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

Enabling Collaborative Clinical Diagnosis of Infectious Keratitis by Integrating Expert Knowledge and Interpretable Data-driven Intelligence (2401.08695v1)

Published 14 Jan 2024 in cs.AI, cs.CV, and cs.HC

Abstract: Although data-driven AI in medical image diagnosis has shown impressive performance in silico, the lack of interpretability makes it difficult to incorporate the "black box" into clinicians' workflows. To make the diagnostic patterns learned from data understandable by clinicians, we develop an interpretable model, knowledge-guided diagnosis model (KGDM), that provides a visualized reasoning process containing AI-based biomarkers and retrieved cases that with the same diagnostic patterns. It embraces clinicians' prompts into the interpreted reasoning through human-AI interaction, leading to potentially enhanced safety and more accurate predictions. This study investigates the performance, interpretability, and clinical utility of KGDM in the diagnosis of infectious keratitis (IK), which is the leading cause of corneal blindness. The classification performance of KGDM is evaluated on a prospective validation dataset, an external testing dataset, and an publicly available testing dataset. The diagnostic odds ratios (DOR) of the interpreted AI-based biomarkers are effective, ranging from 3.011 to 35.233 and exhibit consistent diagnostic patterns with clinic experience. Moreover, a human-AI collaborative diagnosis test is conducted and the participants with collaboration achieved a performance exceeding that of both humans and AI. By synergistically integrating interpretability and interaction, this study facilitates the convergence of clinicians' expertise and data-driven intelligence. The promotion of inexperienced ophthalmologists with the aid of AI-based biomarkers, as well as increased AI prediction by intervention from experienced ones, demonstrate a promising diagnostic paradigm for infectious keratitis using KGDM, which holds the potential for extension to other diseases where experienced medical practitioners are limited and the safety of AI is concerned.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (49)
  1. McKinney, S. M. et al. International evaluation of an ai system for breast cancer screening. \JournalTitleNature 577, 89–94, DOI: 10.1038/s41586-019-1799-6 (2020).
  2. Tschandl, P. et al. Human–computer collaboration for skin cancer recognition. \JournalTitleNature Medicine 26, 1229–1234, DOI: 10.1038/s41591-020-0942-0 (2020).
  3. Bai, X. et al. Advancing covid-19 diagnosis with privacy-preserving collaboration in artificial intelligence. \JournalTitleNature Machine Intelligence 3, 1081–1089, DOI: 10.1038/s42256-021-00421-z (2021).
  4. Liu, X. et al. A comparison of deep learning performance against health-care professionals in detecting diseases from medical imaging: A systematic review and meta-analysis. \JournalTitleThe Lancet Digital Health 1, e271–e297, DOI: 10.1016/S2589-7500(19)30123-2 (2019).
  5. Vasey, B. et al. Association of clinician diagnostic performance with machine learning–based decision support systems: A systematic review. \JournalTitleJAMA Network Open 4, e211276, DOI: 10.1001/jamanetworkopen.2021.1276 (2021).
  6. Explainable artificial intelligence (xai) in deep learning-based medical image analysis. \JournalTitleMedical Image Analysis 79, 102470, DOI: 10.1016/j.media.2022.102470 (2022).
  7. Infectious keratitis: An update on epidemiology, causative microorganisms, risk factors, and antimicrobial resistance. \JournalTitleEye 35, 1084–1101, DOI: 10.1038/s41433-020-01339-3 (2021).
  8. Infectious keratitis: A review. \JournalTitleClinical & Experimental Ophthalmology 50, 543–562, DOI: 10.1111/ceo.14113 (2022).
  9. Clemens, L. E. et al. Designed host defense peptides for the treatment of bacterial keratitis. \JournalTitleInvestigative Ophthalmology & Visual Science 58, 6273–6281, DOI: 10.1167/iovs.17-22243 (2017).
  10. Juárez, M. M. et al. Acanthamoeba in the eye, can the parasite hide even more? latest developments on the disease. \JournalTitleContact Lens & Anterior Eye: The Journal of the British Contact Lens Association 41, 245–251, DOI: 10.1016/j.clae.2017.12.017 (2018).
  11. Dalmon, C. et al. The clinical differentiation of bacterial and fungal keratitis: A photographic survey. \JournalTitleInvestigative Opthalmology & Visual Science 53, 1787, DOI: 10.1167/iovs.11-8478 (2012).
  12. Redd, T. K. et al. Expert performance in visual differentiation of bacterial and fungal keratitis. \JournalTitleOphthalmology 129, 227–230, DOI: 10.1016/j.ophtha.2021.09.019 (2022).
  13. Achieving diagnostic excellence for infectious keratitis: A future roadmap. \JournalTitleFrontiers in Microbiology 13, 1020198, DOI: 10.3389/fmicb.2022.1020198 (2022).
  14. Xu, Y. et al. Deep sequential feature learning in clinical image classification of infectious keratitis. \JournalTitleEngineering 7, 1002–1010, DOI: 10.1016/j.eng.2020.04.012 (2021).
  15. Deep learning for discrimination between fungal keratitis and bacterial keratitis: Deepkeratitis. \JournalTitleCornea 41, 616–622, DOI: 10.1097/ICO.0000000000002830 (2022).
  16. Redd, T. K. et al. Image-based differentiation of bacterial and fungal keratitis using deep convolutional neural networks. \JournalTitleOphthalmology Science 2, DOI: 10.1016/j.xops.2022.100119 (2022).
  17. Very deep convolutional networks for large-scale image recognition, DOI: 10.48550/arXiv.1409.1556. 1409.1556.
  18. Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770–778, DOI: 10.1109/CVPR.2016.90 (IEEE, Las Vegas, NV, USA, 2016).
  19. Densely connected convolutional networks. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2261–2269, DOI: 10.1109/CVPR.2017.243 (IEEE, Honolulu, HI, 2017).
  20. Chen, C. et al. This looks like that: Deep learning for interpretable image recognition. In Advances in Neural Information Processing Systems, vol. 32 (Curran Associates, Inc., 2019).
  21. Evidential deep learning to quantify classification uncertainty. In Advances in Neural Information Processing Systems, vol. 31 (Curran Associates, Inc., 2018).
  22. Koh, P. W. et al. Concept bottleneck models. In Proceedings of the 37th International Conference on Machine Learning, 5338–5348 (PMLR, 2020).
  23. Dosovitskiy, A. et al. An image is worth 16x16 words: Transformers for image recognition at scale, DOI: 10.48550/arXiv.2010.11929 (2021). 2010.11929.
  24. Chua, M. et al. Tackling prediction uncertainty in machine learning for healthcare. \JournalTitleNature Biomedical Engineering DOI: 10.1038/s41551-022-00988-x (2022).
  25. Update on acanthamoeba keratitis: Diagnosis, treatment, and outcomes. \JournalTitleCornea 35, 713–720, DOI: 10.1097/ICO.0000000000000804 (2016).
  26. Current thoughts in fungal keratitis: Diagnosis and treatment. \JournalTitleCurrent Fungal Infection Reports 7, 209–218, DOI: 10.1007/s12281-013-0150-110.1007/s12281-013-0150-1 (2013).
  27. Mycotic keratitis: Epidemiology, diagnosis and management. \JournalTitleClinical Microbiology and Infection: The Official Publication of the European Society of Clinical Microbiology and Infectious Diseases 19, 210–220, DOI: 10.1111/1469-0691.12126 (2013).
  28. Variable kernel density estimation. \JournalTitleThe Annals of Statistics 1236–1265 (1992).
  29. Bansal, G. et al. Does the whole exceed its parts? the effect of ai explanations on complementary team performance, DOI: 10.48550/arXiv.2006.14779 (2021). 2006.14779.
  30. Inkpen, K. et al. Advancing human-ai complementarity: The impact of user expertise and algorithmic tuning on joint decision making (2022). arXiv:2208.07960.
  31. Patel, B. N. et al. Human–machine partnership with artificial intelligence for chest radiograph diagnosis. \JournalTitlenpj Digital Medicine 2, 1–10, DOI: 10.1038/s41746-019-0189-7 (2019).
  32. Ahn, J. S. et al. Association of artificial intelligence–aided chest radiograph interpretation with reader performance and efficiency. \JournalTitleJAMA Network Open 5, e2229289, DOI: 10.1001/jamanetworkopen.2022.29289 (2022).
  33. Leibig, C. et al. Combining the strengths of radiologists and ai for breast cancer screening: A retrospective analysis. \JournalTitleThe Lancet Digital Health 4, e507–e519, DOI: 10.1016/S2589-7500(22)00070-X (2022).
  34. Human-ai collaboration in healthcare: A review and research agenda. In Hawaii International Conference on System Sciences, DOI: 10.24251/HICSS.2021.046 (2021).
  35. Xprotonet: Diagnosis in chest radiography with global and local explanations. In 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 15714–15723, DOI: 10.1109/CVPR46437.2021.01546 (IEEE, Nashville, TN, USA, 2021).
  36. Barnett, A. J. et al. A case-based interpretable deep learning model for classification of mass lesions in digital mammography. \JournalTitleNature Machine Intelligence 3, 1061–1070, DOI: 10.1038/s42256-021-00423-x (2021).
  37. OpenAI. Gpt-4 technical report, DOI: 10.48550/arXiv.2303.08774 (2023). arXiv:2303.08774.
  38. Kirillov, A. et al. Segment anything, DOI: 10.48550/arXiv.2304.02643 (2023). arXiv:2304.02643.
  39. Koyama, A. et al. Determination of probability of causative pathogen in infectious keratitis using deep learning algorithm of slit-lamp images. \JournalTitleScientific Reports 11, 22642, DOI: 10.1038/s41598-021-02138-w (2021).
  40. Kuo, M.-T. et al. A deep learning approach in diagnosing fungal keratitis based on corneal photographs. \JournalTitleScientific Reports 10, 14424, DOI: 10.1038/s41598-020-71425-9 (2020).
  41. Kuo, M.-T. et al. Deep learning approach in image diagnosis of pseudomonas keratitis. \JournalTitleDiagnostics 12, 2948, DOI: 10.3390/diagnostics12122948 (2022).
  42. Li, Z. et al. Preventing corneal blindness caused by keratitis using artificial intelligence. \JournalTitleNature Communications 12, 3738, DOI: 10.1038/s41467-021-24116-6 (2021).
  43. Rudin, C. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. \JournalTitleNature Machine Intelligence 1, 206–215, DOI: 10.1038/s42256-019-0048-x (2019).
  44. Liu, Z. et al. Swin transformer: Hierarchical vision transformer using shifted windows, DOI: 10.48550/arXiv.2103.14030 (2021). 2103.14030.
  45. Aligning language models to follow instructions. https://openai.com/research/instruction-following.
  46. Decoding artificial intelligence to achieve diagnostic excellence: Learning from experts, examples, and experience. \JournalTitleJAMA 328, 709, DOI: 10.1001/jama.2022.13735 (2022).
  47. Jøsang, A. Subjective Logic. Artificial Intelligence: Foundations, Theory, and Algorithms (Springer International Publishing, Cham, 2016).
  48. Zeydan, B. et al. Imaging biomarkers of alzheimer disease in multiple sclerosis. \JournalTitleAnnals of Neurology 87, 556–567, DOI: 10.1002/ana.25684 (2020).
  49. Paszke, A. et al. Pytorch: An imperative style, high-performance deep learning library. In Advances in Neural Information Processing Systems, vol. 32 (Curran Associates, Inc., 2019).

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets