TREMOR AI

iPhone / Medecine

Purpose

To quantify the degree of tremor in patients with tremor or cerebellar ataxia. This allows you to follow the changes of symptoms due to treatment quantitatively.
Artificial intelligence (AI) predicts whether the type of tremor is close to one of the three types: cerebellar, tremor, or normal.


Method of operation

Print the specified PDF file on A4 paper, and trace the spiral from the outside in red (such as a sign pen).
Launch this app and take a photo with your smartphone.


About the results

"Length" indicates the ratio between the printed spiral and the handwriting red pen spiral length. If it is 105% or more, it is estimated to be abnormal, but up to 110% is estimated to be within the normal range for elderly people.
"Deviation" indicates the area of the “deviation” between the printed spiral and the handwriting red pen spiral. If it is 1000mm2 or more, it is estimated to be abnormal, but it is estimated that up to 1500mm2 is within the normal range for elderly people.

Probability of speculated diagnosis

* Cerebellar type (CD): tremor type caused by cerebellar ataxia.
* Tremor type (ET): postural tremor type including essential tremor and enlarged physiological tremor.
* Normal type (NL): within normal range.

The above three probabilities are displayed. Please note that the probability of diagnosis may vary depending on the device used.
Because the accuracy of AI diagnosis remains at a low level of 70-80%, AI diagnosis is just one of assistant tools. Please consider visiting a medical institution or introducing to a specialist if necessary.
For all of the above results, the creator of this app is not responsible for anything.

Quoi de neuf dans la dernière version ?

Updated the learning model to improve the judgment accuracy.