{"id":29,"date":"2024-07-04T14:11:42","date_gmt":"2024-07-04T12:11:42","guid":{"rendered":"https:\/\/odas.ujep.cz\/?page_id=29"},"modified":"2025-11-06T11:10:02","modified_gmt":"2025-11-06T10:10:02","slug":"cinnost","status":"publish","type":"page","link":"https:\/\/odas.ujep.cz\/en\/cinnost\/","title":{"rendered":"Research Activities"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">N\u00e1\u0161 t\u00fdm se v\u011bnuje pokro\u010dil\u00e9mu v\u00fdzkumu a&nbsp;v\u00fdvoji <strong>algoritm\u016f <\/strong>pro zpracov\u00e1n\u00ed a&nbsp;anal\u00fdzu <strong>medic\u00ednsk\u00fdch dat<\/strong>, matematick\u00e9mu <strong>modelov\u00e1n\u00ed <\/strong>a&nbsp;simulaci komplexn\u00edch syst\u00e9m\u016f. Na t\u00e9to str\u00e1nce naleznete p\u0159ehled nab\u00edzen\u00fdch v\u00fdzkumn\u00fdch aktivit na\u0161eho odd\u011blen\u00ed.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"cinnost-medanal\">Po\u010d\u00edta\u010dov\u00e1 anal\u00fdza medic\u00ednsk\u00fdch dat<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Specializujeme se na v\u00fdvoj a&nbsp;implementaci algoritm\u016f pro zpracov\u00e1n\u00ed a&nbsp;anal\u00fdzu medic\u00ednsk\u00fdch dat, p\u0159edev\u0161\u00edm pro <strong>diagnostick\u00e9 \u00fa\u010dely<\/strong>. Zam\u011b\u0159ujeme se nap\u0159\u00edklad na <strong>detekci specifick\u00fdch ud\u00e1lost\u00ed<\/strong> v&nbsp;sign\u00e1lech, <strong>extrapolaci <\/strong>komplexn\u00edch informac\u00ed z&nbsp;omezen\u00fdch datov\u00fdch soubor\u016f, korelaci, segmentaci bun\u011bk ze sn\u00edmk\u016f po\u0159\u00edzen\u00fdch mikroskopy a&nbsp;tvorbu specializovan\u00fdch <strong>datab\u00e1z\u00ed <\/strong>pro dal\u0161\u00ed v\u00fdzkumn\u00e9 vyu\u017eit\u00ed. V&nbsp;oblasti fyziologick\u00fdch sign\u00e1l\u016f m\u00e1me zku\u0161enosti se zpracov\u00e1n\u00edm <strong>EKG, EEG, EMG<\/strong> a&nbsp;sign\u00e1l\u016f zobrazuj\u00edc\u00edch nitrolebn\u00ed (ICP) a&nbsp;arteri\u00e1ln\u00ed krevn\u00ed tlak (ABP). U&nbsp;obrazov\u00fdch dat nej\u010dast\u011bji pracujeme se sn\u00edmky bun\u011bk a&nbsp;tk\u00e1n\u00ed. Pro zpracov\u00e1n\u00ed vyu\u017e\u00edv\u00e1me \u0161irok\u00e9 spektrum n\u00e1stroj\u016f \u2013 od klasick\u00fdch programovac\u00edch jazyk\u016f (nap\u0159. Python) a\u017e po technologie Big Data (nap\u0159. Apache Hadoop, Cassandra). D\u00e1le se zam\u011b\u0159ujeme na implementaci metod <strong>strojov\u00e9ho u\u010den\u00ed<\/strong>, v\u010detn\u011b vyu\u017eit\u00ed specializovan\u00e9ho hardwaru a&nbsp;masivn\u011b paraleln\u00edch p\u0159\u00edstup\u016f.&nbsp;<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong><a href=\"https:\/\/odas.ujep.cz\/projekty\/\">\u0158e\u0161en\u00e9 projekty<\/a><\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>OP JAK: <\/strong><em><a href=\"https:\/\/odas.ujep.cz\/digitech\/\">DIGITECH<\/a><\/em>, 2025\u2013dosud.<\/li>\n\n\n\n<li><strong>FN Bulovka, FN Kr\u00e1lovsk\u00e9 Vinohrady, FBMI \u010cVUT, KI UJEP<\/strong>: <em><a href=\"https:\/\/odas.ujep.cz\/2025\/03\/11\/odas-se-zapojil-do-strategicke-spoluprace-v-oblasti-medicinskych-dat\/\">PhysioConnect.cz<\/a><\/em>, 2023\u2013dosud.<\/li>\n\n\n\n<li><strong>IGA Krajsk\u00e1 zdravotn\u00ed, a.s.<\/strong>: <em>Um\u011bl\u00e1 inteligence v&nbsp;segmentaci a&nbsp;hodnocen\u00ed radiologick\u00e9ho obrazu<\/em>, 2022\u20132024.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"cinnosti-predikce\">Prediktivn\u00ed modelov\u00e1n\u00ed<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Zam\u011b\u0159ujeme se na aplikaci metod matematick\u00e9ho modelov\u00e1n\u00ed pro anal\u00fdzu a&nbsp;predikci <strong>\u010dasov\u00fdch \u0159ad<\/strong>. Pracujeme s&nbsp;daty z&nbsp;firemn\u00edch proces\u016f a&nbsp;podnikatelsk\u00e9ho prost\u0159ed\u00ed, ale i&nbsp;s&nbsp;\u010dasov\u00fdmi \u0159adami vznikl\u00fdmi v&nbsp;r\u00e1mci dal\u0161\u00edch v\u00fdzkumn\u00fdch aktivit. Vyu\u017e\u00edv\u00e1me jak klasick\u00e9 statistick\u00e9 p\u0159\u00edstupy, tak i&nbsp;modern\u00ed metody strojov\u00e9ho u\u010den\u00ed \u2013 nap\u0159\u00edklad <strong>neuronov\u00e9 s\u00edt\u011b<\/strong> \u2013 p\u0159\u00edpadn\u011b jejich kombinace.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Modelov\u00e1n\u00ed \u010dasov\u00fdch \u0159ad<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Modelov\u00e1n\u00ed \u010dasov\u00fdch \u0159ad prov\u00e1d\u00edme pomoc\u00ed klasick\u00fdch statistick\u00fdch metod, nap\u0159\u00edklad autoregresn\u00edch model\u016f (ARMA, ARIMA), jejich sez\u00f3nn\u00edch roz\u0161\u00ed\u0159en\u00ed (SARIMA) \u010di variant s&nbsp;extern\u00edmi regresory ([S]ARIMAX). Vhodn\u00fd typ modelu a&nbsp;jeho parametry vol\u00edme na z\u00e1klad\u011b d\u016fkladn\u00e9 anal\u00fdzy p\u0159\u00edslu\u0161n\u00e9 \u010dasov\u00e9 \u0159ady, v\u010detn\u011b posouzen\u00ed jej\u00edch statistick\u00fdch vlastnost\u00ed.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Dal\u0161\u00ed skupinu model\u016f tvo\u0159\u00ed metody strojov\u00e9ho u\u010den\u00ed. Vyu\u017e\u00edv\u00e1me bu\u010f hotov\u00e9 aplika\u010dn\u00ed bal\u00edky, nap\u0159\u00edklad Prophet, zalo\u017een\u00fd na aditivn\u00edm regresn\u00edm modelu a\u00a0bayesovsk\u00e9m p\u0159\u00edstupu, nebo navrhujeme a\u00a0implementujeme vlastn\u00ed neuronov\u00e9 s\u00edt\u011b pro predikci konkr\u00e9tn\u00edch \u010dasov\u00fdch \u0159ad. Nej\u010dast\u011bji pracujeme s\u00a0rekurentn\u00edmi s\u00edt\u011bmi typu <strong>LSTM<\/strong> nebo <strong>GRU<\/strong>, implementovan\u00fdmi v\u00a0b\u011b\u017en\u011b dostupn\u00fdch knihovn\u00e1ch, jako jsou <strong>TensorFlow <\/strong>\u010di <strong>PyTorch<\/strong>. \u00dasp\u011b\u0161n\u00e9 u\u010den\u00ed neuronov\u00fdch s\u00edt\u00ed v\u0161ak vy\u017eaduje dostate\u010dn\u011b rozs\u00e1hl\u00e1 a\u00a0kvalitn\u00ed data. Zab\u00fdv\u00e1me se tak\u00e9 vysv\u011btlov\u00e1n\u00edm vlivu jednotliv\u00fdch regresor\u016f na v\u00fdvoj \u010dasov\u00fdch \u0159ad, nap\u0159\u00edklad pomoc\u00ed interpretovateln\u00fdch model\u016f nebo citlivostn\u00edch anal\u00fdz.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong><a href=\"https:\/\/odas.ujep.cz\/projekty\/\">\u0158e\u0161en\u00e9 projekty<\/a><\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>ADLER Czech, a.s.:&nbsp;<\/strong><em>Datov\u00e1 anal\u00fdza ve spole\u010dnosti ADLER Czech, a.s.,&nbsp;<\/em>2019\u20132020.<\/li>\n\n\n\n<li><strong>Severo\u010desk\u00e9 doly a.s.:&nbsp;<\/strong><em>Anal\u00fdza vyu\u017eit\u00ed metod strojov\u00e9ho u\u010den\u00ed pro monitoring a&nbsp;predikci v\u00fdvoje stavu spodn\u00ed vody v&nbsp;dob\u00fdvac\u00edm \u00fazem\u00ed,&nbsp;<\/em>2020\u20132021.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"cinnosti-simulace\">Po\u010d\u00edta\u010dov\u00e1 simulace komplexn\u00edch syst\u00e9m\u016f<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Specializujeme se na simulace komplexn\u00edch syst\u00e9m\u016f pomoc\u00ed \u010d\u00e1sticov\u00fdch i\u00a0spojit\u00fdch p\u0159\u00edstup\u016f. V\u00a0r\u00e1mci \u010d\u00e1sticov\u00e9ho modelov\u00e1n\u00ed vyu\u017e\u00edv\u00e1me zejm\u00e9na <strong>molekul\u00e1rn\u00ed dynamiku<\/strong> k\u00a0popisu fyzik\u00e1ln\u011b-chemick\u00fdch jev\u016f \u2013 nap\u0159\u00edklad p\u0159i simulac\u00edch kompozitn\u00edch materi\u00e1l\u016f na b\u00e1zi polymer\u016f nebo separa\u010dn\u00edch proces\u016f v\u00a0membr\u00e1n\u00e1ch. Zab\u00fdv\u00e1me se tak\u00e9 simulac\u00ed pohybu chodc\u016f. Pou\u017e\u00edv\u00e1me jak diskr\u00e9tn\u00ed modely zalo\u017een\u00e9 na celul\u00e1rn\u00edch automatech, tak spojit\u00e9 p\u0159\u00edstupy vyu\u017e\u00edvaj\u00edc\u00ed metody <strong>computational fluid dynamics<\/strong>. V&nbsp;r\u00e1mci t\u011bchto metod jsou chodci modelov\u00e1ni jako inteligentn\u00ed tekutina.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Po\u010d\u00edta\u010dov\u00e9 simulace kompozitn\u00edch materi\u00e1l\u016f na b\u00e1zi polymer\u016f<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Po\u010d\u00edta\u010dov\u00fd design nanomateri\u00e1l\u016f vy\u017eaduje zohledn\u011bn\u00ed \u0161irok\u00e9ho spektra parametr\u016f a&nbsp;jejich mo\u017en\u00fdch kombinac\u00ed. Simulace p\u0159edstavuj\u00ed \u00fa\u010dinn\u00fd n\u00e1stroj pro systematick\u00e9 zkoum\u00e1n\u00ed t\u011bchto parametr\u016f a&nbsp;jejich optimalizaci p\u0159ed experiment\u00e1ln\u00edm ov\u011b\u0159en\u00edm. Vzhledem k&nbsp;vysok\u00e9mu po\u010dtu stup\u0148\u016f volnosti vyu\u017e\u00edv\u00e1me zjednodu\u0161en\u00e9 (zhrubnut\u00e9) modely polymer\u016f a&nbsp;dal\u0161\u00edch slo\u017eek kompozitn\u00edch materi\u00e1l\u016f. Pro zv\u00fd\u0161en\u00ed v\u00fdpo\u010detn\u00ed efektivity zapojujeme tak\u00e9 <strong>GPU akceleraci<\/strong>. V\u011bnujeme se p\u0159edev\u0161\u00edm modelov\u00e1n\u00ed samoorganizace kopolymern\u00edch syst\u00e9m\u016f \u2013 v&nbsp;objemov\u00e9 f\u00e1zi, na povrchu pevn\u00fdch f\u00e1z\u00ed (nap\u0159\u00edklad nano\u010d\u00e1stic) nebo v&nbsp;nerovnov\u00e1\u017en\u00fdch podm\u00ednk\u00e1ch, nap\u0159\u00edklad za p\u0159\u00edtomnosti toku.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong><a href=\"https:\/\/odas.ujep.cz\/projekty\/\">\u0158e\u0161en\u00e9 projekty<\/a><\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>GA \u010cR:&nbsp;<\/strong><em>\u0158\u00edzen\u00e1 asociace polyelektrolytov\u00fdch kopolymer\u016f se surfaktanty ve vodn\u00fdch roztoc\u00edch,&nbsp;<\/em>2013\u20132016.<\/li>\n\n\n\n<li><strong>ERDF\/ESF<\/strong>:&nbsp;<em>UniQSurf \u2013<\/em> <em>Centre of Biointerfaces and Hybrid Functional Materials<\/em>, 2018\u20132022.<\/li>\n\n\n\n<li><strong>TA \u010cR \u2013<\/strong> <strong>Kappa:&nbsp;<\/strong><em>METAMORPH \u2013<\/em> <em>Advanced Hybrid Organic-Inorganic Nanofibers for CO\u2082 Capture and Photocatalysis,&nbsp;<\/em>2021\u20132023.<\/li>\n\n\n\n<li><strong>GA \u010cR:&nbsp;<\/strong><em>Separace racemick\u00fdch sm\u011bs\u00ed membr\u00e1nov\u00fdmi procesy,&nbsp;<\/em>2017\u20132019.<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n","protected":false},"excerpt":{"rendered":"<p>N\u00e1\u0161 t\u00fdm se v\u011bnuje pokro\u010dil\u00e9mu v\u00fdzkumu a&nbsp;v\u00fdvoji algoritm\u016f pro zpracov\u00e1n\u00ed a&nbsp;anal\u00fdzu medic\u00ednsk\u00fdch dat, matematick\u00e9mu modelov\u00e1n\u00ed a&nbsp;simulaci komplexn\u00edch syst\u00e9m\u016f. Na t\u00e9to str\u00e1nce naleznete p\u0159ehled nab\u00edzen\u00fdch v\u00fdzkumn\u00fdch aktivit na\u0161eho odd\u011blen\u00ed. Po\u010d\u00edta\u010dov\u00e1 anal\u00fdza medic\u00ednsk\u00fdch dat Specializujeme se na v\u00fdvoj a&nbsp;implementaci algoritm\u016f pro zpracov\u00e1n\u00ed a&nbsp;anal\u00fdzu medic\u00ednsk\u00fdch dat, p\u0159edev\u0161\u00edm pro diagnostick\u00e9 \u00fa\u010dely. Zam\u011b\u0159ujeme se nap\u0159\u00edklad na detekci specifick\u00fdch ud\u00e1lost\u00ed v&nbsp;sign\u00e1lech, [&hellip;]<\/p>\n","protected":false},"author":2,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"folder":[],"class_list":["post-29","page","type-page","status-publish","hentry"],"featured_image_src":null,"_links":{"self":[{"href":"https:\/\/odas.ujep.cz\/en\/wp-json\/wp\/v2\/pages\/29","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/odas.ujep.cz\/en\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/odas.ujep.cz\/en\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/odas.ujep.cz\/en\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/odas.ujep.cz\/en\/wp-json\/wp\/v2\/comments?post=29"}],"version-history":[{"count":26,"href":"https:\/\/odas.ujep.cz\/en\/wp-json\/wp\/v2\/pages\/29\/revisions"}],"predecessor-version":[{"id":755,"href":"https:\/\/odas.ujep.cz\/en\/wp-json\/wp\/v2\/pages\/29\/revisions\/755"}],"wp:attachment":[{"href":"https:\/\/odas.ujep.cz\/en\/wp-json\/wp\/v2\/media?parent=29"}],"wp:term":[{"taxonomy":"folder","embeddable":true,"href":"https:\/\/odas.ujep.cz\/en\/wp-json\/wp\/v2\/folder?post=29"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}