Reference stimulus
A stimulation, biological technology, applied in the field of signal processing, can solve the problem of not reaching the measurement of pain and so on
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0248] (Example 1: Closed-eye sample magnification)
[0249] In this embodiment, an experiment with closed-eye sample magnification (Long short-term memory (LSTM) 4 class) is performed. The method and the like are shown below.
[0250] (method 1)
[0251] The 4-level discrimination of "no pain / pain / noise and no pain / noise and pain" was performed using LSTM. Noise tests and noise tests in pain stimuli were performed for the labeling of classes containing noise. The subjects were asked to close their eyes for a portion of the trial (eye-closure task). The conditions under which the experiments were performed are shown below.
[0252] ·Experimental trial:
[0253] (1) artifact1: noise test (close the eyes strongly, stretch the body, read aloud), open the eyes
[0254] (2) artifact2: noise test (close the eyes strongly, stretch the body, read aloud), open the eyes
[0255] (3) artifact_pain1: Noise test (spontaneous response to noise entering), eye opening
[0256] (4) art...
Embodiment 2
[0327] (Example 2: Comparison of 4-level and 2-level LSTMs)
[0328] This example compares 4-level and 2-level LSTMs.
[0329] (method)
[0330] In the positive and negative 2-level classification problem, the classification is performed as follows based on the predicted result of the classifier and the actual result. For example, let TP (TruePositive) be the number of data that is actually positive and the prediction result is also positive, and TN (TrueNegative) that is the number of data that is actually negative and the prediction result is also negative The number of data that is actually negative and the prediction result is positive is FP (FalsePositive), and the number of data that is actually positive and the prediction result is negative is FN (FalseNegative) .
[0331] (Evaluation Criteria)
[0332] Hereinafter, four evaluation criteria are defined as follows ( Figure 22 ).
[0333] Positive solution rate (accuracy): Among the data predicted to be positive or...
Embodiment 3
[0347] (Example 3: 2-level LSTM parsing)
[0348] In this example, 2-level LSTM parsing is performed. Figure 28 The flow of 2-level LSTM parsing is shown.
[0349] (result)
[0350] The results are shown below.
[0351] Figure 29 Raw data for the artifact1 (noise test (eyes tightly closed, body stretching, reading aloud), eyes open) conditions are shown. Figure 30 The following results of offline time-series data analysis are shown. In level 2, it should be judged to be no pain (0), but it was misjudged.
[0352] Figure 31 Raw data for the artifact2 (noise test (eyes tightly closed, body stretching, reading aloud), eyes open) conditions are shown. Figure 32 Offline time series data analysis for modeling is shown. Since it is grade 2, it should be judged as having no pain (0), but it was erroneously judged as having pain.
[0353] Figure 33 Raw data is shown for the condition artifact_pain1 (noise test (spontaneous response to noise entry), eyes open) on painful...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


