Pith. sign in

REVIEW 4 major objections 6 minor 26 references

Suicide Risk Assessment Using Multimodal Speech Features: A Study on the SW1 Challenge Dataset

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper claims that weighting attention across speech, text, and handcrafted acoustic features, together with mixup regularization, generalizes best for suicide-risk classification, reaching 56% test accuracy on the SW1 Challenge…

desk verdict A clean, competent challenge report whose central claim—that weighted attention plus mixup drove better generalization—rests on a 3-sample test-set difference and a metric table that does not add up. read the letter →

arxiv 2505.13069 v1 pith:4APTQRAH submitted 2025-05-19 cs.CL cs.LGcs.SDeess.AS

Ambre Marie , Ilias Maoudj , Guillaume Dardenne , Gwenolé Quellec This is my paper · ORCID
classification cs.CLcs.LGcs.SDeess.AS
keywords suicideriskassessmentmultimodalfusionspeechembeddingsweightedattentionmixupregularizationacousticfeaturesadolescentMINI-KID
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper asks whether combining automatic transcription, language-model embeddings, speech embeddings, and handcrafted acoustic features can classify suicide risk in adolescents from voice recordings, and which way of fusing those modalities generalizes best. On the SW1 Challenge dataset of 600 Mandarin-speaking adolescents labelled by the MINI-KID interview, the authors compare early concatenation, modality-specific attention, and weighted attention with mixup regularization. They report that the simplest fusion reaches 70% development accuracy but collapses to 45% on the held-out test set, while weighted attention with spectral-contrast and pitch features holds 69% development accuracy and rises to 56% test accuracy, the best generalization. The practical stakes: if the result holds, clinician-usable speech screening could favour explicit modality weighting and handcrafted prosody over raw concatenated embeddings, though the authors caution the findings are confined to current-risk labels under the MINI-KID framework.

What carries the argument

The load-bearing mechanism is a weighted attention fusion layer combined with mixup regularization. Each modality—audio embeddings from WavLM, text embeddings from Chinese RoBERTa, and handcrafted acoustic features—passes through its own feedforward layer, then the model learns an attention weight per modality before merging the transformed embeddings; mixup regularizes training by interpolating embeddings from different samples. The handcrafted acoustic branch, whose third version adds spectral contrast and pitch-related statistics derived from a fundamental-frequency estimation algorithm, is what carries the extra generalization beyond the self-supervised audio and text representations.

What would settle it

Transcribe a held-out subset of the 100 test recordings by hand, feed those human transcripts through the same RoBERTa and audio pipelines under the best submission, and compare accuracy; if accuracy does not rise above 56% and the attention weights shift toward the audio branch, the linguistic modality is contributing little and the claimed fusion benefit is mostly acoustic.

Watch

Extended reading notes

Core claim

The central discovery, as the authors state it, is that a multimodal suicide-risk classifier using WavLM audio embeddings, Chinese RoBERTa text embeddings, and handcrafted acoustic features performs best on unseen data when the modalities are combined by learned weighted attention with mixup regularization rather than by early concatenation or simple modality-specific attention. Expanding the acoustic features from MFCC means to include spectral contrast and pitch statistics (fundamental frequency and voiced probability) raised held-out test accuracy to 56%, the highest among the three submissions, even though the development accuracy of 69% was one point below the early-fusion model's 70%. The paper interprets this as evidence that learned modality weighting and prosodic handcrafted features reduce overfitting to dataset-specific patterns. It also reports that the gap between development and test accuracy remains substantial, and that all results are strictly limited to classifying current suicide risk as defined by the MINI-KID scale.

Load-bearing premise

The paper assumes that the WhisperX transcriptions and the Chinese RoBERTa embeddings preserve the suicide-relevant content of adolescent Mandarin speech, yet it never checks transcription accuracy or embedding quality on this dataset.

Editorial extensions

If this is right

  • Early concatenation of WavLM and RoBERTa embeddings overfits, dropping from 70% development accuracy to 45% on the test set.
  • Adding handcrafted MFCC, spectral-contrast, and pitch features improves held-out accuracy even when development accuracy falls.
  • Learned modality weighting with mixup regularization yields the best test accuracy (56%) among the three fusion strategies.
  • The model separates at-risk from non-risk groups more clearly in its penultimate layer than in raw embeddings, per the paper's t-SNE visualization.
  • The reported accuracies should not be read as prediction of future suicidal behavior; the labels reflect current MINI-KID risk status only.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If WhisperX transcription is lossy for this adolescent Mandarin corpus, the text modality may be adding noise; swapping in human transcripts would reveal whether the fusion gain is linguistic or purely acoustic.
  • The attention weights learned by the best submission could be visualized per interview task (emotional regulation, reading, image description) to see which speech tasks carry the suicide-risk signal and which modalities the model learns to ignore.
  • The persistent dev–test gap suggests that even the best fusion strategy would need calibration or domain adaptation before deployment on a new adolescent population.
  • Because the acoustic branch is the one that changes between the second and third submissions, the 56% result can be read as evidence that task-specific handcrafted features still matter alongside self-supervised embeddings, a point the paper leaves implicit.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. This paper describes three submissions to the 1st SpeechWellness Challenge (Interspeech 2025) for suicide risk classification from Mandarin adolescent speech. The system pipeline combines WhisperX transcriptions, Chinese RoBERTa text embeddings, WavLM audio embeddings, and handcrafted acoustic features, with three fusion strategies: early concatenation, modality-specific processing with attention, and weighted attention with mixup regularization. On the challenge's balanced 400/100/100 train/dev/test split, the authors report dev accuracies of 70%, 63%, and 69% for the three submissions and test accuracies of 45%, 53%, and 56%, concluding that the weighted-attention version generalizes best. The discussion interprets these results in terms of embedding choices, acoustic feature design, and fusion architecture, and the limitations section correctly restricts the conclusions to the MINI-KID assessment framework.

Significance. If the reported results were statistically robust, the paper would provide a useful, if incremental, system description for speech-based suicide risk assessment on a public challenge benchmark. The study has several genuine strengths: it is evaluated on an externally held-out challenge test set whose labels were not available to the authors, it uses fixed pretrained embeddings rather than fitting the target labels through a learned representation, and it is honest about the MINI-KID-specific nature of the labels. The comparison among fusion strategies is a relevant question for the speech-and-language community. However, the central claim rests on a three-sample difference on a 100-sample test set, with no uncertainty quantification and with multiple architectural changes varied simultaneously, so the paper's current evidential value is limited to that of a descriptive challenge report rather than a demonstrated methodological conclusion.

major comments (4)
  1. [Section 4.1.3 and Table 1] The central claim that Submission 3 achieves 'the highest generalization performance' is based on test accuracies of 56% versus 53% and 45% on a balanced test set of 100 subjects. This is a difference of three correct labels. No confidence interval, bootstrap, or significance test is reported, and the observed drop from 69% dev accuracy to 56% test accuracy suggests substantial sampling or distribution shift. A 95% Wilson interval for 56/100 spans roughly 46% to 66%, which includes 53%, so the reported margin is fully consistent with chance variation. The authors should provide an uncertainty estimate for the test scores and either temper the generalization claim or support it with a statistically meaningful comparison.
  2. [Table 1] For each of the three submissions, Dev Accuracy, Dev F1, and Dev AUROC are reported as numerically identical to two decimal places (0.70, 0.70, 0.70; 0.63, 0.63, 0.63; 0.69, 0.69, 0.69). This is implausible for three genuinely distinct metrics and strongly suggests either a reporting error or that the three columns are all the same metric. Since the discussion draws interpretative conclusions from dev behavior, the authors must clarify what each column actually measures and, ideally, provide confusion matrices and per-class precision/recall for the dev and test sets.
  3. [Sections 2.3 and 2.4] Even if the test scores are accepted at face value, the comparison across submissions is confounded: the three systems differ in WavLM Base+ versus Large, chunk size and overlap, pooling strategy, MFCC-only versus MFCC plus spectral contrast and pitch features, fusion architecture, mixup regularization, number of training folds, and ensembling. Consequently, the improvement from 53% to 56% cannot be attributed specifically to weighted attention, mixup, or the expanded acoustic feature set. A controlled ablation, or at minimum an explicit statement that such attribution is not possible, is required to support the causal language used in Sections 4.2 and 5.
  4. [Sections 2.2 and 2.3.2] The text modality is built from WhisperX transcriptions and Chinese RoBERTa embeddings, but the paper reports no evaluation of transcription accuracy or downstream embedding quality on this adolescent Mandarin corpus. If the ASR pipeline or the pretrained embeddings fail to preserve suicide-relevant linguistic markers, the text modality may contribute noise rather than signal. Since all three submissions share this text pipeline, the issue does not invalidate the relative comparison, but it does weaken the paper's broader claim that 'integrating linguistic and paralinguistic features enhances classification performance.' A small transcription-quality check, such as word error rate on a sample or a complementary text-only baseline, would clarify this point.
minor comments (6)
  1. [Section 2.4] There is a typo: 'expolored' should be 'explored'.
  2. [Section 2.3.3] The phrase 'the acoustic representation was expended' should be 'expanded'; also, 'averaging each one over time' is awkward and should be reworded.
  3. [Sections 2.4.2 and 4.2] Section 2.4.2 states that the attention mechanism 'dynamically assigned weights to each modality,' while Section 4.2 later says the second approach failed due to 'lack of explicit weighting mechanisms.' These statements are inconsistent and should be reconciled.
  4. [Abstract and Section 4.2] The abstract says weighted attention achieved '69% accuracy on the development set,' but Submission 1 had 70% dev accuracy; the claim of 'best generalization' should be tied to test accuracy (56%), not dev accuracy, and the wording should make this distinction clear.
  5. [Figure 2] The t-SNE visualization is presented only for Submission 2's audio embeddings and is descriptive. It would be more informative to show all submissions or provide a quantitative separation measure; otherwise, the visual claim of 'clearer distinction' should be softened.
  6. [Reproducibility] No code, configuration details, or per-sample predictions are released, which limits reproducibility of the three-submission comparison. Providing these artifacts, or at least detailed hyperparameters and ensemble counts, would strengthen the paper.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper reports an empirical comparison of three fusion systems on a held-out challenge test set, with no fitted parameter or self-citation chain that defines the reported outcome.

full rationale

The paper is a challenge system description rather than a derivation or prediction-from-first-principles study. Its central claim is that Submission 3 achieved the highest held-out test accuracy (56%) among three compared systems. This claim rests on external evaluation by the SpeechWellness Challenge organizers against undisclosed test labels, and the systems were selected using the development set, which is standard practice. Nothing in the manuscript defines the target result in terms of a fitted quantity: the embeddings are pretrained off-the-shelf models (WhisperX, Chinese RoBERTa, WavLM), the acoustic features are standard handcrafted descriptors, and the fusion architectures are described as alternative designs compared on a fixed split. There is no parameter that is fitted to the test set and then renamed a prediction, no uniqueness theorem imported from the authors' own prior work, and no equation in which an output quantity is identical by construction to an input quantity. The self-citations that do appear are ordinary references to the challenge definition [12] and to established toolkits such as WhisperX [13], WavLM [15], and RoBERTa [16], none of which is load-bearing in a circular way; the cited pretrained models are external, independently trained resources. The paper's known weakness is statistical: the decisive test-accuracy gap between Submission 3 and Submission 2 is three subjects out of 100, with no significance testing, and the submissions differ along multiple architectural axes simultaneously. That is a validity or evidential concern, not a circularity concern, because the comparison remains an empirical evaluation against an externally defined held-out benchmark rather than a reduction of the claim to its own inputs. The manuscript also candidly states its limitations, including the MINI-KID labeling framework and the dev-test generalization gap, which further supports the conclusion that the reported results are presented as measured outcomes rather than as consequences of the method's definitions.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No new entities, parameters, or forces are introduced. The paper relies entirely on existing models and dataset labels.

assumptions (4)
  • domain assumption WhisperX produces accurate enough Mandarin transcriptions for downstream linguistic analysis.
    Section 2.2 uses WhisperX large-v2 for all transcriptions without evaluating transcription quality on this adolescent speech corpus.
  • domain assumption WavLM and Chinese RoBERTa embeddings preserve suicide-risk-relevant acoustic and semantic cues.
    Sections 2.3.1 and 2.3.2 assume these pretrained models transfer to the SW1 domain without task-specific validation.
  • domain assumption The MINI-KID-based binary labels (at-risk vs not) are the ground truth for the classification task.
    Section 2.1 defines labels from MINI-KID; the paper acknowledges in Section 4.3 that this reflects current self-reported risk, not future behavior.
  • domain assumption The 400/100/100 split is balanced and the test labels are unbiased.
    Section 2.1 states the sets are balanced, but the paper does not verify this beyond the challenge description.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Suicide Risk Assessment Using Multimodal Speech Features: A Study on the SW1 Challenge Dataset." pith.science (2026). https://pith.science/paper/4APTQRAH

@misc{pith2026250513069,
  author       = {Pith},
  title        = {Pith review of: Suicide Risk Assessment Using Multimodal Speech Features: A Study on the SW1 Challenge Dataset},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4APTQRAH}},
  note         = {Machine review of arXiv:2505.13069}
}
read the original abstract

The 1st SpeechWellness Challenge conveys the need for speech-based suicide risk assessment in adolescents. This study investigates a multimodal approach for this challenge, integrating automatic transcription with WhisperX, linguistic embeddings from Chinese RoBERTa, and audio embeddings from WavLM. Additionally, handcrafted acoustic features -- including MFCCs, spectral contrast, and pitch-related statistics -- were incorporated. We explored three fusion strategies: early concatenation, modality-specific processing, and weighted attention with mixup regularization. Results show that weighted attention provided the best generalization, achieving 69% accuracy on the development set, though a performance gap between development and test sets highlights generalization challenges. Our findings, strictly tied to the MINI-KID framework, emphasize the importance of refining embedding representations and fusion mechanisms to enhance classification reliability.

Figures

Figures reproduced from arXiv: 2505.13069 by the authors.

Figure 1
Figure 1. Architecture of the Submission 3 classification pipeline. speech recordings from anonymized Chinese adolescents, with 50% identified as having SR based on validated psychological scales. In this work, we present a multimodal approach for assess￾ing SR, integrating automatic transcription, text embeddings, audio embeddings, and acoustic feature extraction. We explore different fusion strategies for combining these mo… view at source ↗
Figure 2
Figure 2. Visualization of AuE for Submission 2. Left: Raw AuE extracted from WavLM Large. Right: Pre-logits AuE. Purple : label 0 (non-risk), Green : label 1 (at-risk). 3.2. Embedding Representation To illustrate how the learned representations evolve from raw embeddings to the classifier’s penultimate layer, we applied t￾SNE to visualize both sets of embeddings [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 20 canonical work pages

  1. [1]

    Early identification of individuals at risk is critical for timely in- tervention, yet reliable and accessible risk assessment remains a persistent challenge

    Introduction Suicide is recognized as a significant global health issue, rank- ing among the leading causes of death in adolescents [1, 2]. Early identification of individuals at risk is critical for timely in- tervention, yet reliable and accessible risk assessment remains a persistent challenge. Traditional suicide risk (SR) assess- ments are based prim...

    work page 2025

  2. [2]

    Suicide Risk Assessment Using Multimodal Speech Features: A Study on the SW1 Challenge Dataset

    Materials and methods Each submission differs in its choice of embeddings and fusion strategies:(1)Uses only audio (WavLM) and text (RoBERTa) embeddings, with an early fusion approach;(2)Introduces acoustic features (MFCCs) alongside audio and text, with a modality-specific fusion strategy;(3)Expands acoustic embed- dings by incorporating spectral and pit...

    work page Pith review arXiv 2025

  3. [3]

    Overview of Performance Metrics Table 1 presents the main performance metrics for all three submissions

    Results 3.1. Overview of Performance Metrics Table 1 presents the main performance metrics for all three submissions. Each differs in its choice of embeddings, fusion strategies, and model variants. The table reports accuracy, F1- score, and AUROC on the dev set, along with accuracy on the held-out test set. Table 1:Performance metrics on the dev and test...

    work page

  4. [4]

    Embeddings The choice of embeddings significantly influenced classifica- tion performance, with each modality contributing complemen- tary information to the SR assessment task

    Discussion 4.1. Embeddings The choice of embeddings significantly influenced classifica- tion performance, with each modality contributing complemen- tary information to the SR assessment task. 4.1.1. Audio Embeddings AuE derived from WavLM provided a foundation for model- ing speech characteristics. In the first version, WavLM Base+ was used with TE, whi...

    work page

  5. [5]

    Our findings suggest that combining self-supervised speech representations with hand- crafted acoustic and linguistic features enhances classification performance

    Conclusion In this study, we explored a multimodal approach to suicide risk assessment using the SW1 Challenge dataset, integrating lin- guistic and paralinguistic features. Our findings suggest that combining self-supervised speech representations with hand- crafted acoustic and linguistic features enhances classification performance. Among the fusion st...

    work page

  6. [6]

    Suicide prevention in childhood and adolescence: a narrative review of current knowledge on risk and protective factors and effectiveness of interventions,

    D. Wasserman, V . Carli, M. Iosue, A. Javed, and H. Herrman, “Suicide prevention in childhood and adolescence: a narrative review of current knowledge on risk and protective factors and effectiveness of interventions,”Asia-Pacific Psychiatry, vol. 13, no. 3, p. e12452, Sep. 2021. [Online]. Available: https://onlinelibrary.wiley.com/doi/10.1111/appy.12452

    work page doi:10.1111/appy.12452 2021

  7. [7]

    Suicide and suicide risk in adolescents

    L. L. Hua, J. Lee, M. H. Rahmandar, and E. J. Sigel, “Suicide and suicide risk in adolescents.”Pediatrics, 2023

    work page 2023

  8. [8]

    Limitations of a single- item assessment of suicide attempt history: Implications for stan- dardized suicide risk assessment

    M. A. Hom, T. Joiner, and R. Bernert, “Limitations of a single- item assessment of suicide attempt history: Implications for stan- dardized suicide risk assessment.”Psychological assessment, vol. 28 8, pp. 1026–30, 2016

    work page 2016

Show all 26 references
  1. [9]

    The role of the clinical interview in suicide risk assessment,

    D. Wasserman, C. Wasserman, M. Schechter, and J. T. Malts- berger, “The role of the clinical interview in suicide risk assessment,” inOxford Textbook of Suicidology and Suicide Pre- vention. Oxford University Press, 01 2021. [Online]. Available: https://doi.org/10.1093/med/978...

    2021

  2. [10]

    How do healthcare professionals interview patients to assess suicide risk?

    R. McCabe, I. Sterno, S. Priebe, R. Barnes, and R. Byng, “How do healthcare professionals interview patients to assess suicide risk?” BMC Psychiatry, vol. 17, 2017

    2017

  3. [11]

    A review of depression and suicide risk assessment using speech analysis,

    N. Cummins, S. Scherer, J. Krajewski, S. Schnieder, J. Epps, and T. Quatieri, “A review of depression and suicide risk assessment using speech analysis,”Speech Commun., vol. 71, pp. 10–49, 2015

    2015

  4. [12]

    Content analysis of suicidal notes - the verbal behaviour,

    J. D. Fekete and S. Fekete, “Content analysis of suicidal notes - the verbal behaviour,”European Psychiatry, 2024

    2024

  5. [13]

    Linguistic research on specific features of suicidal communication—past, present, and future,

    S. Fekete, P. Osvath, J. Fekete, and V . V oros, “Linguistic research on specific features of suicidal communication—past, present, and future,”Clinical Psychological Science, vol. 6, pp. 759 – 760, 2018

    2018

  6. [14]

    Investigat- ing voice quality as a speaker-independent indicator of depression and ptsd,

    S. Scherer, G. Stratou, J. Gratch, and L.-P. Morency, “Investigat- ing voice quality as a speaker-independent indicator of depression and ptsd,” pp. 847–851, 2013

    2013

  7. [15]

    Examination of speech analysis to predict suicidal behavior in depression,

    S. Y ¨unden, M. Ak, M. Sert, S. Gica, O. C ¸ inar, and Y . A. Acar, “Examination of speech analysis to predict suicidal behavior in depression,”European Psychiatry, 2024

    2024

  8. [16]

    Acoustic and language analysis of speech for suicidal ideation among us veterans,

    A. Belouali, S. Gupta, V . Sourirajan, J. Yu, N. Allen, A. Alaoui, M. Dutton, and M. Reinhard, “Acoustic and language analysis of speech for suicidal ideation among us veterans,”BioData Mining, vol. 14, 2020

    2020

  9. [17]

    The 1st speechwellness challenge: Detecting suicidal risk among adolescents,

    W. Wu, Z. Cui, C. Lei, Y . Duan, D. Qu, J. Wu, B. Zhou, R. Chen, and C. Zhang, “The 1st speechwellness challenge: Detecting suicidal risk among adolescents,” 2025. [Online]. Available: https://arxiv.org/abs/2501.06474

    2025

  10. [18]

    Whisperx: Time-accurate speech transcription of long-form audio,

    M. Bain, J. Huh, T. Han, and A. Zisserman, “Whisperx: Time-accurate speech transcription of long-form audio,”INTER- SPEECH 2023, 2023

    2023

  11. [19]

    Detection of suicidal ideation on social media: Multimodal, rela- tional, and behavioral analysis,

    D. Ram ´ırez-Cifuentes, A. Freire, R. Baeza-Yates, J. Punt ´ı, P. Medina-Bravo, D. Velazquez, J. M. Gonfaus, and J. Gonz`alez, “Detection of suicidal ideation on social media: Multimodal, rela- tional, and behavioral analysis,”Journal of Medical Internet Re- search, vol. 22, 2020

    2020

  12. [20]

    Wavlm: Large-scale self-supervised pre-training for full stack speech processing,

    S. Chen, C. Wang, Z. Chen, Y . Wu, S. Liu, Z. Chen, J. Li, N. Kanda, T. Yoshioka, X. Xiao, J. Wu, L. Zhou, S. Ren, Y . Qian, Y . Qian, J. Wu, M. Zeng, X. Yu, and F. Wei, “Wavlm: Large-scale self-supervised pre-training for full stack speech processing,”IEEE Journal of Selected...

    2022

  13. [21]

    Roberta: A robustly optimized bert pretraining approach,

    Y . Liu, M. Ott, N. Goyal, J. Du, M. Joshi, D. Chen, O. Levy, M. Lewis, L. Zettlemoyer, and V . Stoyanov, “Roberta: A robustly optimized bert pretraining approach,” 2019. [Online]. Available: https://arxiv.org/abs/1907.11692

    2019 arXiv

  14. [22]

    Bert: Pre-training of deep bidirectional transformers for language understanding,

    J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, “Bert: Pre-training of deep bidirectional transformers for language understanding,” 2019. [Online]. Available: https://arxiv.org/abs/ 1810.04805

    2019 arXiv

  15. [23]

    Characterizing depres- sive related speech with mfcc,

    S. Suwannakhun and T. Yingthawornsuk, “Characterizing depres- sive related speech with mfcc,”2019 14th International Joint Symposium on Artificial Intelligence and Natural Language Pro- cessing (iSAI-NLP), pp. 1–6, 2019

    2019

  16. [24]

    Association between suicidal ideation and acoustic parameters of university students’ voice and speech: a pilot study,

    C. F. Saavedra, T. O. Hern ´andez, C. A. Godoy, A. R. P ´erez, D. F. Salinas, and R. L. Hern´andez, “Association between suicidal ideation and acoustic parameters of university students’ voice and speech: a pilot study,”Logopedics Phoniatrics Vocology, vol. 46, pp. 55 – 62, 2020

    2020

  17. [25]

    Investigating the speech characteristics of suicidal adolescents,

    S. Scherer, J. Pestian, and L.-P. Morency, “Investigating the speech characteristics of suicidal adolescents,”2013 IEEE Inter- national Conference on Acoustics, Speech and Signal Processing, pp. 709–713, 2013

    2013

  18. [26]

    Agreement between clinical and mini diagnoses in outpatients with mood and anxiety disorders

    F. Verhoeven, L. Swaab, I. Carlier, A. V . van Hemert, F. Zit- man, H. Ruh´e, R. Schoevers, and E. Giltay, “Agreement between clinical and mini diagnoses in outpatients with mood and anxiety disorders.”Journal of affective disorders, vol. 221, pp. 268–274, 2017

    2017

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.