Learning from Few Subjects with Large Amounts of Voice Monitoring - - PowerPoint PPT Presentation

Jose Javier Gonzalez Ortiz

Learning from Few Subjects with Large Amounts of Voice Monitoring Data

John V. Guttag Robert E. Hillman Daryush D. Mehta Jarrad H. Van Stan Marzyeh Ghassemi

Jose Javier Gonzalez Ortiz

• Few subjects and large amounts of data

→ Overfitting to subjects

• No obvious mapping from signal to features

→ Feature engineering is labor intensive

• Subject-level labels

→ In many cases, no good way of getting sample specific annotations

1

Challenges of Many Medical Time Series

Jose Javier Gonzalez Ortiz

• Few subjects and large amounts of data

→ Overfitting to subjects

• No obvious mapping from signal to features

→ Feature engineering is labor intensive

• Subject-level labels

→ In many cases, no good way of getting sample specific annotations

2

Challenges of Many Medical Time Series

Unsupervised feature extraction Multiple Instance Learning

Jose Javier Gonzalez Ortiz

128 x 64 128 x 64

Conv + BatchNorm + ReLU Pooling Dense Upsampling Sigmoid

3

Learning Features

• Segment signal into windows
• Compute time-frequency representation
• Unsupervised feature extraction

Jose Javier Gonzalez Ortiz 4

Raw Waveform Spectrogram Encoder Logistic Regression

• % Positive

Prediction

Per Window Per Subject Per Window Per Subject

Classification Using Multiple Instance Learning

• Logistic regression on learned features with subject labels
• Aggregate prediction using % positive windows per subject

Jose Javier Gonzalez Ortiz 5

Application: Voice Monitoring Data

• Voice disorders affect 7% of the US population
• Data collected through neck placed accelerometer

1 week = ~4 billion samples ~100

Jose Javier Gonzalez Ortiz

Previous work relied on expert designed features[1]

6

Results

AUC Accuracy

Expert LR Train 0.70 ± 0.05 0.71 ± 0.04 Test 0.68 ± 0.05 0.69 ± 0.04 Ours Train 0.73 ± 0.06 0.72 ± 0.04 Test 0.69 ± 0.07 0.70 ± 0.05

[1] Marzyeh Ghassemi et al. Learning to detect vocal hyperfunction from ambulatory neck-surface acceleration features: initial results for vocal fold nodules. IEEE Trans. Biomed Engineering

Comparable performance wi with thout t task-specific feature engineering!

Jose Javier Gonzalez Ortiz

• Our method learns features from large time series data
• Reduces the need for laborious task-specific feature engineering
• Applied to large voice monitoring dataset
• Comparable performance to previous work that relied on

expert engineered features

7

Summary