Abstract
This article describes the principles of creating an intelligent system using a JSM method of automated research support (JSM method ARS) to predict the feasibility of computed tomography application. The procedures of JSM research that are adopted to increase the reliability of the empirical regularities obtained in the system are described. The obtained empirical regularities and their expert ratings are given.
Notes
The minimum hypotheses refer to the hypotheses such that no other hypothesis obtained consists of a subset of the features of these hypotheses. Formal description of minimum hypotheses is given, for example, in [14].
6In this case”, “–” designates the operation of Boolean difference.
REFERENCES
Shesternikova, O.P., Finn, V.K., Vinokurova, L.V., Les’ko, K.A., Varvanina, G.G., and Tyulyaeva, E.Yu., An intelligent system for diagnostics of pancreatic diseases, Autom. Doc. Math. Linguist., 2019, vol. 53, no. 5, pp. 288–294. https://doi.org/10.3103/s000510551905008x
Vinokurova, L.V., Finn, V.K., Les’ko, K.A., Shesternikova, O.P., and Varvanina, G.G., First results of studying the capabilities of the intelligent system of automated research support in differential diagnostics of pancreas cancer, Materialy 47-i nauchnoi sessii TsNII gastroenterologii Gastroenterologiya v menyayushcheisya real’nosti (Proc. 47th Sci. Session of the Central Research Institute of Gastroenterology on Gastroenterology in Changing Reality), Moscow: Terapevticheskii Arkhiv, 2021, p. 93.
Zhang, J., Zuo, Ch.-J., Jia, N.-Ya., Wang, J.-H., Hu, Sh.-P., Yu, Zh.-F., Zheng, Yu., Zhang, A., and Feng, X.-Yu., Cross-modality PET/CT and contrast-enhanced CT imaging for pancreatic cancer, World J. Gastroenterol., 2015, vol. 21, no. 10, pp. 2988–2996. https://doi.org/10.3748/wjg.v21.i10.2988
Sostoyanie onkologicheskoi pomoshchi naseleniyu Rossii v 2019 godu (State-of-the-art of the Oncological Treatment to the Russian Population in 2019), Kaprin, A.D., Starinskii, V.V., and Shakhzadova, A.O., Eds., Moscow: MNIOI im. P.A. Gertsena, 2020.
Sung, H., Ferlay, J., Siegel, R.L., Laversanne, M., Soerjomataram, I., Jemal, A., and Bray, F., Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries, CA: A Cancer J. Clinicians, 2021, vol. 71, no. 3, pp. 209–249. https://doi.org/10.3322/caac.21660
Baranov, A.A., Namazova-Baranova, L.S., Smirnov, I.V., Devyatkin, D.A., Shelmanov, A.O., Vishneva, E.A., Antonova, E.V., Smirnov, V.I., and Latyshev, A.V., Methods and tools of complex intelligent analysis of medical data, Tr. Inst. Sist. Anal. Ross. Akad. Nauk, 2015, vol. 65, no. 2, pp. 81–93. https://doi.org/10.15690/vramn663
Luger, G.F., Artificial Intelligence: Structures and Strategies for Complex Problem Solving, Addison-Wesley, 2002, 4th ed.
Linden, A. and Yarnold, P.R., Using data mining techniques to characterize participation in observational studies, J. Eval. Clin. Pract., 2016, vol. 22, no. 6, pp. 839–847. https://doi.org/10.1111/jep.12515
Yu, E., Wesselius, A., Sinhart, C., Wolk, A., Stern, M., Jiang, X., Tang, L., Marshall, J., Kellen, E., Van Den Brandt, P., Lu, C., Pohlabeln, H., Steineck, G., Allam, M., Karagas, M., La Vecchia, C., Porru, S., Carta, A., Golka, K., Johnson, K., Benhamou, S., Zhang, Z., Bosetti, C., Taylor, J., Weiderpass, E., Grant, E., White, E., Polesel, J., and Zeegers, M., A data mining approach to investigate food groups related to incidence of bladder cancer in the BLadder cancer Epidemiology and Nutritional Determinants International Study, Br. J. Nutr., 2020, vol. 124, no. 6, pp. 611–619. https://doi.org/10.1017/s0007114520001439
Khan, S., Mohan, H., Liu, Yi., Batchuluun, B., Gohil, H., Al Rijjal, D., Manialawy, Yo., Cox, B.J., Gunderson, E.P., and Wheeler, M.B., The discovery of novel predictive biomarkers and early-stage pathophysiology for the transition from gestational diabetes to type 2 diabetes, Diabetologia, 2019, vol. 62, no. 4, pp. 687–703. https://doi.org/10.1007/s00125-018-4800-2
Finn, V.K. and Shesternikova, O.P., The heuristics of detection of empirical regularities by JSM reasoning, Autom. Doc. Math. Linguist., 2018, vol. 52, no. 5, pp. 215–247. https://doi.org/10.3103/s0005105518050023
Finn, V.K., Distributive lattices of inductive JSM procedures, Autom. Doc. Math. Linguist., 2014, vol. 48, no. 6, pp. 265–295. https://doi.org/10.3103/S0005105514060028
Pankratova, E.S. and Vinogradov, D.V., Formal description of adaptation of intelligent JSM-systems for clinical and laboratory data analysis, Autom. Doc. Math. Linguist., 2011, vol. 45, no. 5, pp. 213–217. https://doi.org/10.3103/s0005105511050049
Gusakova, S.M. and Mikheyenkova, M.A., Knowledge discovery as a tool for the formation of the structure of society, Autom. Doc. Math. Linguist., 2016, vol. 50, no. 5, pp. 179–188. https://doi.org/10.3103/S0005105516050022
Klein, A.P., Pancreatic cancer epidemiology: Understanding the role of lifestyle and inherited risk factors, Nat. Rev. Gastroenterol. Hepatol., 2016, vol. 18, no. 7, pp. 493–502. https://doi.org/10.1038/s41575-021-00457-x
Midha, S., Chawla, S., and Garg, P.K., Modifiable and non-modifiable risk factors for pancreatic cancer: A review, Cancer Lett., 2016, vol. 381, no. 1, pp. 269–277. https://doi.org/10.1016/j.canlet.2016.07.022
Finn, V.K., Exact epistemology and artificial intelligence, Autom. Doc. Math. Linguist., 2020, vol. 54, no. 3, pp. 140–173. https://doi.org/10.3103/S0005105520030073
Funding
This work was supported by ongoing institutional funding. No additional grants to carry out or direct this particular research were obtained.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
The authors declare that they have no conflict of interest.
Additional information
Translated by L. Solovyova
Publisher’s Note.
Allerton Press remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
APPENDICES
APPENDICES
Appendix 1. Similarity Predicate Definitions
\(M_{{a,n}}^{\sigma }(V,W)\) (σ = +, –, 0, τ)| are minimal predicates that implement inductive inferences. They can be supplemented with more informative improvementsFootnote 26
Ban on counterexamples:
\({{(b)}^{ + }}\,\,\forall \,\,X\,\,\forall \,\,Y(((\begin{array}{*{20}{c}} V&X \end{array})\& {\text{(}}\begin{array}{*{20}{c}} W&Y \end{array})) \to \)\( \to ({{J}_{{(1,n{\text{)}}}}}(X \Rightarrow \,{{\,}_{1}}Y) \vee {{J}_{{(\tau ,n)}}}(X \Rightarrow \,{{\,}_{1}}Y))),\)\({{{\text{(}}b)}^{ + }}\,\,\forall \,\,X\,\,\forall \,\,Y(((\begin{array}{*{20}{c}} V&X \end{array})\& {\text{(}}\begin{array}{*{20}{c}} W&Y \end{array})) \to \) \( \to ({{J}_{{( - 1,n{\text{)}}}}}(X \Rightarrow \,{{\,}_{1}}Y) \vee {{J}_{{(\tau ,n)}}}(X \Rightarrow \,{{\,}_{1}}Y))),\)
\(M_{{ab,n}}^{ + }(V,W)\,\,\,\mathop \to \limits^ \leftarrow {\kern 1pt} M_{{a,n}}^{ + }(V,W)\& {{(b)}^{ + }},\) \(M_{{ab,n}}^{ - }(V,W)\mathop \to \limits^ \leftarrow {\kern 1pt} {\kern 1pt} M_{{a,n}}^{ - }(V,W)\& {{(b)}^{ - }},\)
Method of difference:
\({{({{d}_{0}})}^{ + }}\,\,\forall \,\,X\,\,\forall \,\,Y\,\,\forall \,\,Z\,\,\forall \,\,U(({{J}_{{(1,n{\text{)}}}}}(X \Rightarrow \,{{\,}_{1}}Y)\) \(\& (W\,\,Y)\& (V\,\,X)\& ((X\,\, - V)\,\,Z)\) \(\& ((X\,\, - V) \ne \emptyset )\& \neg (V\,\,Z)) \to \) \( \to \neg {{J}_{{{\text{(1}}{\text{,n)}}}}}(Z \Rightarrow \,{{\,}_{1}}U) \vee \neg (W\,\,U))){{,}^{6}}\)
\({{({{d}_{0}})}^{ - }}\,\,\forall \,\,X\,\,\forall \,\,Y\,\,\forall \,\,Z\,\,\forall \,\,U(({{J}_{{( - 1,n{\text{)}}}}}(X \Rightarrow \,{{\,}_{1}}Y)\) \(\& (W\,\,Y)\& (V\,\,X)\& ((X\,\, - V)\,\,Z)\) \(\& ((X\,\, - V) \ne \emptyset )\& \neg (V\,\,Z)) \to \)\( \to \neg {{J}_{{{\text{(}} - {\text{1}}{\text{,}}n{\text{)}}}}}(Z \Rightarrow \,{{\,}_{1}}U) \vee \neg (\begin{array}{*{20}{c}} W&U \end{array}))),\)
\(M_{{ad0,n}}^{ + }(V,W)\,\,\, \leftarrow \to {\kern 1pt} M_{{a,n}}^{ + }(V,W)\& {{({{d}_{0}})}^{ + }},\) \(M_{{ad0,n}}^{ - }(V,W)\,\,\, \leftarrow \to {\kern 1pt} M_{{a,n}}^{ - }(V,W)\& {{({{d}_{0}})}^{ - }}.\)
Method of difference with a ban on counterexamples:
\(M_{{ad0b,n}}^{ + }(V,W)\,\,\, \leftarrow \to {\kern 1pt} M_{{a,n}}^{ + }(V,W)\& {{(b)}^{ + }}\& {{({{d}_{0}})}^{ + }},\) \(M_{{ad0b,n}}^{ - }(V,W)\,\,\, \leftarrow \to {\kern 1pt} M_{{a,n}}^{ - }(V,W)\& {{(b)}^{ - }}\& {{({{d}_{0}})}^{ - }}.\)
Appendix 2. List of Regularities Obtained
Signs and abbreviations:
• Ages are divided into decades: 3: 30–39, 4: 40–49, 5: 50–59, 6: 60–69, 7: 70–79, 8: 80–89;
• Alcohol (alcohol addiction) only has yes or no values;
• Tobacco smoking (tobacco addiction) only has yes or no values;
• Body mass index (BMI) has the following values: Below: below the norm, Norm: normal, Above: above the norm;
• The duration of the disease was compared across key points: 1 year, 2 years, 3 years, 5 years;
• Pain (severity of pain) has the possible values yes (severe pain) and no (mild pain);
• Diabetes mellitus (DM) only has yes or no values.
k is the number of “parents” of the hypothesis (examples that generate it).
If the corresponding regularity is generated by several strategies, the largest one was chosen (since the strategies by which the regularities were obtained in our experiments turned out to be comparable). Thus, each empirical regularity was assigned one rating.
Experiment 1. Positive regularities | ||||||||||
〈b, Strab,ab〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 (Strong) | DM | k | Expert’s rating | |
2105 | Norm | <3 Years | Yes | 17 | 3 | |||||
2168 | No | <3 Years | Yes | 16 | 3 | |||||
132 | No | <2 Years | Yes | 13 | 3 | |||||
205 | <1 Years | Yes | 12 | 3 | ||||||
2114 | 5 | <3 Years | Yes | 6 | 3 | |||||
3145 | No | No | From 1 to 2 years | 6 | 3 | |||||
3321 | 6 | <2 Years | Yes | 6 | 3 | |||||
2178 | 5 | No | Norm | Yes | 5 | 3 | ||||
2538 | 5 | No | Yes | No | 4 | 3 | ||||
2548 | 5 | Yes | <3 Years | No | 4 | 3 | ||||
3028 | 5 | From 1 to 3 years | No | 4 | 3 | |||||
3314 | 6 | Yes | <5 Years | No | 4 | 3 | ||||
3315 | Yes | <1 Years | No | 4 | 3 | |||||
9 | Yes | No | <5 Years | Yes | No | 3 | 3 | |||
1520 | 5 | Norm | Yes | Yes | 3 | 3 | ||||
〈b, Strab,abd0〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 (Strong) | DM | k | Expert’s rating | |
1537 | 5 | No | Yes | |||||||
〈a, Strab,ab〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 (Strong) | DM | k | Expert’s rating | |
33 | Norm | Yes | 23 | 3 | ||||||
689 | Norm | <2 Years | 19 | 2 | ||||||
207 | No | <2 Years | 18 | 3 | ||||||
1089 | No | Yes | 18 | 2 | ||||||
1280 | No | <5 Years | Yes | 17 | 2 | |||||
1832 | No | <2 Years | 16 | 3 | ||||||
1267 | No | Yes | No | 15 | 2 | |||||
1837 | <1 Years | 14 | 3 | |||||||
795 | 5 | <5 Years | 9 | 2 | ||||||
794 | Yes | Norm | <5 Years | 8 | 2 | |||||
1178 | 6 | <5 Years | Yes | 8 | 3 | |||||
1428 | 5 | No | 8 | 2 | ||||||
2120 | >2 Years | Yes | No | 8 | 1 | |||||
3618 | 6 | <3 Years | No | 8 | 2 | |||||
1834 | 6 | No | <5 Years | No | 7 | 3 | ||||
1644 | Yes | No | No | 6 | 2 | |||||
2552 | No | Yes | <3 Years | 6 | 3 | |||||
3609 | 6 | Norm | <3 Years | 6 | 3 | |||||
69 | No | Above | No | 5 | 2 | |||||
681 | 5 | Yes | Norm | 5 | 3 | |||||
3411 | 4 | Yes | >1 Years | No | 5 | 1 | ||||
2276 | No | Above | No | 4 | 1 | |||||
3314 | 6 | Yes | <5 Years | No | 4 | 2 | ||||
3424 | 4 | Yes | >1 Years | 4 | 1 | |||||
〈n, Strab,abd0〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 (Strong) | DM | k | Expert’s rating | |
23073 | 6 | Yes | From 2 to 5 years | No | 3 | 2 | ||||
〈a, Strab,abd0〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 | DM | k | Expert’s rating | |
694 | Norm | <5 Years | 23 | 1 | ||||||
1265 | No | Yes | 21 | 2 | ||||||
257 | No | <5 Years | 20 | 2 | ||||||
4 | No | No | 19 | 1 | ||||||
802 | No | <5 Years | 19 | 2 | ||||||
2744 | No | No | Norm | 11 | 1 | |||||
1843 | 6 | <5 Years | No | 10 | 2 | |||||
2103 | >2 Years | Yes | 10 | 1 | ||||||
76 | 5 | Yes | 8 | 2 | ||||||
54 | 5 | Yes | 7 | 2 | ||||||
1274 | 4 | >1 Years | No | 6 | 1 | |||||
2154 | From 2 to 5 years | No | 5 | 1 | ||||||
478 | Yes | Above | >3 Years | 4 | 2 | |||||
3424 | 4 | Yes | >1 Years | 4 | 2 | |||||
1724 | Above | >3 Years | No | 3 | 2 | |||||
1286 | 4 | No | >1 Years | 2 | 2 | |||||
Experiment 1. Negative regularities | ||||||||||
〈a, Strabd0,a〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 | DM | k | Expert’s rating | |
69 | 3 | >3 Years | 7 | 2 | ||||||
353 | No | >3 Years | 7 | 2 | ||||||
269 | >1 Years | No | Yes | 6 | 3 | |||||
546 | No | >5 Years | No | 6 | 3 | |||||
308 | 5 | No | 3 | 3 | ||||||
399 | 4 | Norm | >5 Years | Yes | 3 | 2 | ||||
267 | 6 | Above | >1 Years | No | 2 | 3 | ||||
480 | 6 | Yes | From 1 to 2 years | No | 2 | 3 | ||||
672 | 4 | No | Yes | Norm | >2 Years | 2 | 2 | |||
Experiment 1. Negative hypotheses that are not regularities | ||||||||||
〈Strab,ad0〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 | DM | k | Expert’s rating | |
271 | >1 Years | 69 | 3 | |||||||
〈Strab,a〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 | DM | k | Expert’s rating | |
62 | No | 70 | 2 | |||||||
224 | No | 51 | 1 | |||||||
4097 | Norm | 47 | 1 | |||||||
4100 | Yes | 45 | 1 | |||||||
364 | <5 Years | 33 | 1 | |||||||
4108 | Yes | 33 | 1 | |||||||
1249 | Above | 25 | 1 | |||||||
1879 | Below | 8 | 1 | |||||||
Experiment 2. Positive regularities | ||||||||||
〈a, Strab,ab〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 | DM | k | Expert’s rating | Present in experiment 1 |
3315 | Yes | <1 Years | No | 4 | 3 | Yes | ||||
〈a, Stra,ab〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 | DM | k | Expert’s rating | Present in experiment 1 |
884 | No | Norm | 20 | 2 | No | |||||
1103 | No | Norm | 15 | 2 | No | |||||
2520 | No | Yes | 13 | 2 | No | |||||
2101 | No | >2 Years | 12 | 2 | No | |||||
1835 | Norm | <1 Years | 11 | 3 | No | |||||
4010 | No | Yes | 9 | 2 | No | |||||
1857 | 6 | Norm | <5 Years | 8 | 3 | No | ||||
2728 | Norm | <2 Years | Yes | 8 | 2 | No | ||||
1830 | Yes | <1 Years | 6 | 3 | No | |||||
〈a, Strab,abd0〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 | DM | k | Expert’s rating | Present in experiment 1 |
3145 | No | No | From 1 to 2 years | 6 | 3 | Yes | ||||
3315 | Yes | <1 Years | No | 4 | 3 | Yes | ||||
〈n, Strab,abd0〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 | DM | k | Expert’s rating | Present in experiment 1 |
23055 | 5 | No | Below | >1 Years | 3 | 2 | No | |||
〈n, Stra,abd0〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 | DM | k | Expert’s rating | Present in experiment 1 |
37 | No | >1 Years | 22 | 2 | No | |||||
75 | 5 | 14 | 1 | No | ||||||
121 | No | >1 Years | 14 | 1 | No | |||||
494 | Yes | 14 | 2 | No | ||||||
2520 | No | Years | 13 | 2 | No | |||||
1855 | 6 | <5 Years | 12 | 3 | No | |||||
486 | Above | 11 | 1 | No | ||||||
998 | Yes | No | 7 | 1 | No | |||||
601 | 3 | Yes | Norm | 5 | 1 | No | ||||
Experiment 2. Negative regularities | ||||||||||
〈a, Strabd0,a〉 | ||||||||||
# | Age (decade) | Alcohol addiction | Tobacco smoking | BMI | Duration | Pain > 6 | DM | k | Expert’s rating | Present in experiment 1 |
473 | Norm | No | 7 | 3 | No | |||||
474 | Yes | No | 8 | 3 | No | |||||
479 | Yes | <5 Years | No | 6 | 2 | No |
Appendix 3. List of Correct Predictions and Errors by Strategies for the Last Expansion
For experiment 1. | ||||||||
10 | а | b | с | |||||
+ | - | + | - | + | - | + | - | |
Strab,ab | 8 | 3 | 3 | 2 | 1 | 1 | 2 | |
Stra,abd0 | 11 | 9 | ||||||
Strab,bd0 | 10 | 4 | 1 | 5 | ||||
Stra,ab | 9 | 5 | 2 | 4 | ||||
Strad0,ad0 | 9 | 5 | 3 | 3 | ||||
Strabd0,ad0 | 9 | 8 | 3 | |||||
Strad0,a | 9 | 8 | 3 | |||||
Strabd0,a | 9 | 11 | ||||||
Strab,ad0 | 3 | 5 | 1 | 7 | 4 | |||
Strad0,ab | 2 | 4 | 1 | 1 | 7 | 5 | ||
Strab,a | 5 | 1 | 10 | 4 | ||||
Strabd0,ab | 4 | 2 | 9 | 5 | ||||
Strad0,abd0 | 3 | 8 | 9 | |||||
Stra,ad0 | 3 | 8 | 9 | |||||
Strabd0,abd0 | 11 | 9 | ||||||
Stra,a | 11 | 9 | ||||||
For experiment 2. | ||||||||
10 | а | b | с | |||||
+ | - | + | - | + | - | + | - | |
Stra,abd0 | 7 | 3 | ||||||
Strab,abd0 | 5 | 2 | 3 | |||||
Strab,ab | 4 | 1 | 1 | 1 | 1 | 2 | ||
Stra,ab | 5 | 2 | 2 | 1 | ||||
Strab,ad0 | 2 | 2 | 3 | 2 | 1 | |||
Strab,a | 3 | 2 | 5 | |||||
Strad0,a, Strabd0,a | 3 | 7 | ||||||
Stra,ad0 | 2 | 1 | 5 | 2 | ||||
Strad0,ad0, Strabd0,ad0 | 2 | 5 | 2 | 1 | ||||
Strad0,ab, Strabd0,ab | 1 | 2 | 5 | 2 | ||||
Strad0,abd0, Strabd0,abd0 | 7 | 3 | ||||||
Stra,a | 7 | 3 |
About this article
Cite this article
Shesternikova, O.P., Finn, V.K., Lesko, K.A. et al. Intelligent System for Predicting the Feasibility of Using Computed Tomography. Sci. Tech. Inf. Proc. 50, 464–474 (2023). https://doi.org/10.3103/S0147688223050131
Received:
Published:
Issue Date:
DOI: https://doi.org/10.3103/S0147688223050131