K.O. Tutsenko1, A.N. Narkevich2, V.G. Abramov3

1,2 FSBEI НЕ Prof. V.F. Voino-Yasenetsky KrasSMU MOH Russia (Krasnoyarsk, Russia)

2 South Ural State Medical University (Chelyabinsk, Russi)

3 Federal Siberian Research Clinical Center, FMBA of Russia (Krasnoyarsk, Russia)

Parkinson's disease is characterized by the destruction of neurons in the dopaminergic system of the brain. This disease is one of the most common neurodegenerative diseases of the elderly and is characterized by many symptoms. This pathology has a long asymptomatic period, during which a significant number of neurons die and dopamine deficiency develops. Preclinical diagnostics is the most important direction when working with patients with Parkinson's disease.

Parkinson's disease has symptoms similar to essential tremor. Errors in the differential diagnosis of these pathologies are observed quite often, which is associated with the absence of specific diagnostic signs, a short observation period, and a lack of clinical data. Such a diagnostic problem can be solved through the study of the dopaminergic system of the brain using positron emission tomography with the 18F-DOPA radiopharmaceutical. If a decrease in the activity of the radiopharmaceutical is detected, the patient is diagnosed with Parkinson's disease; with essential tremor, neuronal destruction and dopamine deficiency are not observed.

The purpose of the study: to determine the ability of various clustering methods to differentiate patients with PD from other study groups.

We analyzed 16 absolute and 24 relative indicators of radiopharmaceutical activity in the brain on both sides. The k-means method implemented in the IBM SPSS Statistics 26 program was used for data clustering. The elastic map method implemented in the
VidaExpert program was used for data clustering and visualization.

The results of data clustering using the k-means method showed that this method in its classical form cannot be used to solve the data clustering problem described in this article. The method of elastic maps was used to visualize multidimensional data. To assess the quality of clustering using the method of elastic maps, the internal coordinates of points on the presented plane were extracted and the k-means method was applied. The method of elastic maps and the method of k-means for the internal coordinates of the elastic map showed a high quality of clustering of the studied groups, which indicates the distinguishability of the groups in terms of radiopharmaceutical activity.

In the course of the work, the inapplicability of the classical k-means method for clustering 18F-DOPA activity data was proved, which may be due to ignoring the dependencies between the indicators. Using the method of elastic maps, the distinguishability of the studied groups was checked according to the activity of the radiopharmaceutical. Thus, indicators of activity in different areas of the brain make it possible to differentiate between patients with Parkinson's disease and healthy people, as well as patients with Parkinson's disease and essential tremor. And clustering by the method of elastic maps in combination with the k-means method gives high clustering accuracy.

Tutsenko K.O., Narkevich A.N., Abramov V.G. Possibilities of differential diagnosis of Parkinson's disease using clustering methods. Technologies of Living Systems. 2023. V. 20. № 2. Р. 34-41. DOI: https://doi.org/10.18127/j20700997-202302-04 (In Russian)

1. Baranov O.A., Martyn I.A. Sravnitelnyy analiz dvukh metodov klasternogo analiza. Informatsionnyye tekhnologii i sistemy: upravleniye. ekonomika. transport. pravo. 2022. № 1(41). S. 30–34. (in Russian).
2. Zinovyev A.Yu. Vizualizatsiya proizvolnykh dannykh metodom uprugikh kart. Radіoyelektronіka. іnformatika. upravlіnnya. 2000. № 1(3). S. 76–85. (in Russian).
3. Katunina E.A. Blokhin V.E., Nodel M.R., Moskalets E.R., Ugryumov E.R., Ugryumov M.V. Otsenka rannikh kliniko-biokhimicheskikh markerov bolezni Parkinsona s pomoshchyu pozitronno-emissionnoy tomografii. Byulleten Natsionalnogo obshchestva po izucheniyu bolezni Parkinsona i rasstroystv dvizheniy. 2022. T. 2. S. 82–86. (in Russian).
4. Magzhanov R.V., Ibatullin R.A., Davletova A.I. Differentsialnaya diagnostika essentsialnogo i parkinsonicheskogo vidov tremora. Vestnik Rossiyskoy Voyenno-meditsinskoy akademii. 2018. T. S3. S. 83–84. (in Russian).
5. Mosaleva E.I., Zhumzhanov I.M., Alekseyenko P.V., Ismailova S.B., Prokopenko S.V. Kognitivnyye fluktuatsii na fone terapii u patsiyentov s boleznyu Parkinsona. Sibirskoye meditsinskoye obozreniye. 2021. № 1. S. 63–67. (in Russian).
6. Plokhikh A.B., Masaleva I.O., Lyubortseva Yu.O. Nefarmakologicheskiye metody lecheniya trevozhnykh narusheniy pri bolezni Parkinsona (obzor literatury). Mezhdunarodnyy nauchno-issledovatelskiy zhurnal. 2022. № 6-2(120). S. 103–106. (in Russian).
7. Poverennova I.E., Kalinin V.A., Shpileva S.A. Diagnostika nemotornykh narusheniy pri bolezni Parkinsona. Sibirskoye meditsinskoye obozreniye. 2018. № 3. S. 93–97. (in Russian).
8. Selikhova M.V., Katunina E.A., Voun A. Pozitronnaya emissionnaya i odnofotonnaya emissionnaya kompyuternaya tomografiya v otsenke sostoyaniya monoaminergicheskikh sistem mozga pri ekstrapiramidnykh rasstroystvakh. Annaly klinicheskoy i eksperimentalnoy nevrologii. 2019. T. 13(2). S. 69–78. (in Russian).
9. Tkach V.V., Frantsishko O.B., Tkach A.V. Neyrokhirurgicheskoye lecheniye bolezni Parkinsona. Modern Science. 2021. V. 4(4). P. 206–211. (in Russian).
10. Trufanov E.A. Differentsialnyy diagnoz bolezni Parkinsona i essentsialnogo tremora. Meditsinskiy vestnik yuga Rossii. 2013. № 1. S. 73–79. (in Russian).
11. Tutsenko K.O., Narkevich A.N., Abramov V.G. Analiz aktivnosti radiofarmpreparata 18F-DOPA dlya differentsialnoy diagnostiki ekstrapiramidnykh patologiy. Vestnik novykh meditsinskikh tekhnologiy. 2022. № 3. S. 105–108. (in Russian).
12. Ugryumov M.V., Kozina E.A., Khaindrava V.G., Kudrin V.S., Kucheryanu V.G., Klodt P.D., Rayevskiy K.S. Modelirovaniye parkinsonizma u myshey pri pomoshchi MFTP: ot ranney dosimptomnoy do pozdney simptomnoy stadii. Tekhnologii zhivykh sistem. 2011. T. 8(8). S. 3–13. (in Russian).
13. Iacono D., Geraci-Erck M., Rabin M.L., Adler C.H., Serrano G., Beach T.G., Kurlan R. Parkinson disease and incidental Lewy body disease: just a question of time? Neurology. 2015. V. 85. P. 1670–1679.
14. Löhle M., Wolz M., Beuthien-Baumann B., Oehme L., Hoff J., Kotzerke J., Reichmann H., Storch A. Olfactory dysfunction correlates with putaminal dopamine turnover in early de novo Parkinson’s disease. Journal of Neural Transmission. 2020. V. 127(1). P. 9–16.
15. Savica R., Grossardt B.R., Bower J.H., Ahlskog J.E., Rocca W.A. Incidence and pathology of synucleinopathies and tauopathies related to parkinsonism. Journal of the American Medical Association. 2013. V. 70. P. 859–866.