Speech analysis phone app can detect early signs of hepatic encephalopathy

A study conducted at the University of Michigan suggests that one day a phone app will be able to detect the presence of hepatic encephalopathy in patients simply by listening to their speech.

This is stated in a paper published in the journal Hepatology.

Liver encephalopathy, a brain disorder caused by the liver’s inability to filter toxins from the blood, often affects the quality of life of patients with cirrhosis.

Minimal hepatic encephalopathy, a less severe form of the disease, can be a precursor to overt hepatic encephalopathy, which can lead to complications such as coma.

Minimal hepatic encephalopathy may be accompanied by confusion, sleep disturbance and personality changes, as well as other symptoms.

However, despite its impact on quality of life, the minimal form can be difficult to detect.

Between 10% and 20% of patients with cirrhosis may have minimal hepatic encephalopathy, but the physician will not know this because a psychometric test for hepatic encephalopathy is needed to make the diagnosis. However, this never happens in clinical practice, so such people are not diagnosed.

Liver encephalopathy often affects speech as well.

Hoping to create a simpler test, the researchers recorded the performance of speech tasks by 200 patients with cirrhosis.

The recordings were then analyzed to identify hundreds of speech variables.

The analysis resulted in a multivariate model called the Hear-MHE, which was compared to validated psychometric tests to determine its ability to detect moderate hepatic encephalopathy.

The Hear-MHE model was highly comparable to validated tests and demonstrated its ability to predict time to development of overt hepatic encephalopathy.

Speech tasks included reading a text, describing a picture, and a test to determine the name of an animal.

The text-reading test was the most effective and relatively simple – patients recorded reading for 30-40 seconds, instead of the 15-20 minute pen-and-paper test – leading researchers to hope that simplified testing may one day lead to more diagnoses and better treatment for patients who have been diagnosed earlier.

An additional benefit of the text reading test is its uniformity, making it easier to compare one patient’s tests to others or to make comparisons between patients.

The researchers’ next steps involve evaluating the tool in a larger and more diverse clinical setting to confirm its effectiveness in detecting hepatic encephalopathy.

The researchers also hope to develop a home version of the web-based application that can be sent to patients via text message.