{"id":7249,"date":"2026-07-14T11:34:24","date_gmt":"2026-07-14T09:34:24","guid":{"rendered":"https:\/\/ai4healthcro.eu\/?post_type=ht_kb&p=7249"},"modified":"2026-07-14T11:41:17","modified_gmt":"2026-07-14T09:41:17","slug":"from-winning-the-challenge-to-building-the-data-behind-it","status":"publish","type":"ht_kb","link":"https:\/\/ai4healthcro.eu\/?ht_kb=from-winning-the-challenge-to-building-the-data-behind-it","title":{"rendered":"From winning the challenge to building the data behind it"},"content":{"rendered":"\t\t
\n\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t

Karlo Pintari\u0107 first met the AI4Health.Cro Innovation Competition as a member of the winning MediBoost team. Two years later, he helped shape the 2026 diabetes challenge from the inside, preparing the real-world health data that allowed teams to build AI solutions on a clinically meaningful foundation.<\/h2>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
<\/div>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t

In 2024, Karlo Pintari\u0107 was on the competitor\u2019s side of the AI4Health.Cro Innovation Competition. His team, MediBoost, won first place with an application designed to predict early hospital readmissions, taking the top award in a competition that brought together medical and computing expertise around a concrete healthcare problem.<\/p>

By 2026, Pintari\u0107 had moved to the other side of the challenge. As a physician and data scientist at the Croatian Institute of Public Health (HZJZ) he became one of the key people behind the competition\u2019s data pipeline. The 2026 edition focused on type 2 diabetes and medication adherence, with Insulogic winning for an AI platform designed to identify patients at risk of discontinuing therapy and support follow-up through an AI voice assistant.<\/p>

That transition, from winning participant to organiser and data lead, captures one of the strongest features of the AI4Health.Cro model. The competition did more than reward a promising prototype. It helped create a community of young clinicians, engineers and data specialists who returned to strengthen the next generation of challenges.<\/p>

Why the data pipeline mattered<\/strong><\/h2>

For Pintari\u0107, the defining lesson of the 2026 competition was direct – an AI challenge in healthcare depends on the quality of its data.<\/p>

The teams were asked to predict medication adherence in a cohort of patients with type 2 diabetes using real clinical data. To make that possible, Pintari\u0107 and the AI4Health.Cro data team worked on harmonising primary-care and hospitalisation data into the OMOP Common Data Model, an international standard widely used for observational health data.<\/p>

This was a technical step with scientific consequences. Data prepared in a common model can be reused, compared and transferred more easily across OMOP-compatible environments. It also gives teams a more reliable basis for building solutions that can later be tested beyond the original challenge dataset.<\/p>

Using OHDSI tools, Pintari\u0107 deployed a phenotyping algorithm, extracted patient cohorts and prepared features for the baseline model. The predictive target was calculated using Proportion of Days Covered, PDC, a standard pharmacoepidemiological measure used to estimate medication adherence from prescription patterns and real-world data.<\/p>

These choices shaped the scientific quality of the competition. Participants were not simply given a spreadsheet and asked to optimise a model. They worked with data prepared through recognised methods, structured around a clinically relevant question and aligned with international approaches to secondary use of health data.<\/p>

From competition format to research infrastructure<\/strong><\/h2>

The 2026 challenge showed how innovation competitions can become more than short, high-energy events. When health data are prepared carefully, a competition can introduce participants to the same questions that define real-world clinical AI, which is, how cohorts are constructed, how outcomes are measured, how data formats influence future transferability and how predictive targets reflect clinical behaviour.<\/p>

AI4health.Cro data team\u2019s work connected these elements into a usable pipeline. The OMOP model created a common structure. OHDSI tools supported cohort and feature extraction. PDC provided a defensible adherence target. Together, these components gave participants a foundation that reflected the scientific and regulatory direction of European health-data use.<\/p>

The approach also aligns with the broader vision of the European Health Data Space<\/strong>, where the secondary use of health data<\/strong> is expected to support research, innovation and evidence-based healthcare <\/strong>under clearer governance<\/strong> and technical standards<\/strong>.<\/p>

For AI4Health.Cro team, this is a best-practice lesson in challenge design. The value of the competition lies in the data environment as much as in the final ranking. A clinically grounded challenge teaches teams how healthcare AI is built when the input data, outcome definition and evaluation context are treated seriously.<\/p>

A physician working at the edge of data science<\/strong><\/h2>

Pintari\u0107\u2019s role reflects a professional profile that is becoming increasingly important in digital health. He graduated in medicine from the University of Zagreb School of Medicine in 2023 and is now a physician and data scientist at the Croatian Institute of Public Health. There, his work focuses on health informatics, biostatistics and the application of artificial intelligence and machine learning in medicine.<\/p>

His expertise includes transforming health data into the OMOP Common Data Model, developing automated ETL pipelines and applying advanced analytics in clinical research. He is also active as a lecturer in workshops on the smart use of health data.<\/p>

That combination of clinical training and data-science practice shaped his contribution to the 2026 competition. He understood the medical stakes of adherence in diabetes care and the technical requirements needed to turn fragmented health data into a structured challenge dataset.<\/p>

A full-circle testimonial<\/strong><\/h2>

Pintari\u0107\u2019s experience with AI4Health.Cro now spans the full evolution of the innovation challenge.<\/p>

He first joined as a competitor and winner with MediBoost in 2024. He then contributed to the second challenge, where his role included processing mammography images. In 2026, through his work at HZJZ, he was involved from the earliest stages, including brainstorming, data preparation and final execution.<\/p>

That continuity matters. It shows how AI4Health.Cro can retain talent within its ecosystem and turn former participants into contributors, mentors and technical leads.<\/p>

For Pintari\u0107, the 2026 challenge confirmed the importance of high-quality, clean and validated data.<\/strong> It also demonstrated the value of real-world data <\/strong>for AI development and clinical research<\/strong> when supported by standard formats and transparent methods<\/strong>.<\/p>

His trajectory offers a clear best-practice message<\/strong> for future health innovation programmes: competitions should be designed as learning systems. The strongest participants should not disappear after the awards ceremony. They can become the people who raise the quality of the next dataset, the next challenge and the next generation of teams.<\/p>

AI4Health.Cro\u2019s Innovation Competition<\/strong> provides that pathway. It gives young clinicians, engineers and data scientists an entry point into practical healthcare AI, then creates opportunities for them to return with deeper expertise. In Pintari\u0107\u2019s case, the path led from a winning prototype to the data infrastructure behind a national health AI challenge.<\/p>

\u00a0<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"

Karlo Pintari\u0107 first met the AI4Health.Cro Innovation Competition as a member of the winning MediBoost team. Two years later, he helped shape the 2026 diabetes challenge from the inside, preparing the real-world health data that allowed teams to build AI solutions on a clinically meaningful foundation. https:\/\/youtu.be\/I6n9ntCfy3M?si=nX9-lhATIJ5RQ5qs In 2024, Karlo Pintari\u0107 was on the competitor\u2019s …<\/p>\n

From winning the challenge to building the data behind it<\/span> Read More »<\/a><\/p>\n","protected":false},"author":3,"comment_status":"open","ping_status":"closed","template":"","format":"standard","meta":{"_seopress_robots_primary_cat":"","_seopress_titles_title":"","_seopress_titles_desc":"","_seopress_robots_index":"","_bbp_topic_count":0,"_bbp_reply_count":0,"_bbp_total_topic_count":0,"_bbp_total_reply_count":0,"_bbp_voice_count":0,"_bbp_anonymous_reply_count":0,"_bbp_topic_count_hidden":0,"_bbp_reply_count_hidden":0,"_bbp_forum_subforum_count":0,"site-sidebar-layout":"no-sidebar","site-content-layout":"page-builder","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"disabled","ast-breadcrumbs-content":"","ast-featured-img":"disabled","footer-sml-layout":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","footnotes":""},"ht-kb-category":[94,93,81,30],"ht-kb-tag":[],"class_list":["post-7249","ht_kb","type-ht_kb","status-publish","format-standard","hentry","ht_kb_category-success-stories","ht_kb_category-best-practices","ht_kb_category-ai4health-cro-inovacijsko-natjecanje-2026","ht_kb_category-events-and-news-2"],"_links":{"self":[{"href":"https:\/\/ai4healthcro.eu\/index.php?rest_route=\/wp\/v2\/ht-kb\/7249","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ai4healthcro.eu\/index.php?rest_route=\/wp\/v2\/ht-kb"}],"about":[{"href":"https:\/\/ai4healthcro.eu\/index.php?rest_route=\/wp\/v2\/types\/ht_kb"}],"author":[{"embeddable":true,"href":"https:\/\/ai4healthcro.eu\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/ai4healthcro.eu\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=7249"}],"version-history":[{"count":9,"href":"https:\/\/ai4healthcro.eu\/index.php?rest_route=\/wp\/v2\/ht-kb\/7249\/revisions"}],"predecessor-version":[{"id":7260,"href":"https:\/\/ai4healthcro.eu\/index.php?rest_route=\/wp\/v2\/ht-kb\/7249\/revisions\/7260"}],"wp:attachment":[{"href":"https:\/\/ai4healthcro.eu\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=7249"}],"wp:term":[{"taxonomy":"ht_kb_category","embeddable":true,"href":"https:\/\/ai4healthcro.eu\/index.php?rest_route=%2Fwp%2Fv2%2Fht-kb-category&post=7249"},{"taxonomy":"ht_kb_tag","embeddable":true,"href":"https:\/\/ai4healthcro.eu\/index.php?rest_route=%2Fwp%2Fv2%2Fht-kb-tag&post=7249"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}