Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
80 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
7 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Towards Knowledge-Infused Automated Disease Diagnosis Assistant (2405.11181v1)

Published 18 May 2024 in cs.AI and cs.CL

Abstract: With the advancement of internet communication and telemedicine, people are increasingly turning to the web for various healthcare activities. With an ever-increasing number of diseases and symptoms, diagnosing patients becomes challenging. In this work, we build a diagnosis assistant to assist doctors, which identifies diseases based on patient-doctor interaction. During diagnosis, doctors utilize both symptomatology knowledge and diagnostic experience to identify diseases accurately and efficiently. Inspired by this, we investigate the role of medical knowledge in disease diagnosis through doctor-patient interaction. We propose a two-channel, knowledge-infused, discourse-aware disease diagnosis model (KI-DDI), where the first channel encodes patient-doctor communication using a transformer-based encoder, while the other creates an embedding of symptom-disease using a graph attention network (GAT). In the next stage, the conversation and knowledge graph embeddings are infused together and fed to a deep neural network for disease identification. Furthermore, we first develop an empathetic conversational medical corpus comprising conversations between patients and doctors, annotated with intent and symptoms information. The proposed model demonstrates a significant improvement over the existing state-of-the-art models, establishing the crucial roles of (a) a doctor's effort for additional symptom extraction (in addition to patient self-report) and (b) infusing medical knowledge in identifying diseases effectively. Many times, patients also show their medical conditions, which acts as crucial evidence in diagnosis. Therefore, integrating visual sensory information would represent an effective avenue for enhancing the capabilities of diagnostic assistants.

PDF HTML Abstract
Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Bookmark Chat Bubble Oval Streamline Icon: https://streamlinehq.com Chat (Pro)
Definition Search Book Streamline Icon: https://streamlinehq.com
References (56)
  1. Use of the internet for health information: United states, 2009. \JournalTitleNCHS data brief 1—8 (2011).
  2. George, P. P. et al. Online elearning for undergraduates in health professions: a systematic review of the impact on knowledge, skills, attitudes and satisfaction. \JournalTitleJournal of global health 4 (2014).
  3. Wei, Z. et al. Task-oriented dialogue system for automatic diagnosis. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), 201–207 (2018).
  4. The interplay of a conversational ontology and ai planning for health dialogue management. In Proceedings of the 36th annual ACM symposium on applied computing, 611–619 (2021).
  5. Liao, K. et al. Task-oriented dialogue system for automatic disease diagnosis via hierarchical reinforcement learning. \JournalTitlearXiv preprint arXiv:2004.14254 (2020).
  6. Refuel: Exploring sparse features in deep reinforcement learning for fast disease diagnosis. \JournalTitleAdvances in neural information processing systems 31, 7322–7331 (2018).
  7. The graph-based mutual attentive network for automatic diagnosis. In Proceedings of the Twenty-Ninth International Conference on International Joint Conferences on Artificial Intelligence, 3393–3399 (2021).
  8. Reinforcement learning in healthcare: A survey. \JournalTitleACM Computing Surveys (CSUR) 55, 1–36 (2021).
  9. Artificial intelligence in disease diagnosis: a systematic literature review, synthesizing framework and future research agenda. \JournalTitleJournal of ambient intelligence and humanized computing 1–28 (2022).
  10. Context-aware symptom checking for disease diagnosis using hierarchical reinforcement learning. In Proceedings of the AAAI Conference on Artificial Intelligence, vol. 32 (2018).
  11. Ramos, J. et al. Using tf-idf to determine word relevance in document queries. In Proceedings of the first instructional conference on machine learning, vol. 242, 29–48 (Citeseer, 2003).
  12. The potential for artificial intelligence in healthcare. \JournalTitleFuture healthcare journal 6, 94 (2019).
  13. Deep learning for healthcare: review, opportunities and challenges. \JournalTitleBriefings in bioinformatics 19, 1236–1246 (2018).
  14. Ventres, W. et al. Physicians, patients, and the electronic health record: an ethnographic analysis. \JournalTitleThe Annals of Family Medicine 4, 124–131 (2006).
  15. Li, Y. et al. Behrt: transformer for electronic health records. \JournalTitleScientific reports 10, 1–12 (2020).
  16. Electronic health records based reinforcement learning for treatment optimizing. \JournalTitleInformation Systems 104, 101878 (2022).
  17. Mnih, V. et al. Playing atari with deep reinforcement learning. \JournalTitlearXiv preprint arXiv:1312.5602 (2013).
  18. Predictive modeling of depression and anxiety using electronic health records and a novel machine learning approach with artificial intelligence. \JournalTitleScientific reports 11, 1–9 (2021).
  19. Med-bert: pretrained contextualized embeddings on large-scale structured electronic health records for disease prediction. \JournalTitleNPJ digital medicine 4, 86 (2021).
  20. Bert: Pre-training of deep bidirectional transformers for language understanding. \JournalTitlearXiv preprint arXiv:1810.04805 (2018).
  21. Med7: A transferable clinical natural language processing model for electronic health records. \JournalTitleArtificial Intelligence in Medicine 118, 102086 (2021).
  22. Benefits and drawbacks of electronic health record systems. \JournalTitleRisk management and healthcare policy 4, 47 (2011).
  23. Inquire and diagnose: Neural symptom checking ensemble using deep reinforcement learning. In NIPS Workshop on Deep Reinforcement Learning (2016).
  24. Dietterich, T. G. Hierarchical reinforcement learning with the maxq value function decomposition. \JournalTitleJournal of artificial intelligence research 13, 227–303 (2000).
  25. Diaformer: Automatic diagnosis via symptoms sequence generation. In Proceedings of the AAAI Conference on Artificial Intelligence, vol. 36, 4432–4440 (2022).
  26. Semi-supervised classification with graph convolutional networks. \JournalTitlearXiv preprint arXiv:1609.02907 (2016).
  27. Veličković, P. et al. Graph attention networks. \JournalTitlearXiv preprint arXiv:1710.10903 (2017).
  28. A generalization of transformer networks to graphs. \JournalTitleArXiv abs/2012.09699 (2020).
  29. Rampášek, L. et al. Recipe for a general, powerful, scalable graph transformer. \JournalTitlearXiv preprint arXiv:2205.12454 (2022).
  30. Simple spectral graph convolution. In International conference on learning representations (2021).
  31. Training graph neural networks with 1000 layers. In International conference on machine learning, 6437–6449 (PMLR, 2021).
  32. How attentive are graph attention networks? \JournalTitlearXiv preprint arXiv:2105.14491 (2021).
  33. Zhang, Z. et al. Ernie: Enhanced language representation with informative entities. \JournalTitlearXiv preprint arXiv:1905.07129 (2019).
  34. Qa-gnn: Reasoning with language models and knowledge graphs for question answering. \JournalTitlearXiv preprint arXiv:2104.06378 (2021).
  35. Zhang, X. et al. Greaselm: Graph reasoning enhanced language models for question answering. \JournalTitlearXiv preprint arXiv:2201.08860 (2022).
  36. Yasunaga, M. et al. Deep bidirectional language-knowledge graph pretraining. \JournalTitlearXiv preprint arXiv:2210.09338 (2022).
  37. Finding structural knowledge in multimodal-bert. \JournalTitlearXiv preprint arXiv:2203.09306 (2022).
  38. Liu, J. et al. Generated knowledge prompting for commonsense reasoning. \JournalTitlearXiv preprint arXiv:2110.08387 (2021).
  39. The methodology of dynamic uncertain causality graph for intelligent diagnosis of vertigo. \JournalTitleComputer methods and programs in biomedicine 113, 162–174 (2014).
  40. The cubic dynamic uncertain causality graph: A methodology for temporal process modeling and diagnostic logic inference. \JournalTitleIEEE Transactions on Neural Networks and Learning Systems 31, 4239–4253, DOI: 10.1109/TNNLS.2019.2953177 (2020).
  41. The application of dynamic uncertain causality graph based diagnosis and treatment unification model in the intelligent diagnosis and treatment of hepatitis b. \JournalTitleSymmetry 13, 1185 (2021).
  42. Zhong, C. et al. Hierarchical reinforcement learning for automatic disease diagnosis. \JournalTitleBioinformatics (2022).
  43. Xu, L. et al. End-to-end knowledge-routed relational dialogue system for automatic diagnosis. In Proceedings of the AAAI Conference on Artificial Intelligence, vol. 33, 7346–7353 (2019).
  44. Yan, G. et al. M^ 2-meddialog: A dataset and benchmarks for multi-domain multi-service medical dialogues. \JournalTitlearXiv preprint arXiv:2109.00430 (2021).
  45. Zeng, G. et al. Meddialog: Large-scale medical dialogue dataset. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) (2020).
  46. Liu, W. et al. Meddg: A large-scale medical consultation dataset for building medical dialogue system. \JournalTitleCoRR abs/2010.07497 (2020). 2010.07497.
  47. Statistical methods for rates and proportions (john wiley & sons, 2013).
  48. Self-alignment pretraining for biomedical entity representations. \JournalTitlearXiv preprint arXiv:2010.11784 (2020).
  49. Bert for joint intent classification and slot filling. \JournalTitlearXiv preprint arXiv:1902.10909 (2019).
  50. Deep learning on graphs: A survey. \JournalTitleIEEE Transactions on Knowledge and Data Engineering (2020).
  51. Neural machine translation by jointly learning to align and translate. \JournalTitlearXiv preprint arXiv:1409.0473 (2014).
  52. Linkbert: Pretraining language models with document links. \JournalTitlearXiv preprint arXiv:2203.15827 (2022).
  53. Zhang, S. et al. Knowledge-rich self-supervision for biomedical entity linking. In Findings of the Association for Computational Linguistics: EMNLP 2022, 868–880 (2022).
  54. A knowledge infused context driven dialogue agent for disease diagnosis using hierarchical reinforcement learning. \JournalTitleKnowledge-Based Systems 242, 108292 (2022).
  55. Yuan, Z. et al. Coder: Knowledge-infused cross-lingual medical term embedding for term normalization. \JournalTitleJournal of biomedical informatics 126, 103983 (2022).
  56. Welch, B. L. The generalization ofstudent’s’ problem when several different population variances are involved. \JournalTitleBiometrika 34, 28–35 (1947).
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (3)
  1. Mohit Tomar (2 papers)
  2. Abhisek Tiwari (6 papers)
  3. Sriparna Saha (48 papers)
Citations (1)
View on Semantic Scholar