20.500.12556/RUL-170965 Model za prepoznavanje stopnje prizadetosti s pomočjo testa tapkanja s prsti pri bolnikih s Parkinsonovo boleznijo A model for identifying the degree of impairment using the finger tapping test in patients with Parkinson's disease Parkinsonova bolezen je ena najpogosteje obravnavanih nevroloških bolezni na področju strojnega učenja. Pravočasna obravnava pacienta in ocena stopnje prizadetosti pomembno prispevata k ustrezni terapiji, kar pa je zaradi dolgih čakalnih vrst pogosto oteženo. Razvili smo klasifikacijski model, ki temelji na opisanih merilih lestvice Movement Disorder Society-Unified Parkinson’s Disease Rating Scale (MDS-UPDRS). Glede na opise v lestvici MDS-UPDRS generiramo umetne signale tapkanja s prsti, katere uporabimo za učenje globokih modelov strojnega učenja; razvite modele nato apliciramo na realne podatke, pridobljene iz video posnetkov. Ocene opredeljujejo tri ključne anomalije: upad amplitude, upad hitrosti in prekinitve signala, ki smo jih matematično opisali in ločeno modelirali s pomočjo treh preprostih avtokodirnikov. Latentni prostori avtokodirnikov so služili kot vhodni podatki za metodo k najbližjih sosedov (kNN), s katero smo določali stopnjo prizadetosti glede na posamezno anomalijo. Za nove primere ne uporabljamo učenja, temveč primerjamo signal tapkanja s prsti z naborom umetno generiranih signalov in s pomočjo metode najbližjega soseda (kNN) poiščemo, kateremu od vnaprej definiranih signalov je opazovani signal najbližji, kar omogoča oceno skladno z merili MDS-UPDRS. Največjo točnost je dosegel model za napoved upada hitrosti, čeprav je numerična opredelitev te anomalije otežena zaradi tekstovnega opisa v lestvici MDS-UPDRS. Zbranih je bilo 183 video posnetkov tapkanja s prsti, iz katerih smo s pomočjo orodja MediaPipe Hand pridobili signale tapkanja. Ti signali so omogočili generiranje umetnega nabora podatkov in testiranje klasifikacijskega pristopa. Avtokodirniki in metode kNN prispevajo k razumevanju povezave med matematično opredeljenimi anomalijami v signalih gibanja prstov in ocenami MDS-UPDRS. Parkinson’s disease is one of the most commonly studied neurological disorders in the field of machine learning. Timely patient care and assessment of the severity level significantly contribute to appropriate therapy, which is often hindered by long waiting times. We developed a classification model based on the criteria described in the Movement Disorder Society-Unified Parkinson’s Disease Rating Scale (MDS-UPDRS). According to the descriptions in the MDS-UPDRS scale, we generate synthetic finger-tapping signals, which we use to train deep machine learning models; the developed models are then applied to real data obtained from video recordings. The scores define three key anomalies: decrease in amplitude, decrease in speed, and signal interruptions, which we mathematically described and separately modeled using three simple autoencoders. The latent spaces of the autoencoders served as input data for the k-nearest neighbors (kNN) method, which we used to determine the severity level based on each anomaly. For new cases, we do not perform additional training; instead, we compare the observed finger-tapping signal with a set of synthetically generated signals and use the kNN method to find which predefined signal the observed one is closest to, enabling assessment consistent with the MDS-UPDRS criteria. The highest accuracy was achieved by the model predicting the decrease in speed, although the numerical definition of this anomaly is challenging due to the textual description in the MDS-UPDRS scale. A total of 183 finger-tapping videos were collected, from which tapping signals were extracted using the MediaPipe Hand tool. These signals enabled the generation of a synthetic dataset and the testing of the classification approach. The autoencoders and kNN methods contribute to understanding the relationship between mathematically defined anomalies in finger movement signals and the MDS-UPDRS scores. Parkinsonova bolezen test tapkanja s prsti avtokodirniki metoda najbližjega soseda klasifikacija Parkinson’s disease finger-tapping test autoencoders k-nearest neighbors classification true false false Slovenski jezik Angleški jezik Magistrsko delo/naloga 2025-07-24 07:20:00 2025-07-24 07:20:09 2025-08-04 15:39:44 0000-00-00 00:00:00 2025 0 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 63070 244778499 Krzan_Neza_-_Model_za_prepoznavanje_stopnje_prizadetosti_s_pomocjo_testa_tapkanja_s_prsti_pri_bo.pdf Krzan_Neza_-_Model_za_prepoznavanje_stopnje_prizadetosti_s_pomocjo_testa_tapkanja_s_prsti_pri_bo.pdf 1 735688BD4D4D506B29D6BECEF603ACF7 9c43356a9ff26ecca2d618481694c59b1c0cbf5bc45ea78920bd5326bac64195 d9ae3c27-684d-11f0-9328-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=213268 Fakulteta za elektrotehniko 0 0 0