Performance Evaluation of Machine Learning Frameworks for Aphasia Assessment
Open accessThe DOI matches its Crossref record; 36 of the 40 references it cites were found in a registry or an index; 6 of the 32 sources its claims cite were checked for support.
Abstract
Speech assessment is an essential part of the rehabilitation procedure for patients with aphasia (PWA). It is a comprehensive and time-consuming process that aims to discriminate between healthy individuals and aphasic patients, determine the type of aphasia syndrome, and determine the patients’ impairment severity levels (these are referred to here as aphasia assessment tasks). Hence, the automation of aphasia assessment tasks is essential. In this study, the performance of three automatic speech assessment models based on the speech dataset-type was investigated. Three types of datasets were used: healthy subjects’ dataset, aphasic patients’ dataset, and a combination of healthy and aphasic datasets. Two machine learning (ML)-based frameworks, classical machine learning (CML) and deep neural network (DNN), were considered in the design of the proposed speech assessment models. In this paper, the DNN-based framework was based on a convolutional neural network (CNN). Direct or indirect transformation of these models to achieve the aphasia assessment tasks was investigated. Comparative performance results for each of the speech assessment models showed that quadrature-based high-resolution time-frequency images with a CNN framework outperformed all the CML frameworks over the three dataset-types. The CNN-based framework reported an accuracy of 99.23 ± 0.003% with the healthy individuals’ dataset and 67.78 ± 0.047% with the aphasic patients’ dataset. Moreover, direct or transformed relationships between the proposed speech assessment models and the aphasia assessment tasks are attainable, given a suitable dataset-type, a reasonably sized dataset, and appropriate decision logic in the ML framework.
Overview
CoChat Summary
TL;DR
This paper compares classical machine learning (CML) and a CNN-based deep neural network framework for three aphasia-related speech assessment models built from different dataset types: healthy, aphasic, and combined healthy-aphasic speech. Across all three models, the ResNet-34 CNN with high-resolution time-frequency images outperformed the CML classifiers, with the best reported accuracy of 99.23 ± 0.003% on healthy speech and 67.78 ± 0.047% on aphasic speech.
Key findings
- On the healthy dataset (Model-A), the ResNet-34 CNN with HTD time-frequency images outperformed all CML classifiers; it achieved 99.23 ± 0.003% accuracy on the only-words dataset, versus 95.28 ± 0.79% for the best CML method (LDA). For the larger 26-class dataset, CNN reached 97.39 ± 0.004%, and all classifiers except kNN exceeded 90% accuracy.
- On the aphasic dataset (Model-B), CNN again outperformed the CML methods, but overall performance was much lower; CNN achieved 67.78 ± 0.047% accuracy on the only-words dataset, compared with 49.06 ± 1.91% for the best CML method (random forest).
- On the combined healthy-aphasic dataset (Model-C), CNN remained the top-performing framework, with accuracies of 59.17% on the only-words dataset and 57.29% on the vowels + words dataset; all ML algorithms performed poorly overall on this dataset.
- The paper reports that transforming Model-A into a binary normal-vs-aphasic classifier achieved 100% accuracy when the decision threshold κ was set to 0.9.
Methods
- The study reused Mandarin speech datasets from a prior collection: 34 healthy subjects and 12 aphasic patients, with 6 Mandarin vowels, 10 nouns, and 10 verbs; data were organized into two task-specific subsets per dataset type: 26 classes (vowels + words) and 20 classes (only-words).
- Two framework families were evaluated: CML with five classifiers (QSVM, RBF SVM, LDA, random forest, kNN) using 51 ASR features (MFCCs, energy, formants), and a CNN framework using ResNet-34 fine-tuned on HTD-based time-frequency images resized to 224 × 224 × 3.
- Model-A and Model-B were evaluated with 5-fold cross-validation; Model-C used a train-test split. Performance was measured by accuracy, precision, and recall, and the paper reports mean and standard deviation across folds where applicable.
Limitations
- The paper states that the aphasic dataset is limited and highly variable: the recruited patients had different impairment severity levels, and aphasic syndrome datasets are scarce with small samples per severity group, which contributed to poor performance, especially in Models B and C.
- The authors note that Model-B is not suitable for aphasia syndrome discrimination in its current form because it was trained only on aphasic speech and would need properly labeled data covering multiple aphasia syndromes.
- Evident from the text: the dataset is small overall (34 healthy, 12 aphasic), and the paper itself notes that 5 of the 12 patients recorded vowels only, which likely further restricts some analyses.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures






Verification
Claims the paper makes
4 Valid · 1 Not valid · 12 Not checked · 34 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“However, it can be retrained with properly labeled data that include all aphasia syndromes such as Global aphasia, Broca’s aphasia, Wernicke’s aphasia and amnesic aphasia datasets to discriminate between them [13].”
§ Performance Evaluation of Machine Learning Frameworks for Aphasia Assessment > 5. Discussion > 5.2. Aphasic Dataset: Model-B
What the sources say
- Partly supportedcites [13] Kohlschein, C., Schmitt, M., Schuller, B., Jeschke, S., and Werner, C.J. (2017, January 12–15). A machine learning based system for the automatic evaluation of aphasia speech. Proceedings of the 2017 IEEE 19th International Conference on e-Health Networking, Applications and Services (Healthcom), Dalian, China. sourceExperimental — accuracy not yet measured
The Abstract proposes a system that maps speech to aphasia type, and the Classifier section describes training a classifier to distinguish defined classes. The Aphasia types section names the four syndromes listed in the claim, but also identifies rarer and unclassified cases. The paper does not say that an existing Model-B can be retrained to discriminate among all aphasia syndromes. (LLM, conf=0.84)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Most research on automatic aphasia assessment [5,9,10,12,13,14,18] has focused on a single aphasia assessment task using automatic speech recognition techniques and a machine learning (ML) framework with a fixed dataset-type.”
§ Performance Evaluation of Machine Learning Frameworks for Aphasia Assessment > 1. Introduction
What the sources say
- Not checked — no full text of the sourcecites [5] Shahin, 2014, A comparison of GMM-HMM and DNN-HMM based pronunciation verification techniques for use in the assessment of childhood apraxia of speech, Interspeech source
- Not checked — no full text of the sourcecites [9] Qin, 2020, An End-to-End Approach to Automatic Speech Assessment for Cantonese-speaking People with Aphasia, J. Signal Process. Syst. source
- Not checked — no full text of the sourcecites [10] Mahmoud, 2020, An efficient deep learning based method for speech as-sessment of Mandarin-speaking aphasic patients, IEEE J. Biomed. Health Inform. source
- Not checked — no full text of the sourcecites [12] Le, 2016, Automatic Assessment of Speech Intelligibility for Individuals with Aphasia, IEEE ACM Trans. Audio Speech Lang. Process. source
- Partly supportedcites [13] Kohlschein, C., Schmitt, M., Schuller, B., Jeschke, S., and Werner, C.J. (2017, January 12–15). A machine learning based system for the automatic evaluation of aphasia speech. Proceedings of the 2017 IEEE 19th International Conference on e-Health Networking, Applications and Services (Healthcom), Dalian, China. sourceExperimental — accuracy not yet measured
The proposed system section describes using the Aachen database to train a machine-learning system for automatic aphasia assessment, supporting this paper as an example of that approach. But it proposes rating speech along six dimensions rather than a single assessment task, and describes acoustic-feature extraction rather than automatic speech recognition. The Conclusion also says large portions of the system remain to be implemented. (LLM, conf=0.88)
- Not checked — no full text of the sourcecites [14] Fraser, 2014, Automated classification of primary progressive aphasia subtypes from narrative speech transcripts, Cortex source
- Not checked — no full text of the sourcecites [18] Juhola, 2011, Comparison of machine learning methods for classifying aphasic and non-aphasic speakers, Comput. Methods Programs Biomed. source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“The choice of the CML’s classifiers in this paper was influenced by the comparative performance investigation reported in [19,25], which found that the random forest (RF) and a radial basis function (RBF) kernel SVM outperformed 14 other classification algorithms.”
§ Performance Evaluation of Machine Learning Frameworks for Aphasia Assessment > 3. Materials and Methods > 3.2. Machine Learning Frameworks
What the sources say
- Not matchedcites [19] Gu, Y., Bahrani, M., Billot, A., Lai, S., Braun, E.J., Varkanitsa, M., Bighetto, J., Rapp, B., Parrish, T.B., and Caplan, D. (July, January 30). A machine learning approach for predicting post-stroke aphasia recovery: A pilot study. Proceedings of the 13th ACM International Conference on PErvasive Technologies Related to Assistive Environments, Corfu, Greece.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Partly supportedcites [25] Wainer, J. (2016). Comparison of 14 different families of classification algorithms on 115 binary datasets. arXiv. sourceExperimental — accuracy not yet measured
The Abstract (p. 1) confirms Wainer (2016) is a comparative performance study whose top performers include random forest and RBF (Gaussian-kernel) SVM, and the Conclusion (p. ~14) states random forests, RBF SVM and gradient boosting machines "most likely will result in the highest accuracy." However, the paper tested 14 classification algorithms IN TOTAL, so RF and RBF SVM did not outperform "14 other" algorithms; moreover the top tier is three, not two — gbm is not significantly different from RF and RBF SVM (Bayesian ANOVA against a 0.0112 irrelevance threshold, Discussion §5), a qualifier the citing sentence drops. (LLM, conf=0.82)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Not matched
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“All ML algorithms, including CNN, have significantly low performance on the aphasic patients’ dataset.”
§ Performance Evaluation of Machine Learning Frameworks for Aphasia Assessment > 4. Results > 4.2. ML Performance on Aphasic Dataset: Model-B
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and 67.78 ± 0.047% with the aphasic patients’ dataset.”
- Also stated in: § 4.2. ML Performance on Aphasic Dataset: Model-B · plain
“The CNN model has a higher accuracy of 67.78 ± 0.047% for the only-words dataset compared with 49.06 ± 1.91% for the random forest classifier with the same dataset.”
§ Performance Evaluation of Machine Learning Frameworks for Aphasia Assessment > Abstract Speech assessment is an essential part of the rehabilitation procedure for patients with aphasia (PWA). It is a comprehensive and time-consuming process that aims to discriminate between healthy individuals and aphasic patients, determine the type of aphasia syndrome, and determine the patients’ impairment severity levels (these are referred to here as aphasia assessment tasks). Hence, the automation of aphasia assessment tasks is essential. In this study, the performance of three automatic speech assessment models based on the speech dataset-type was investigated. Three types of datasets were used: healthy subjects’ dataset, aphasic patients’ dataset, and a combination of healthy and aphasic datasets. Two machine learning (ML)-based frameworks, classical machine learning (CML) and deep neural network (DNN), were considered in the design of the proposed speech assessment models. In this paper, the DNN-based framework was based on a convolutional neural network (CNN). Direct or indirect transformation of these models to achieve the aphasia assessment tasks was investigated. Comparative performance results for each of the speech assessment models showed that quadrature-based high-resolution time-frequency images with a CNN framework outperformed all the CML frameworks over the three dataset-types. The CNN-based framework reported an accuracy of 99.23 ± 0.003% with the healthy individuals’ dataset and 67.78 ± 0.047% with the aphasic patients’ dataset. Moreover, direct or transformed relationships between the proposed speech assessment models and the aphasia assessment tasks are attainable, given a suitable dataset-type, a reasonably sized dataset, and appropriate decision logic in the ML framework.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 4.2. ML Performance on Aphasic Dataset: Model-B · plain
26 more claims
Create a free account to see all 51 claims and every source verdict.
- It may state two separate claims as one. 15
- Probably not a substantial claim. 10
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Mahmoud et al. (2021) compared classical machine‐learning frameworks with convolutional neural network approaches, showing that deep learning methods more effectively captured the complexity of aphasic language.”
From Exploring the Effectiveness of Technology‐Based Interventions in Aphasia Rehabilitation: A Systematic Review (2026) · cites this paper as Mahmoud et al. 2021
Experimental — accuracy not yet measured
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 2×
“In order to utilize the weights of the pre-trained ResNet-34 model for transfer learning, all TFD RGB color images were resized to 224 × 224 × 3 pixels and normalized as per ImageNet dataset characteristics, before feeding them to the pre-trained ResNet-34 CNN model [34].”
MethodAdam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1512.03385 · Cited by 4,196 · cited 2×
“In order to utilize the weights of the pre-trained ResNet-34 model for transfer learning, all TFD RGB color images were resized to 224 × 224 × 3 pixels and normalized as per ImageNet dataset characteristics, before feeding them to the pre-trained ResNet-34 CNN model [34].”
MethodCyclical Learning Rates for Training Neural Networks
Smith, L. N. (2017) · 2017 IEEE Winter Conference on Applications of Computer Vision (WACV) · doi:10.1109/wacv.2017.58 · Cited by 2,670 · cited 2×
“A learning rate range test was used to estimate the maximum learning rate’s optimum value [35].”
Method(1992) · Cited by 1,405
Understanding how image quality affects deep neural networks
Dodge, S., & Karam, L. (2016) · 2016 Eighth International Conference on Quality of Multimedia Experience (QoMEX) · doi:10.1109/qomex.2016.7498955 · Cited by 835 · cited 2×
“The HTD has been found to produce a high-resolution TF image Sensors 2021 , 21 , 2582 6 of 15 of Mandarin speech signals; hence, it the speech signal classification can be improved using the CNN model [10,31–33]. neural networks library [38].”
ResultNovel Speech Signal Processing Algorithms for High-Accuracy Classification of Parkinson’s Disease
Tsanas, A., Little, M. A., McSharry, P. E., Spielman, J., & Ramig, L. O. (2012) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2012.2183367 · Cited by 732
Huber, W., Poeck, K., & Willmes, K. (1984) · PubMed · Cited by 338
Development of convolutional neural network and its application in image classification: a survey
Wang, W., Yang, Y., Wang, X., Wang, W., & Li, J. (2019) · Optical Engineering · doi:10.1117/1.oe.58.4.040901 · Cited by 252 · cited 1×
“The HTD has been found to produce a high-resolution TF image Sensors 2021 , 21 , 2582 6 of 15 of Mandarin speech signals; hence, it the speech signal classification can be improved using the CNN model [10,31–33]. neural networks library [38].”
Automated classification of primary progressive aphasia subtypes from narrative speech transcripts
Fraser, K. C., Meltzer, J. A., Graham, N. L., Leonard, C., Hirst, G., Black, S. E., & Rochon, E. (2014) · Cortex · doi:10.1016/j.cortex.2012.12.006 · Cited by 229 · cited 3×
“Speech assessment is used extensively in the diagnosis of Parkinson’s and aphasia diseases [1–14].”
BackgroundMethodVoice Disorder Identification by Using Machine Learning Techniques
Verde, L., De Pietro, G., & Sannino, G. (2018) · IEEE Access · doi:10.1109/access.2018.2816338 · Cited by 156 · cited 1×
“Secondly, for the assessment task of discriminating between normal and aphasic speech [6,7], Model-A can either be retrained to include an aphasic dataset to act as a binary classification problem or the model remains the same.”
BackgroundHussain, Z. M., & Boashash, B. (2002) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2002.800406 · Cited by 141 · cited 1×
“In this research, the hyperbolic T-distribution (HTD) [29,30] was used as a time-frequency-based image input to the CNN model within each model.”
MethodYang, H., Yao, Q., Yu, A., et al. (2019) · IEEE Transactions on Communications · doi:10.1109/tcomm.2019.2894711 · Cited by 128
A comparative study of adaptive, automatic recognition of disordered speech
Christensen, H., Cunningham, S., Fox, C., Green, P., & Hain, T. (2012) · Interspeech 2012 · doi:10.21437/interspeech.2012-484 · Cited by 123 · cited 2×
“In this aphasia assessment task, the performance of this model depends on the size of the training dataset from each aphasia type, which is a real challenge due to domain data scarcity [39].”
BackgroundResultAutomatic quantitative analysis of spontaneous aphasic speech
Le, D., Licata, K., & Mower Provost, E. (2018) · Speech Communication · doi:10.1016/j.specom.2018.04.001 · Cited by 97 · cited 1×
“Speech assessment is used extensively in the diagnosis of Parkinson’s and aphasia diseases [1–14].”
MethodMultigranulation rough-fuzzy clustering based on shadowed sets
Zhou, J., Lai, Z., Miao, D., Gao, C., & Yue, X. (2018) · Information Sciences · doi:10.1016/j.ins.2018.05.053 · Cited by 95 · cited 2×
“However, other classification algorithms such as fuzzy and neuro-fuzzy-based can be used when the data is affected by uncertainty and/or inaccuracies [26 The performance of different classifiers for each speech assessment model were compared.”
MethodAvuçlu, E., & Elen, A. (2020) · Medical & Biological Engineering & Computing · doi:10.1007/s11517-020-02260-3 · Cited by 74
Machine learning‐based multimodal prediction of language outcomes in chronic aphasia
Kristinsson, S., Zhang, W., Rorden, C., Newman‐Norlund, R., Basilakos, A., Bonilha, L., Yourganov, G., Xiao, F., Hillis, A., & Fridriksson, J. (2021) · Human Brain Mapping · doi:10.1002/hbm.25321 · Cited by 64 · cited 3×
“Conventional ASR features were used in the feature stage of the three A set of 51 ASR features based on Mel-Frequency Cepstral Coefficients (MFCCs), energy signals and the formant frequencies were calculated for the proposed models [20–24]. assessment models.”
MethodAmami, R., & Smiti, A. (2017) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2016.08.021 · Cited by 63 · cited 1×
“Secondly, for the assessment task of discriminating between normal and aphasic speech [6,7], Model-A can either be retrained to include an aphasic dataset to act as a binary classification problem or the model remains the same.”
BackgroundAutomatic Assessment of Speech Intelligibility for Individuals with Aphasia
Le, D., Licata, K., Persad, C., & Provost, E. M. (2016) · IEEE/ACM Transactions on Audio, Speech, and Language Processing · doi:10.1109/taslp.2016.2598428 · Cited by 59 · cited 4×
“These tasks are the discrimination between normal and aphasic speech [9], the assessment of the degree of severity of impairment for aphasic patients [10,12], and the classification of aphasia syndromes (such as Global aphasia, Broca’s aphasia, Wernicke’s aphasia and amnesic aphasia) [13,14].”
BackgroundMethodInfluentialComparing Humans and Automatic Speech Recognition Systems in Recognizing Dysarthric Speech
Mengistu, K. T., & Rudzicz, F. (2011) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-21043-3_36 · Cited by 55 · cited 1×
“This finding agrees with what was reported in the literature where data scarcity [39], abnormal speech patterns [40], and speaker variability [41] are challenging to any classification problem.”
ResultTime-frequency analysis of normal and abnormal biological signals
Mahmoud, S. S., Hussain, Z. M., Cosic, I., & Fang, Q. (2006) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2006.02.001 · Cited by 50 · cited 1×
“In this research, the hyperbolic T-distribution (HTD) [29,30] was used as a time-frequency-based image input to the CNN model within each model.”
MethodSelf-citationMatias-Guiu, J. A., Díaz-Álvarez, J., Cuetos, F., Cabrera-Martín, M. N., Segovia-Ríos, I., Pytel, V., Moreno-Ramos, T., Carreras, J. L., Matías-Guiu, J., & Ayala, J. L. (2019) · Cortex · doi:10.1016/j.cortex.2019.05.007 · Cited by 50
Mustafa, M. B., Rosdi, F., Salim, S. S., & Mughal, M. U. (2015) · Expert Systems with Applications · doi:10.1016/j.eswa.2015.01.033 · Cited by 37 · cited 1×
“This finding agrees with what was reported in the literature where data scarcity [39], abnormal speech patterns [40], and speaker variability [41] are challenging to any classification problem.”
ResultAn Efficient Deep Learning Based Method for Speech Assessment of Mandarin-Speaking Aphasic Patients
Mahmoud, S. S., Kumar, A., Tang, Y., Li, Y., Gu, X., Fu, J., & Fang, Q. (2020) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2020.3011104 · Cited by 36 · cited 10×
“In [10], we collected Mandarin datasets from healthy subjects and aphasic patients.”
BackgroundMethodInfluentialSelf-citationShahin, M., Ahmed, B., McKechnie, J., Ballard, K., & Gutierrez-Osuna, R. (2014) · Interspeech 2014 · doi:10.21437/interspeech.2014-377 · Cited by 34 · cited 1×
“Most research on automatic aphasia assessment [5,9,10,12–14 , 18 ] has focused on a single aphasia assessment task using This paper is organized as follows.”
BackgroundA Geometric Fuzzy-Based Approach for Airport Clustering
Postorino, M. N., & Versaci, M. (2014) · Advances in Fuzzy Systems · doi:10.1155/2014/201243 · Cited by 34 · cited 2×
“…reported in [19,25], which found that the random a radial basis function (RBF) kernel SVM outperformed 14 other classification However, other classification algorithms such as fuzzy and neuro-fuzzy-based techniques can be used when the data is affected by uncertainty and/or inaccuracies [26– 28].”
MethodHighly Accurate Mandarin Tone Classification In The Absence of Pitch Information
Ryant, N., Slaney, M., Liberman, M., Shriberg, E., & Yuan, J. (2014) · Speech Prosody 2014 · doi:10.21437/speechprosody.2014-123 · Cited by 32
Automatic Assessment of Speech Impairment in Cantonese-Speaking People with Aphasia
Qin, Y., Lee, T., & Kong, A. P. H. (2020) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2019.2956371 · Cited by 32
Automatic Speech Assessment for People with Aphasia Using TDNN-BLSTM with Multi-Task Learning
Qin, Y., Lee, T., Feng, S., & Kong, A. P. H. (2018) · Interspeech 2018 · doi:10.21437/interspeech.2018-1630 · Cited by 30
A machine learning based system for the automatic evaluation of aphasia speech
Kohlschein, C., Schmitt, M., Schuller, B., Jeschke, S., & Werner, C. J. (2017) · 2017 IEEE 19th International Conference on e-Health Networking, Applications and Services (Healthcom) · doi:10.1109/healthcom.2017.8210766 · Cited by 29 · cited 5×
“However, it can be retrained with properly labeled data that include all aphasia syndromes such as Global aphasia, Broca’s aphasia, Wernicke’s aphasia and amnesic aphasia datasets to discriminate between them [13].”
BackgroundInfluentialAn End-to-End Approach to Automatic Speech Assessment for Cantonese-speaking People with Aphasia
Qin, Y., Wu, Y., Lee, T., & Kong, A. P. H. (2020) · Journal of Signal Processing Systems · doi:10.1007/s11265-019-01511-3 · Cited by 25 · cited 5×
“The CNN models are widely used across various applications including aphasia assessment tasks [9].”
BackgroundMethodInfluentialComparison of machine learning methods for classifying aphasic and non-aphasic speakers
Järvelin, A., & Juhola, M. (2011) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2011.02.015 · Cited by 18 · cited 3×
“For example, in [18], researchers investigated eight machine learning (ML) classifiers’ performance on two aphasia assessment tasks.”
BackgroundA machine learning approach for predicting post-stroke aphasia recovery: a pilot study
Gu, Y., Bahrani, M., Billot, A., Lai, S., Braun, E. J., Varkanitsa, M., Bighetto, J., Rapp, B., Parrish, T. B., Caplan, D., et al. (2020) · Proceedings of the 13th ACM International Conference on PErvasive Technologies Related to Assistive Environments · doi:10.1145/3389189.3389204 · Cited by 12 · cited 3×
“The choice of the CML’s classifiers in this paper was influenced by the performance investigation reported in [19,25], which found that the random a radial basis function (RBF) kernel SVM outperformed 14 other classification However, other classification algorithms such as fuzzy and neuro-fuzzy-based…”
BackgroundMethodTowards Automatic Speech-Language Assessment for Aphasia Rehabilitation
Le, D. (2017) · Deep Blue (University of Michigan) · Cited by 3
Research works printed but not linked: 4
- Mahmoud, S.S., Fang, Q., Tang, Y., Alsulami, M., and Alotaibi, M. (2019, January 28–29). Automatic Mandarin Vowels Recognition Framework for Aphasic Patients Rehabilitation. Proceedings of the 13th IEEE-EMBS International Summer School and Symposium on Medical Devices and Biosensors (MDBS2019), Chengdu, China.
- Li, S. (2014). Speech Therapy, Huaxia Press.
- Rabiner, L.R., and Schafer, R.W. (2010). Theory and Application of Digital Speech Processing, Pearson.
- Suresh, 2018, Prasad Kannojia, Effects of Varying Resolution on Performance of CNN based Image Classification: An Experi-mental Study, Int. J. Comput. Sci. Eng., 6, 451
Cited by
A Survey of Technologies Facilitating Home and Community-Based Stroke Rehabilitation
Sun, X., Ding, J., Dong, Y., Ma, X., Wang, R., Jin, K., Zhang, H., & Zhang, Y. (2022) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2022.2050545 · Cited by 26
“…to rule out conventional time series processing methods and statistical models, such as Hidden Markov Model (HMM) (Le et al., 2016; Lee et al., 2016), and introduce deep learning models (Jayasree-Krishnan et al., 2019; Mahmoud et al., 2021) for their advantages in feature extraction and learning.”
Herath, H. M. D. P. M., Weraniyagoda, W. A. S. A., Rajapaksha, R. T. M., Wijesekara, P. A. D. S. N., Sudheera, K. L. K., & Chong, P. H. J. (2022) · Sensors · doi:10.3390/s22186966 · Cited by 23
“A recent study evaluating the performance of ML models for aphasia assessment reports that the random forest classifier has a better classification performance than the KNN algorithm [107].”
BackgroundMethodRanjith, R., & Chandrasekar, A. (2023) · Computer Speech & Language · doi:10.1016/j.csl.2023.101568 · Cited by 17
Silva, U. D., Madanian, S., Olsen, S., Templeton, J. M., Poellabauer, C., Schneider, S. L., Narayanan, A., & Rubaiat, R. (2025) · Journal of Medical Internet Research · doi:10.2196/63004 · Cited by 31
Sayadi, M., Vijayakumar, V., Langarizadeh, M., Bayazian, G., & Torabinezhad, F. (2022) · Electronics · doi:10.3390/electronics11244235 · Cited by 15
“[26] 2021 Aphasia (Laryngeal) Deep learning (ACC = 99.”
Metadata
- Authors
- Mahmoud, S. S., Kumar, A., Li, Y., Tang, Y., & Fang, Q.
- Affiliations
- Shantou University (CN)
- Corresponding author
- Mahmoud, Seedahmed S.
- Fang, Qiang
- Venue
- Sensors
- Publisher
- MDPI AG
- Published
- 2021-04-07
- Volume
- 21
- Issue
- 8
- Pages
- 2582
- ISSN
- 1424-8220 (ISSN-L)
- PMC
- PMC8067696
- OpenAlex
- W3149523164
- Semantic Scholar
- d0e8187547311caa3901aca262afef16debe280e
- Cited by
- 27
- References
- 40
- Citations by year
- 2026: 2 · 2025: 7 · 2024: 5 · 2023: 5 · 2022: 7 · 2021: 1
- FWCI
- 2.82
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Sensors (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- 2020 Li Ka Shing Foundation Cross-disciplinary Research Grant (2020LKSFG04C)
- Topics
- Health Sciences › Health Professions › Speech and Hearing › Voice and Speech Disorders
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Physical Sciences › Computer Science › Signal Processing › Speech and Audio Processing
- Keywords
- aphasia assessment, convolutional neural networks, classical machine learning, time-frequency images, aphasia severity, speech rehabilitation, model performance comparison
- MeSH terms
- Machine Learning, Aphasia, Humans, Speech, Neural Networks, Computer
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-01, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-01 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.