Machine Learning in Medicine Weekly Roundup #2

Machine learning is revolutionizing healthcare. Here are some highlights of interesting applications of how research in artificial intelligence is helping people.

1) Helping people quit smoking using intelligent AI support messages

Quitting smoking is hard. Very hard. Habits are best changed in a group-support type setting. What these researchers accomplished was to create an AI based messaging platform that sends supportive messages to help people quit smoking.

Read more: Impact of a Collective Intelligence Tailored Messaging System on Smoking Cessation: The Perspect Randomized Experiment

2) Learning walking patterns and disease

The way we move is a valuable insight to health and illness. People suffering from a stroke or parkinson for example have a certain pattern of moving. But what if we could detect or monitor the progression of an illness using data from how a person moves? These researchers describes methods to improve the data that can be used to help computers identify the way people move.

Read more here: A data augmentation methodology for training machine/deep learning gait recognition algorithms

3) Identifying suicide risk based on the sound of your voice

The trained human ear and brain is exquisitely good at picking up cues on emotion. And yet, health care professionals are still unable to predict when someone will commit suicide. In this study, researchers apply machine learning principles to manage audio and text data from patients to identify those at risk of being suicidal, mentally ill but not suicidal, or a control.

Read more here: A Machine Learning Approach to Identifying the Thought Markers of Suicidal Subjects: A Prospective Multicenter Trial