Method for estimating arrangement of electrodes on biological tissue

Patent metadata

Field	Value
Publication number	US-12502143-B2
Application number	US-202017424967-A
Country	US
Kind code	B2
Filing date	Jan 24, 2020
Priority date	Jan 25, 2019
Publication date	Dec 23, 2025
Grant date	Dec 23, 2025

Abstract

A method for estimating an arrangement of electrodes obtains detection signals when an initial electrode array having a sufficient number of electrodes for detection of excitation wave arranged and arrayed in a plane is attached to biological tissue. The method uses a plurality of input data based on detection signals obtained in a plurality of second electrode arrays generated by eliminating a predetermined number of electrodes at random from the initial electrode array, and uses an image of excitation wave in a process of obtaining the detection signals by using the initial electrode array, as teacher data, obtaining a learned model by deep learning. The method selects a second electrode array corresponding to an analysis image that is best matched with the image of the teacher data, among a plurality of analysis images obtained by applying the plurality of input data to the learned model, as a selective electrode array.

First claim

The invention claimed is:

1 . A method for optimizing an arrangement of electrodes on biological tissue to detect excitation waves, the method comprising:
a signal obtaining step of
setting a first electrode array, arranged and arrayed on a plane, the first electrode array having a plurality of electrodes such that the excitation wave can be detected,
with the electrodes being attached to the biological tissue, obtaining first detection signals from respective electrodes of the plurality of electrodes of the first electrode array,
generating a first moving image based on the first detection signals,
setting a plurality of second electrode arrays that are generated by eliminating a predetermined number of electrodes at random from the first electrode array,
with the electrodes being attached to the biological tissue, obtaining a plurality of second detection signals from respective electrodes of each of respective second electrode arrays, and
generating a plurality of second moving images based respectively on each of the plurality of second detection signals,
a learned model obtaining step of obtaining a learned model by
using the plurality of second moving images as input data,
using the first moving image as teacher data, and
learning a relationship between the input data and the teacher data by deep learning;
an electrode array selecting step of selecting a second electrode array from the plurality of second electrode arrays as a selective electrode array by
applying the plurality of second moving images to the learned model to generate a plurality of analysis images;
matching an analysis image among the plurality of analysis images to the first moving image, and
selecting the selective electrode array as the second electrode array corresponding to the matched analysis image;
an electrode arrangement estimating step of estimating a number of and an arrangement of electrodes in a third electrode array based on a number of and an arrangement of the electrodes in the selective electrode array; and implementing the third electrode array on the biological tissue to detect the excitation waves of the biological tissue.

2 . The method for optimizing the arrangement of respective electrodes of the plurality of second electrode arrays on the biological tissue according to claim 1 , wherein
the input data is adjusted, based on interpolation signals between electrodes obtained by interpolating detection signals at the electrodes by a nearest neighbor algorithm.

3 . The method for optimizing the arrangement of the electrodes on the biological tissue according to claim 1 , wherein
the learned model obtaining step and the electrode array selecting step are performed repeatedly by using the selective electrode array selected by the electrode array selecting step, as the first electrode array.

4 . The method for optimizing the arrangement of the electrodes on the biological tissue according to claim 3 , wherein
when an analysis image obtained by the selective electrode array is out of a predetermined range with respect to the image of the teacher data, the electrode arrangement estimating step specifies the selective electrode array selected in a previous cycle of the electrode array selecting step, as the selective electrode array.

5 . The method for optimizing the arrangement of the electrodes on the biological tissue according to claim 2 , wherein
the learned model obtaining step and the electrode array selecting step are performed repeatedly by using the selective electrode array selected by the electrode array selecting step, as the first electrode array.

6 . The method for optimizing the arrangement of the electrodes on the biological tissue according to claim 5 , wherein
when an analysis image obtained by the selective electrode array is out of a predetermined range with respect to the image of the teacher data, the electrode arrangement estimating step specifies the selective electrode array selected in a previous cycle of the electrode array selecting step, as the selective electrode array.

Assignees

• Univ Tokyo

Inventors

• Tomii Naoki
• Sakuma Ichiro
• Yamazaki Masatoshi
• Jiao Jiaming

Classifications

• A61B2562/04

  Arrangements of multiple sensors of the same type · CPC title

• A61B5/684

  Indicating the position of the sensor on the body · CPC title

• A61B2560/0223

  of calibration, e.g. protocols for calibrating sensors · CPC title

• A61B5/7264

  Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems · CPC title

• A61B5/7278 Primary

  Artificial waveform generation or derivation, e.g. synthesizing signals from measured signals · CPC title