Machine learning algorithm to predict mortality in heart failure patients

Researchers from the University of California San Diego developed a machine-learning model by training a boosted decision tree algorithm on de-identified electronic health records data of 5,822 hospitalized or ambulatory patients with heart failure from the University of California San Diego.

This machine learning model is based on eight readily available variables. These include, diastolic blood pressure, creatinine, blood urea nitrogen, haemoglobin, white blood cell count, platelets, albumin, and red blood cell distribution width. This model was able to predict life expectancy in 88% of the patients.

This study is published in the European Journal of Heart Failure.

The tool was additionally tested using data from the University of California, San Francisco, and a database derived from 11 European medical centers.

With the advance of machine learning and artificial intelligence tools, large amounts of health data thanks to electronic health records, and computing power, we able to work on creating more and more accurate risk prediction tools. These tools will be commonplace in clinical practice to help with data-based decisions.

Smartphone app shown to improve medication intake compliance in heart patients

According to a research study presented at the 45th Argentine Congress of Cardiology heart patients using a smartphone app reminder are more likely to take their medication than those who receive written instructions,

This study tested the impact of a smartphone application on medication compliance. A total of 90 heart attack patients admitted to hospital were randomly allocated to the app or detailed written information (standard care). Adherence to medical treatment was measured at 90 days using the Morisky Medical Adherence Scale (MMAS-8). For those assigned to the smartphone group, the prescribed medication schedule was uploaded to the digital application, and an alarm would ring each time a pill should be taken. After taking the pills, patients confirmed it in the application. Doctors could check daily adherence using a professional digital platform linked to the patient’s smartphone.

The average age of patients in the study was 63 years and 75% were men. At 90 days, significantly more patients in the digital application group were correctly taking their pills (65%) compared to those who received standard care (21%; p<0.001). A secondary objective of the study was to examine how many patients in each group were hospitalized for another heart attack or had an unplanned visit to the doctor or emergency department. No differences between groups were found.

Overall this study shown increasing use of digital tools to monitor the treatment delivery. However this study does not prove that taking medications on time and digital monitoring would improve patient outcomes. Further research is needed in that respect.

Machine learning algorithms to speed up image biomarker analysis in heart MRI scans

According to research published in journal Circulation:Cardiaovascular imaging cardiac MRI analysis utilizing machine learning algorithms can be performed significantly faster and with similar precision compared to human experts.

In the study, researchers trained a neural network to read the cardiac MRI scans. Utilizing artificial intelligence, a scan can be analyzed in approximately four seconds compared to 13 minutes for the human reviewer. When the AI was tested for precision researchers found that there was no significant difference compared to humans.

Researchers made available the data utilized for this study at thevolumeresource.com. This resource also intends to test and validate new cardiac MRI post-processing technology and machine learning techniques.