Automatic Hierarchical Attention Neural Network for Detecting AD

Pan, Yilin; Mirheidari, Bahman; Reuber, Markus; Venneri, Annalena; Blackburn, Daniel; Christensen, Heidi

doi:10.21437/interspeech.2019-1799

Cited by 32 publications

(46 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another recent attempt to address the use of neural networks for detecting dementia is the recent work of Pan et al [49]. They have proposed a hierarchical bidirectional neural network induced with an attention mechanism for extracting different levels of features.…”

Section: Related Workmentioning

confidence: 99%

Predicting Prodromal Dementia Using Linguistic Patterns and Deficits

Alkenani

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Language deficiency is evident in the onset of several neurodegenerative disorders yet has barely been investigated when first occurs on the continuum of cognitive impairment for the purpose of early diagnoses. Alzheimer's disease (AD) is a neurodegenerative pathology that develops years prior to clinical manifestations and typically preceded by prodromal stages such as Mild Cognitive Impairment (MCI). Currently, the manual diagnostic procedures of both types are time consuming, following certain clinical criteria and neuropsychological examinations. Our study aims to establish state-of-the-art performance in the automatic identification of different dementia etiologies, including AD, MCI, and Possible AD (PoAD), and to determine whether patients with initial cognitive declines exhibit language deficits through the analysis of language samples deduced with the cookie theft picture description task. Data were derived from the cookie theft picture corpus of DementiaBank, from which all language samples of the identified etiologies were used, with a random subsampling technique that handles the skewness of the classes. Several original lexical and syntactic (i.e., lexicosyntactic) features were introduced and used alongside previously established lexicosyntactics to train machine learning (ML) classifiers against these etiologies. Further, a statistical analysis was conducted to uncover the deficiency across these etiologies. Our models resulted in benchmarks for differentiating all the identified classes with accuracies ranging between 95 to 98% and corresponding F1 values falling between 94 and 98%. The statistical analysis of our lexicosyntactic biomarkers shows that linguistic deviations are associated with prodromal as well as advanced neurodegenerative pathologies, being greatly impacted as cognitive decline increases and suggesting that language biomarkers may aid the early diagnosis of these pathologies.

show abstract

Section: Related Workmentioning

confidence: 99%

Predicting Prodromal Dementia Using Linguistic Patterns and Deficits

Alkenani

et al. 2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Recently, automatic approaches to analysing a person's speech and language have gained traction. Language-based analysis is mostly carried out on either the manual or automatic transcripts [4,5], whereas speechbased analysis would normally be based on the acoustic signal [6][7][8][9][10][11]. In both cases, the performance of a typical classification pipeline is highly dependent on the quality of the front-end features.…”

Section: Introductionmentioning

confidence: 99%

“…In [20], it was found that combining CNNs, LSTMs, and DNNs for speech processing in a unified architecture allowed for the exploitation of their complementary natures. The attention mechanism has lately been used in different fields and achieved a great deal of success [4,21,22]. The main idea behind the attention mechanism is to apply a higher attention weight to the more critical parts of the input for classification.…”

Section: Introductionmentioning

confidence: 99%

Acoustic Feature Extraction with Interpretable Deep Neural Network for Neurodegenerative Related Disorder Classification

Pan¹,

Mirheidari²,

Tu³

et al. 2020

Interspeech 2020

Self Cite

View full text Add to dashboard Cite

Speech-based automatic approaches for detecting neurodegenerative disorders (ND) and mild cognitive impairment (MCI) have received more attention recently due to being non-invasive and potentially more sensitive than current pen-and-paper tests. The performance of such systems is highly dependent on the choice of features in the classification pipeline. In particular for acoustic features, arriving at a consensus for a best feature set has proven challenging. This paper explores using deep neural network for extracting features directly from the speech signal as a solution to this. Compared with hand-crafted features, more information is present in the raw waveform, but the feature extraction process becomes more complex and less interpretable which is often undesirable in medical domains. Using a Sinc-Net as a first layer allows for some analysis of learned features. We propose and evaluate the Sinc-CLA (with SincNet, Convolutional, Long Short-Term Memory and Attention layers) as a task-driven acoustic feature extractor for classifying MCI, ND and healthy controls (HC). Experiments are carried out on an inhouse dataset. Compared with the popular hand-crafted feature sets, the learned task-driven features achieve a superior classification accuracy. The filters of the SincNet is inspected and acoustic differences between HC, MCI and ND are found.

show abstract

“…Further improvements are seen when combining with the rhythm-inspired features. (3) Applying the proposed technique to our linguistic-based system proposed in [22] also improves that system. (4) Finally, combining our acoustic and linguistic systems achieves a further improvement.…”

Section: Introductionmentioning

confidence: 99%

“…Among them, RNNs are good at capturing the temporal evolution of input signals and model the sequence information [23], which is suitable for the acoustic and rhythm feature modelling. The attention mechanism has lately been used in different fields and achieved a great deal of success [22,24]. The main idea behind the attention mechanism is applying a higher attention weight to the more critical parts of the input for classification.…”

Section: Introductionmentioning

confidence: 99%

Improving Detection of Alzheimer’s Disease Using Automatic Speech Recognition to Identify High-Quality Segments for More Robust Feature Extraction

Pan¹,

Mirheidari²,

Reuber³

et al. 2020

Interspeech 2020

Self Cite

View full text Add to dashboard Cite

Speech and language based automatic dementia detection is of interest due to it being non-invasive, low-cost and potentially able to aid diagnosis accuracy. The collected data are mostly audio recordings of spoken language and these can be used directly for acoustic-based analysis. To extract linguistic-based information, an automatic speech recognition (ASR) system is used to generate transcriptions. However, the extraction of reliable acoustic features is difficult when the acoustic quality of the data is poor as is the case with DementiaBank, the largest opensource dataset for Alzheimer's Disease classification. In this paper, we explore how to improve the robustness of the acoustic feature extraction by using time alignment information and confidence scores from the ASR system to identify audio segments of good quality. In addition, we design rhythminspired features and combine them with acoustic features. By classifying the combined features with a bidirectional-LSTM attention network, the F-measure improves from 62.15% to 70.75% when only the high-quality segments are used. Finally, we apply the same approach to our previously proposed hierarchical-based network using linguistic-based features and show improvement from 74.37% to 77.25%. By combining the acoustic and linguistic systems, a state-of-the-art 78.34% F-measure is achieved on the DementiaBank task.

show abstract

Automatic Hierarchical Attention Neural Network for Detecting AD

Cited by 32 publications

References 17 publications

Predicting Prodromal Dementia Using Linguistic Patterns and Deficits

Predicting Prodromal Dementia Using Linguistic Patterns and Deficits

Acoustic Feature Extraction with Interpretable Deep Neural Network for Neurodegenerative Related Disorder Classification

Improving Detection of Alzheimer’s Disease Using Automatic Speech Recognition to Identify High-Quality Segments for More Robust Feature Extraction

Contact Info

Product

Resources

About