UK develops robots with realistic expressions that cut doctors' errors and bias
Robots
Scientists in the UK have developed a tool that enables robots to display more realistic and accurate expressions of pain during the physical examination of painful areas.
The Imperial College London team has developed a new physical examination approach that could reduce doctors' errors and bias.
Research published in the journal Scientific Reports suggests that this could also teach trainee doctors how to use clues hidden in patients' facial expressions to minimise the force used in physical exams, while also detecting and correcting early signs of bias in medical students by exposing them to a wide variety of patient identities.
"Improving the accuracy of facial expressions of pain on these robots is a key step in improving the quality of physical examination training for medical students," said Sibylle Rerolle, from Imperial's Dyson School of Design Engineering.
Undergraduate students participated in the study by performing a physical exam on a robotic patient's abdomen. In order to recreate pain-related facial expressions, data about forces applied to the abdomen were used to trigger changes in six different regions of the MorphFace robot face.
The method revealed the order in which different parts of a robotic face, known as facial activation units (AUs), must fire to accurately express pain. Additionally, the study determined the optimum speed and magnitude of AU activation.
Also read |Even a worm can make decisions, says a study
Researchers discovered that the most realistic facial expressions were created by activating the upper face AUs (around the eyes) before activating the lower face AUs (around the mouth). The most natural results were obtained by allowing the Jaw Drop AU to take longer to activate.
It is crucial for doctors to consider patient facial expressions when performing physical examinations of painful areas.
Most medical training simulators, however, cannot show real-time facial expressions related to pain and only include a limited number of patient identities according to ethnicity and gender.
Based on studies demonstrating racial bias in the ability to recognize facial expressions of pain, the researchers state that these limitations may lead to biased practices on the part of medical students.
"Underlying biases could lead doctors to misinterpret the discomfort of patients - increasing the risk of mistreatment, negatively impacting doctor-patient trust, and even causing mortality, a said co-author Thilina Lalitharatne, from the Dyson School of Design Engineering.
"In the future, a robot-assisted approach could be used to train medical students to normalise their perceptions of pain expressed by patients of different ethnicity and gender."
(With inputs from agencies)
