miles short of runway 36 at Douglas Municipal Airport in Charlotte, NC [1]. There was little or no wind, and the visibility was limited due to patchy dense ground fog. Of the 82 people on board, 11 survived. Notably, 5 flights preceded Flight 212 onto runway 36 without difficulty that morning. Partly based on the cockpit voice recorder, the National Transportation Safety Board determined that the likely cause of the crash was “the flight crew’s lack of altitude awareness at critical points during the approach due to poor cockpit discipline in that the crew did not follow prescribed procedures” [1]. Specific issues with discipline and prescribed procedures were as follows: “During the descent, until about 2 minutes and 30 seconds prior to the sound of impact, the flight crew engaged in conversations . . . (that) covered a number of subjects, from politics to used cars, and both crew members expressed strong views and mild aggravation concerning the subjects discussed. The Safety Board believes that these conversations were distractive and reflected a casual mood and lax cockpit atmosphere, which continued throughout the remainder of the approach and which contributed to the accident” [1]. In 1981, in response to aviation accidents, the Federal Aviation Administration imposed the “Sterile Cockpit Rule,” which states that pilots are to refrain from nonessential activities or conversations that could distract or interfere with their duties during critical phases of flight and operations below 10,000 feet [2]. Surgical errors and adverse events include wrong or delayed operations and judgment lapses that lead to incorrect procedures [3–7]. It is estimated that 54% of the adverse events in patients undergoing operations surgery are preventable [7]. In patients undergoing coronary artery bypass grafting, for whom the risk-adjusted mortality rate ranges from 1.3% to 3.1%, approximately one-third of associated deaths may be preventable, with most occurring in the operating room and intensive care unit [6]. Surgical outcomes are often attributed primarily to the technical skills of the surgeon: when errors are made, the surgeon’s competence is questioned [3, 4, 8–10]. The notion that the surgeon is often held solely accountable is

Multispectral imaging burn wound tissue classification system: a comparison of test accuracies between several common machine learning algorithms

The clinical judgment of expert burn surgeons is currently the standard on which diagnostic and therapeutic decisionmaking regarding burn injuries is based. Multispectral imaging (MSI) has the potential to increase the accuracy of burn depth assessment and the intraoperative identification of viable wound bed during surgical debridement of burn injuries. A highly accurate classification model must be developed using machine-learning techniques in order to translate MSI data into clinically-relevant information. An animal burn model was developed to build an MSI training database and to study the burn tissue classification ability of several models trained via common machine-learning algorithms. The algorithms tested, from least to most complex, were: K-nearest neighbors (KNN), decision tree (DT), linear discriminant analysis (LDA), weighted linear discriminant analysis (W-LDA), quadratic discriminant analysis (QDA), ensemble linear discriminant analysis (EN-LDA), ensemble K-nearest neighbors (EN-KNN), and ensemble decision tree (EN-DT). After the ground-truth database of six tissue types (healthy skin, wound bed, blood, hyperemia, partial injury, full injury) was generated by histopathological analysis, we used 10-fold cross validation to compare the algorithms’ performances based on their accuracies in classifying data against the ground truth, and each algorithm was tested 100 times. The mean test accuracy of the algorithms were KNN 68.3%, DT 61.5%, LDA 70.5%, W-LDA 68.1%, QDA 68.9%, EN-LDA 56.8%, EN-KNN 49.7%, and EN-DT 36.5%. LDA had the highest test accuracy, reflecting the bias-variance tradeoff over the range of complexities inherent to the algorithms tested. Several algorithms were able to match the current standard in burn tissue classification, the clinical judgment of expert burn surgeons. These results will guide further development of an MSI burn tissue classification system. Given that there are few surgeons and facilities specializing in burn care, this technology may improve the standard of burn care for patients without access to specialized facilities.

Physician Burnout: Are We Treating the Symptoms Instead of the Disease?

Despite increasing recognition of physician burnout, its incidence has only increased in recent years, with nearly half of physicians suffering from symptoms of burnout in the most recent surveys. Unfortunately, most burnout research has focused on its profound prevalence rather than seeking to identify the root cause of the burnout epidemic. Health care organizations throughout the United States are implementing committees and support groups in an attempt to reduce burnout among their physicians, but these efforts are typically focused on increasing resilience and wellness among participants rather than combating problematic changes in how medicine is practiced by physicians in the current era. This report provides a brief review of the current literature on the syndrome of burnout, a summary of several institutional approaches to combating burnout, and a call for a shift in the focus of these efforts toward one proposed root cause of burnout.

Implantation of Transcatheter Aortic Prosthesis in 3 Patients With Mitral Annular Calcification

Mitral annular calcification (MAC) is a chronic degenerative process at the fibrous base of the mitral valve. It is a feared diagnosis in the context of mitral valve operations because of the risk of severe adverse events such as atrioventricular disruption, injury to the circumflex artery during debridement, and difficult placement of annular sutures. We report a series of 3 consecutive female patients with severe circular MAC who underwent successful mitral valve replacement through a lateral minithoracotomy with use of an inverted transcatheter aortic valve.

Systematic review of transcatheter aortic valve replacement after previous mitral valve surgery

From the Baylor Scott & White Research Institute, Departments of Cardiology, and Cardiothoracic Surgery, The Heart Hospital Baylor Plano, Plano; and Department of General Surgery, Baylor University Medical Center, Dallas, Tex. Disclosures: Authors have nothing to disclose with regard to commercial support. Received for publication May 12, 2017; revisions received Aug 11, 2017; accepted for publication Aug 24, 2017; available ahead of print Oct 5, 2017. Address for reprints: J. Michael DiMaio, MD, Baylor Scott & White Research Institute, The Heart Hospital Baylor Plano, Plano, TX 75093 (E-mail: jmdimaio@yahoo.com). J Thorac Cardiovasc Surg 2018;155:63-5 0022-5223/

Quantifying regional left ventricular contractile function: Leave it to the machines?

36.00 Copyright 2017 by The American Association for Thoracic Surgery https://doi.org/10.1016/j.jtcvs.2017.08.129 Reported combinations of transcatheter aortic valve replacement valves implanted with previous mitral prostheses.

Cerebral protection during deep hypothermic circulatory arrest: Can a molecular approach via microRNA inhibition improve on a millennia-old strategy?

In this issue of the Journal, Henn and colleagues from Washington University in St Louis present a novel method to quantify and localize regional left ventricle (LV) contractile function in coronary artery disease by means of cardiac magnetic resonance imaging with radiofrequency tissue tagging. They have developed a mechanism by which the quantified LV function can be compared against a normalized standard to determine the presence and severity of an individual patient’s pathologic contractile dysfunction. Henn and colleagues deserve praise for developing an automated, quantitative solution for a problem that has only been addressed qualitatively in the past. The ability to both quantify and localize LV contractile dysfunction should improve clinical outcomes in several ways by addressing limitations of the current nonquantitative metrics of regional LV function. First, this method creates an objective standard that will increase consistency and accuracy across the board by reducing the interuser and intertemporal variability that has plagued echocardiographic interpretation of LV function in the past. Furthermore, the simplicity of interpreting and displaying these results may increase patients’ understanding of their coronary artery disease. A key component of this study is the methodology to extrapolate a single measurement of cardiac wall motion (z score) from detailed and multidimensional data and to construct a reference range model by ‘‘normalizing’’ the LV function z score of a healthy cohort, a process similar to what is known as machine learning in engineering circles. Machine learning typifies a ubiquitous engineering approach to solving complex biological questions—and its application is only increasing in frequency within the biomedical realm. Machine learning takes advantage of computer models that can be taught to perform a desired task. Training data are collected to develop and