{"id":265,"date":"2023-10-04T19:05:43","date_gmt":"2023-10-04T17:05:43","guid":{"rendered":"https:\/\/slf.dgstage.se\/sfmr\/imago-medica\/?p=265"},"modified":"2023-10-19T19:03:22","modified_gmt":"2023-10-19T17:03:22","slug":"sfmrs-foretagsstipendier-2023-neuro","status":"publish","type":"post","link":"https:\/\/slf.se\/sfmr\/imago-medica\/sfmrs-foretagsstipendier-2023-neuro\/","title":{"rendered":"SFMR:s f\u00f6retagsstipendier 2023: Neuro"},"content":{"rendered":"\n

Stipendium fr\u00e5n Siemens Healthineers tilldelas Maria Correia de Verdier.<\/h3>\n\n\n\n

Titel<\/strong>: F\u00f6r utveckling av nya metoder inom bildgivande diagnostik inom det kardiovaskul\u00e4ra eller neurovaskul\u00e4ra omr\u00e5det. <\/p>\n\n\n\n

Hj\u00e4rnmetastaser \u00e4r den vanligaste tum\u00f6ren i centrala nervsystemet. Framsteg inom magnetkameraunders\u00f6kning (MR) har resulterat i b\u00e4ttre och mer h\u00f6guppl\u00f6sta sekvenser, vilket leder till l\u00e4ttare uppt\u00e4ckt av sm\u00e5 metastaser. Tillsammans med effektivare systemisk och lokal onkologisk behandling lever patienter med hj\u00e4rnmetastaser l\u00e4ngre. D\u00e4rmed l\u00f6per de \u00f6kad risk att under sin livstid utveckla nya och sm\u00e5 hj\u00e4rnmetastaser. Att uppt\u00e4cka sm\u00e5 metastaser och karakterisera hur de f\u00f6r\u00e4ndras \u00e4r en tidskr\u00e4vande uppgift f\u00f6r radiologer som kr\u00e4ver h\u00f6g sensitivitet och specificitet f\u00f6r att s\u00e4kerhetsst\u00e4lla korrekt diagnos. I och med utvecklingen av artificiell intelligens (AI) har dess anv\u00e4ndbarhet f\u00f6r analys av radiologiska bilder \u00f6kat. En till\u00e4mpning av AI f\u00f6r automatisk uppt\u00e4ckt, kvantifiering och longitudinell bed\u00f6mning av hj\u00e4rnmetastaser har potential att f\u00f6rb\u00e4ttra och effektivisera bed\u00f6mningen av MR-bilder och m\u00f6jligg\u00f6ra mer korrekt diagnos och effektivare behandling.<\/p>\n\n\n\n

Specifika m\u00e5l:<\/p>\n\n\n\n

– Utveckla AI-tekniker f\u00f6r detektion och kvantifiering av hj\u00e4rnmetastaser vid en tidpunkt samt f\u00f6r longitudinell bed\u00f6mning av hj\u00e4rnmetastaser.<\/p>\n\n\n\n

– Kliniskt validera dessa metoder genom att testa om de kan f\u00f6rb\u00e4ttra radiologers noggrannhet och effektivitet f\u00f6r att uppt\u00e4cka och longitudinellt bed\u00f6ma hj\u00e4rnmetastaser.<\/p>\n\n\n\n

Slutligt m\u00e5l:<\/p>\n\n\n\n

– Minska m\u00e4nsklig bias, \u00f6ka arbetsfl\u00f6deseffektiviteten och f\u00f6rb\u00e4ttra patientens kliniska utfall genom att implementera AI i radiologisk vardag.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Stipendium fr\u00e5n Canon tilldelas Woosung Seo.<\/h3>\n\n\n\n

Titel<\/strong>: Estimating ventricular CSF volume pre and post shunt surgery using automatic and manual methods in iNPH patients

Background<\/strong>: Idiopathic normal pressure hydrocephalus (iNPH) is a condition that is frequently overlooked and insufficiently treated. Common symptoms include walking difficulties, cognitive decline and incontinence. INPH can be treated by inserting a shunt, preferably at an early stage to prevent permanent damage.

The aim of the study is to evaluate an AI program\u2019s ability to automatically quantify the CSF volume of the lateral ventricles in CT brain scans, compared to manual segmentations performed in an online dicom interface. Secondly, a comparison will be made between preoperative CT brain image to the routine follow-up after 6 weeks.

Methods: <\/strong>49 iNPH patients with documented postsurgical clinical improvement on the iNPH scale were included. CT brain were performed pre and post shunt surgery and analysed by radiologists. A deep-learning-based, AI software developed by collaborators in Gothenburg was used for the automated segmentation of ventricular CSF (VCSF) in a total of 98 CT brain images. This was compared to manually derived VCSF labels from iNPH patients at Uppsala university hospital.

Preliminary results: <\/strong>Both manual and automatic volumetry of lateral ventricles were successfully performed. Pearson correlation coefficient showed R=0.94. The median volume reduction post shunt surgery showed 12.1% decrease in ventricular volume in automatic measurement while in manual 12.2%.

Conclusion: <\/strong>Our results demonstrate strong correlation between automatic and manual CSF volumetric measurements on CT brain. Both excellent interrater reliability and positive relationship to clinical improvement. Hence, there is an evident potential of CT-derived volumetric measures to aid in the clinical diagnosis of iNPH.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n\n\n

Stipendium fr\u00e5n Canon tilldelas Dimitrios Toumpanakis.<\/h3>\n\n\n\n

Titel<\/strong>: Development of algorithms for the automatic evaluation of iNPH using Deep Learning and quantitative image analysis methods on radiological imaging<\/p>\n\n\n\n

Idiopathic normal pressure hydrocephalus (iNPH) is a form of communicating hydrocephalus accompanied often by the clinical triad of gait disturbance, cognitive impairment and urinary incontinence. Its incidence in various studies ranges from 1.8 to 7.3\/100,000, while it is believed to be under-diagnosed and under-treated. Early diagnosis is of paramount importance since iNPH is a potentially reversible cause of dementia.<\/p>\n\n\n\n

According to the international guidelines for iNPH, the diagnosis requires concrete clinical and imaging criteria. However, many of the individual imaging parameters described in various studies lack adequate sensitivity, specificity or interobserver agreement. In addition, the already high volume of dementia imaging examinations, a significant part of which concerns iNPH, are only going to increase as the population gets older.<\/p>\n\n\n\n

By building upon the power of convolutional neural networks and deep learning, our team will synthesize different approaches into one clinically applicable automatic algorithm that can function as a decision support tool for more confident, accurate and faster iNPH diagnosis based on radiology. Specifically, our project aims first to develop a novel multi-modular algorithm for the automatic evaluation of currently established imaging biomarkers for iNPH by combining Deep Learning and quantitative image analysis methods. Furthermore, we will develop a convolutional neural network that predicts the radiological probability for iNPH and we will implement explainable AI methods on the above neural network to allow interpretability and potentially discover\/extract novel imaging biomarkers for iNPH.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Stipendium fr\u00e5n Bracco tilldelas Sara Shams.<\/h3>\n\n\n\n

Titel<\/strong>: Hur p\u00e5verkar sjukdom av sm\u00e5 k\u00e4rl i hj\u00e4rnan en frisk population. <\/p>\n\n\n\n

Cerebral sm\u00e5k\u00e4rlssjuka \u00e4r sjukdom av mikroskopiska k\u00e4rl i hj\u00e4rnan. Sm\u00e5k\u00e4rlssjuka \u00e4r vanligt inom \u00e5ldrande, kognitiv svikt, stroke, diabetes och njursvikt visat genom v\u00e5ra studier. Sm\u00e5k\u00e4rlssjuka \u00e4r samtidigt en av de vanligaste anledningarna bakom hemorragisk stroke och tros kunna leda till demenssjukdom. \u00c4nnu r\u00e5der dock stor ovisshet kring cerebral sm\u00e5k\u00e4rlssjuka och dess riktiga inneb\u00f6rd och implikationer.<\/p>\n\n\n\n

I v\u00e5ra studier unders\u00f6ker vi sm\u00e5k\u00e4rlssjuka i tv\u00e4rsnitt samt longitudinellt och f\u00f6ljer friska populationer, patienter med diabetes och njursvikt samt patienter med demenssjukdom \u00f6ver tid p\u00e5 magnetkamera i relation till kliniska associationer, CSF-prover och andra labbv\u00e4rden. Data analyseras f\u00f6r samband i multivariabla regressionsanalyser.<\/p>\n\n\n\n

Hittills har vi visat att sm\u00e5k\u00e4rlssjuka \u00e4r associerat med h\u00f6gre amyloidinlagring i hj\u00e4rnan i s\u00e4rskilda regioner av hj\u00e4rnan, att sjukdomen \u00e4r associerad med \u00f6kad morbiditet och mortalitet samt nedsatt kognition. Patienter med sm\u00e5k\u00e4rlssjukdom och antikoagulans har 4 g\u00e5nger \u00f6kad risk f\u00f6r hemorragisk stroke.  <\/p>\n\n\n\n

Vi hoppas att v\u00e5ra vidare studier kan bringa fortsatt klarhet i vad sm\u00e5k\u00e4rlsjuka faktiskt \u00e4r f\u00f6r sjukdom och vad de enskilda mark\u00f6rerna i hj\u00e4rnan p\u00e5 MR faktiskt betyder och vilka associationer som existerar. Denna kunskap hoppas vi bringar mer klarhet i vardaglig granskning av MR-bilder av hj\u00e4rnan och bidrar till mer f\u00f6rst\u00e5else kring hur enskilda MR-mark\u00f6rer i hj\u00e4rnan kommer att p\u00e5verka patienterna vi ser p\u00e5 daglig basis.<\/p>\n\n\n\n

\"\"<\/figure>\n","protected":false},"excerpt":{"rendered":"

Stipendium fr\u00e5n Siemens Healthineers tilldelas Maria Correia de Verdier. Titel: F\u00f6r utveckling av nya metoder inom bildgivande diagnostik inom det kardiovaskul\u00e4ra eller neurovaskul\u00e4ra omr\u00e5det. Hj\u00e4rnmetastaser \u00e4r den vanligaste tum\u00f6ren i centrala nervsystemet. Framsteg inom magnetkameraunders\u00f6kning (MR) har resulterat i b\u00e4ttre och mer h\u00f6guppl\u00f6sta sekvenser, vilket leder till l\u00e4ttare uppt\u00e4ckt av sm\u00e5 metastaser. Tillsammans med effektivare … L\u00e4s hela<\/a><\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[15],"class_list":["post-265","post","type-post","status-publish","format-standard","hentry","category-neuroradiologi"],"acf":[],"_links":{"self":[{"href":"https:\/\/slf.se\/sfmr\/imago-medica\/wp-json\/wp\/v2\/posts\/265","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/slf.se\/sfmr\/imago-medica\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/slf.se\/sfmr\/imago-medica\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/slf.se\/sfmr\/imago-medica\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/slf.se\/sfmr\/imago-medica\/wp-json\/wp\/v2\/comments?post=265"}],"version-history":[{"count":2,"href":"https:\/\/slf.se\/sfmr\/imago-medica\/wp-json\/wp\/v2\/posts\/265\/revisions"}],"predecessor-version":[{"id":317,"href":"https:\/\/slf.se\/sfmr\/imago-medica\/wp-json\/wp\/v2\/posts\/265\/revisions\/317"}],"wp:attachment":[{"href":"https:\/\/slf.se\/sfmr\/imago-medica\/wp-json\/wp\/v2\/media?parent=265"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/slf.se\/sfmr\/imago-medica\/wp-json\/wp\/v2\/categories?post=265"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}