Explainable solutions from artificial intelligence for health-care support systems

Abstract

For decades, efforts to standardize medical care have struggled to fundamentally reduce errors and unjustified variations in medical practice, largely due to the influence of the human factor. The formalization of clinical guidelines and computer-assisted interpretation makes it possible to provide decision-support tools to improve health-care quality. They can better influence clinician behavior than narrative guidelines. Medical ontologies and algorithms based on such ontologies allow the interpretation of formalized clinical documents (guidelines). To support health professionals as consultants, systems must provide reliable knowledge and rely on approaches explicitly explaining their recommendations. Integrating software engineering, knowledge engineering, and artificial intelligence advancements can provide health-care professionals with computer-interpretable clinical guidelines. These should be decision-support complexes combined under a common terminological framework capable of understanding patient health documents. The research focuses on an emerging concept of manufacturing systems working with digital clinical guidelines. The paper presents an architectural principle, a new technology for creating viable clinical decision support systems. It presents a development environment for constructing and controlling the system’s improvements. The main contributions of the study include the automation of multiple physician tasks by filling a single structured “medical history,” integration of formalized knowledge from clinical guidelines and other reliable sources to satisfy both the relevance of the methods used and personalization to patient, transparency of all applicable knowledge, explainability of advice based on the essence of the knowledge and linked to the source, and the integrability of decision-support complexes with neural network services, capable of inputting data from a structured medical history.

Keywords

Explanatory decision support system / Knowledge-enabled system / Interpretable clinical guideline / Knowledge ontology / Viable system

Funding

The work was supported by the Ministry of Education and Science of Russia (No.: FWFW-2021-0004).

Conflict of interest

The authors declare they have no competing interests.

References

[1] Peleg M. Computer-interpretable clinical guidelines: A methodological review. J Biomed Inform. 2013; 46(4):744-763. doi: 10.1016/j.jbi.2013.06.009

[2] Young O, Shahar Y, Liel Y, et al. Runtime application of Hybrid-Asbru clinical guidelines. J Biomed Inform. 2007; 40(5):507-526. doi: 10.1016/j.jbi.2006.12.004

[3] Novais P, Oliveira T, Satoh K, Neves J. The role of ontologies and decision frameworks in computer-interpretable guideline execution. In: Nalepa G, Baumeister J, editors. Synergies between Knowledge Engineering and Software Engineering. Advances in Intelligent Systems and Computing. Vol. 626. Cham: Springer; 2018. p. 197-216.

[4] Deng C, Ji X, Rainey C, Zhang J, Lu W. Integrating machine learning with human knowledge. iScience. 2020; 23(11):101656. doi: 10.1016/j.isci.2020.101656

[5] Müller L, Gangadharaiah R, Klein SC, et al. An open access medical knowledge base for community driven diagnostic decision support system development. BMC Med Inform Decis Mak. 2019; 19:93. doi: 10.1186/s12911-019-0804-1

[6] Musen M. The protégé project: A look back and a look forward. AI Matters. 2015; 1(4):4-12. doi: 10.1145/2757001.2757003

[7] Mortensen J, Minty E, Januszyk M, et al. Using the wisdom of the crowds to find critical errors in biomedical ontologies: A study of SNOMED CT. J Am Med Inform Assoc. 2015; 22(3):640-648. doi: 10.1136/amiajnl-2014-002901

[8] Loh HW, Ooi CP, Seoni S, Barua PD, Molinari F, Acharya UR. Application of explainable artificial intelligence for healthcare: A systematic review of the last decade (2011-2022). Comput Methods Programs Biomed. 2022; 226:107161. doi: 10.1016/j.cmpb.2022.107161

[9] Yang G, Ye Q, Xia J. Unbox the black-box for the medical explainable AI via multi-modal and multi-centre data fusion: A mini-review, two showcases and beyond. Inf Fusion. 2022; 77:29-52. doi: 10.48550/arXiv.2102.01998

[10] Gartner Identifies the Top Strategic Technology Trends for 2021. STAMFORD, Conn; 2020. Available from: https://www.gartner.com/en/newsroom/press-releases/2020-10-19-gartner-identifies-the-top-strategic-technology-trends-for-2021

[11] Pressman RS. Architectural Design, Software Engineering: Practitioner’s Approach. 7th ed. New York, NY, USA: McGraw-Hill; 2010. p. 242-275.

[12] Islam M, Katiyar V. Development of a software maintenance cost estimation model: 4th GL perspective. Int J Techn Res Appl. 2014; 2(6):65-68.

[13] Izurieta C, Bieman JM. A multiple case study of design pattern decay, grime, and rot in evolving software systems. Software Qual J. 2013; 21(2):289-323. doi: 10.1007/s11219-012-9175-x

[14] Breivold HP, Crnkovic I, Eriksson PJ. Analyzing Software Evolvability. COMPSAC 2008: 32nd Annual IEEE International Computer Software and Applications Conference. Turku, Finland; 2008. p. 327-330.

[15] Finn VK. About Data Mining. Artificial Intelligence News. (In Russ.); 2004. p.3-18. Available from: https://masters.donntu.ru/2006/kita/balabanov/library/articles/art010.pdf

[16] Weissler EH, Naumann T, Andersson T, et al. The role of machine learning in clinical research: Transforming the future of evidence generation. Trials. 2021; 22:537. doi: 10.1186/s13063-021-05489-x

[17] Lin Z, Cheng YT, Cheung BMY. Machine learning algorithms identify hypokalaemia risk in people with hypertension in the United States National Health and Nutrition Examination Survey 1999-2018. Ann Med. 2023; 55(1):2209336. doi: 10.1080/07853890.2023.2209336

[18] Kang SH, Baek H, Cho J, et al. Management of cardiovascular disease using an mHealth tool: A randomized clinical trial. NPJ Digit Med. 2021; 4:165. doi: 10.1038/s41746-021-00535-z

[19] Zhang Y, Hu N, Li Z, Ji X, et al. Lumbar Spine Localisation Method Based on Feature Fusion. In: CAAI Transactions on Intelligence Technology 2022. doi: 10.1049/cit2.12137

[20] Wang Y, Li S, Zhao B, et al. A ResNet-based approach for accurate radiographic diagnosis of knee osteoarthritis. CAAI Trans Intell Technol. 2022; 7(3):512-521. doi: 10.1049/cit2.12079

[21] Golenkov VV, Gulyakina NA, Grakova NV, Nikulenka VY, Eremeev AP, Tarasov VB. From training intelligent systems to training their development means. Open Semantic Technol Intell Syst. 2018; 2:81-98.

[22] Nikolenko SI, Tulupiev AL. Self-learning Systems. Moscow, Russia: MCNMO; 2009. p. 288. (In Russ).

[23] Dodonov AG, Landje DV. Viability of Information Systems. Kiev, Ukraine: Naukova Dumka; 2011. p. 256. (In Russ)

[24] Kryazhych OO. Ensuring the viability of the information in time under its processing in the DSS. Math Mach Syst. 2015; 2:170-176. (In Russ)

[25] Gribova VV, Kleshchev AS. Paradigm for controlled intelligent systems. Manag Syst Inform Technol. 2016; 3(65):32-38. (In Russ)

[26] Chhabra JK. Improving modular structure of software system using structural and lexical dependency. Inform Softw Technol. 2017; 82:96-120. doi: 10.1016/j.infsof.2016.09.011

[27] Gribova VV, Kleschev AS, Moskalenko FM, Timchenko VA, Shalfeeva EA. Extensible toolkit for the development of viable systems with knowledge bases. Softw Eng. 2018; 9(8):339-348. doi: 10.17587/prin.9.339-348

[28] Gribova V, Shalfeeva E, Petryaeva M. Cloud infrastructure for creation of interpretable diagnostic knowledge bases of diseases regardless their etiology. Atlantis Highlights Comput Sci. 2019; 3:79-82. doi: 10.2991/csit-19.2019.13

[29] Kleschev AS, Chernyakhovskaya MY, Shalfeeva EA. The paradigm of an intellectual professional activity automation. Part 1. The features of an intellectual professional activity. Ontol Design. 2013; 3(9):53-69. (In Russ).

[30] Gribova V, Moskalenko P, Petryaeva M, Okun D. Cloud environment for development and use of software systems for clinical medicine and education. Adv Intell Syst Res. 2019; 166:225-229. doi: 10.2991/itids-19.2019.40