Artificial intelligence using a deep learning versus expert computed tomography human reading in calcium score and coronary artery calcium data and reporting system classification.

BACKGROUND: Artificial intelligence (AI) applied to cardiac imaging may provide improved processing, reading precision and advantages of automation. Coronary artery calcium (CAC) score testing is a standard stratification tool that is rapid and highly reproducible. We analyzed CAC results of 100 studies in order to determine the accuracy and correlation between the AI software (Coreline AVIEW, Seoul, South Korea) and expert level-3 computed tomography (CT) human CAC interpretation and its performance when coronary artery disease data and reporting system (coronary artery calcium data and reporting system) classification is applied. METHODS: A total of 100 non-contrast calcium score images were selected by blinded randomization and processed with the AI software versus human level-3 CT reading. The results were compared and the Pearson correlation index was calculated. The CAC-DRS classification system was applied, and the cause of category reclassification was determined using an anatomical qualitative description by the readers. RESULTS: The mean age was age 64.5 years, with 48% female. The absolute CAC scores between AI versus human reading demonstrated a highly significant correlation (Pearson coefficient R  = 0.996); however, despite these minimal CAC score differences, 14% of the patients had their CAC-DRS category reclassified. The main source of reclassification was observed in CAC-DRS 0-1, where 13 were recategorized, particularly between studies having a CAC Agatston score of 0 versus 1. Qualitative description of the errors showed that the main cause of misclassification was AI underestimation of right coronary calcium, AI overestimation of right ventricle densities and human underestimation of right coronary artery calcium. CONCLUSION: Correlation between AI and human values is excellent with absolute numbers. When the CAC-DRS classification system was adopted, there was a strong correlation in the respective categories. Misclassified were predominantly in the category of CAC = 0, most often with minimal values of calcium volume. Additional algorithm optimization with enhanced sensitivity and specificity for low values of calcium volume will be required to enhance AI CAC score utilization for minimal disease. Over a broad range of calcium scores, AI software for calcium scoring had an excellent correlation compared to human expert reading and in rare cases determined calcium missed by human interpretation.

Anonymity: an impediment to performance in healthcare.

Many teaching hospitals employ a care team structure composed of a broad range of healthcare providers with different skill sets. Each member of this team has a distinct role and a different level of training ranging from attending physician to resident, intern, and medical student. Often times, these different roles lead to greater complexity and confusion for both patients and nursing staff. It has been demonstrated that patients have a great degree of difficulty in identifying members of their care team. This anonymity also exists between nursing staff and other care providers. In order to better understand the magnitude of anonymity within the teaching hospital, a ten-question survey was sent to nurses across three different departments. Results from this survey demonstrated that 71% of nurses are "Always" or "Often" able to identify which care team is responsible for their patients, while 79% of nurses reported that they either "Often" or "Sometimes" page a provider who is not currently caring for a given patient. Furthermore, 33% of nurses felt that they were either "Rarely" or "Never" able to recognize, by face and name, attending level providers. Residents were "Rarely" or "Never" recognized by face and name 37% of the time, and interns 42% of the time. Contacting the wrong provider repeatedly leads to de facto delays in medication, therapy, and diagnosis. Additionally, these unnecessary interruptions slow workflow for both nurses and members of the care team, making hospital care less efficient and safe overall. Technological systems should focus on reducing anonymity within the hospital in order to enhance healthcare delivery.

Quadricuspid aortic valve defined by echocardiography and cardiac computed tomography.

A 54 year old female presented with lower extremity edema, fatigue, and shortness of breath with physical findings indicative of advanced aortic insufficiency. Echocardiography showed severe aortic regurgitation and a probable quadricuspid aortic valve. In anticipation of aortic valve replacement, cardiac computed tomography (Cardiac CT) was performed using 100 kV, 420 mA which resulted in 6 mSv of radiation exposure. Advanced computing algorithmic software was performed with a non-linear interpolation to estimate potential physiological movement. Surgical photographs and in-vitro anatomic pathology exam reveal the accuracy and precision that preoperative Cardiac CT provided in this rare case of a quadricuspid aortic valve. While there have been isolated reports of quadricuspid diagnosis with Cardiac CT, we report the correlation between echocardiography, Cardiac CT, and similar appearance at surgery with confirmed pathology and interesting post-processed rendered images. Cardiac CT may be an alternative to invasive coronary angiography for non-coronary cardiothoracic surgery with the advantage of providing detailed morphological dynamic imaging and the ability to define the coronary arteries non-invasively. The reduced noise and striking depiction of the valve motion with advanced algorithms will require validation studies to determine its role.