Vanja Mišković

Vanja Mišković

Auto-generated profile, based on publicly available information

Vanja Mišković

Medical Oncology Department, Fondazione IRCCS Istituto Nazionale dei Tumori di Milano, Milan, Italy

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 December 2023 in Nanotechnology
Verified
Journal Article
Open access
PDF file
Teodora VicenticTeodora Vicentic
Teodora Vicentic
IGImmacolata Greco
Immacolata Greco
CICarlo Saverio Iorio
Carlo Saverio Iorio
VMVanja Miskovic
Vanja Miskovic
Danica Bajuk-BogdanovićDanica Bajuk-Bogdanović
Danica Bajuk-Bogdanović
Marko SpasenovicMarko Spasenovic
Marko Spasenovic
Laser-induced graphene (LIG) possesses desirable properties for numerous applications. However, LIG formation on biocompatible substrates is needed to further augment the integration of LIG-based technologies into nanobiotechnology. Here, LIG formation on cross-linked sodium alginate is reported. The LIG is systematically investigated, providing a comprehensive understanding of the physicochemical characteristics of the material. Raman spectroscopy, scanning electron microscopy with energy-dispersive x-ray analysis, x-ray diffraction, transmission electron microscopy, Fourier-transform infrared spectroscopy and x-ray photoelectron spectroscopy techniques confirm the successful generation of oxidized graphene on the surface of cross-linked sodium alginate. The influence of laser parameters and the amount of crosslinker incorporated into the alginate substrate is explored, revealing that lower laser speed, higher resolution, and increased CaCl2 content leads to LIG with lower electrical resistance. These findings could have significant implications for the fabrication of LIG on alginate with tailored conductive properties, but they could also play a guiding role for LIG formation on other biocompatible substrates.

17 April 2023 in IEEE Transactions on Biomedical Engineering
Verified
Journal Article
Closed access
ARAndrea Rozo
Andrea Rozo
VMVanja Miskovic
Vanja Miskovic
TRThomas Rose
Thomas Rose
EKElkana Keersebilck
Elkana Keersebilck
CICarlo Iorio
Carlo Iorio
CVCarolina Varon
Carolina Varon
Objective: An accurate and timely diagnosis of burn severity is critical to ensure a positive outcome. Laser Doppler imaging (LDI) has become a very useful tool for this task. It measures the perfusion of the burn and estimates its potential healing time. LDIs generate a 6-color palette image, with each color representing a healing time. This technique has very high costs associated. In resource-limited areas, such as low- and middle-income countries or remote locations like space, where access to specialized burn care is inadequate, more affordable and portable tools are required. This study proposes a novel image-to-image translation approach to estimate burn healing times, using a digital image to approximate the LDI. Methods: This approach consists of a U-net architecture with a VGG-based encoder and applies the concept of ordinal classification. Paired digital and LDI images of burns were collected. The performance was evaluated with 10-fold cross-validation, mean absolute error (MAE), and color distribution differences between the ground truth and the estimated LDI. Results: Results showed a satisfactory performance in terms of low MAE ( $0.2370 \pm 0.0086$ ). However, the unbalanced distribution of colors in the data affects this performance. Significance: This novel and unique approach serves as a basis for developing more accessible support tools in the burn care environment in resource-limited areas.

8 August 2024 in npj Microgravity
Verified
Journal Article
Open access
PDF file
VMVanja Mišković
Vanja Mišković
IGImmacolata Greco
Immacolata Greco
CMChristophe Minetti
Christophe Minetti
FCFrancesca Cialdai
Francesca Cialdai
MMMonica Monici
Monica Monici
CICarlo Saverio Iorio
Carlo Saverio Iorio
Exposure to altered gravity influences cellular behaviour in cell cultures. Hydrogels are amongst the most common materials used to produce tissue-engineering scaffolds, and their mechanical properties play a crucial role in cell-matrix interaction. However, little is known about the influence of altered gravity on hydrogel properties. Here we study the mechanical properties of Poly (ethylene glycol) diacrylate (PEGDA) and PEGDA incorporated with graphene oxide (GO) by performing tensile tests in micro and hypergravity during a Parabolic flight campaign, and by comparing them to the same tests performed in Earth gravity. We show that gravity levels do not result in a statistically significant difference in Young’s modulus.

13 September 2026 in Nature Medicine
Verified
Journal Article
Open access
PDF file
APARSELA PRELAJ
ARSELA PRELAJ
VMVanja Miskovic
Vanja Miskovic
MSMatteo Sacco
Matteo Sacco
AFAlberto Ferrarin
Alberto Ferrarin
CLCristina Maria Licciardello
Cristina Maria Licciardello
GSGiovanni Scoazec
Giovanni Scoazec
Despite a decade in, immunotherapy (IO) treatment selection in non-small cell lung cancer (NSCLC) remains largely guided by subgroup analyses and imperfect programmed death ligand 1 (PD-L1) and clinical scores. To our knowledge, I 3 LUNG ( NCT05537922 ) is currently the largest international, real-world, multimodal, artificial intelligence (AI)-based study, enrolling 2,396 patients. We integrated real-world clinical and blood (CB) data, computed tomography (CT) images, digital pathology (DP), and genomics into machine learning early fusion (MLEF) and deep learning intermediate fusion (DLIF) models. Machine learning (ML) and deep learning (DL) CB-only models achieved consistent performance across outcomes with area under the curve (AUC) up to 0.77 in the test (TEST) set. Performance drop in external validation (EXVAL) likely reflects population differences (AUC range: 0.55–0.72). AI models significantly surpassed PD-L1, Eastern Cooperative Oncology Group performance status (ECOG PS), neutrophil-to-lymphocyte ratio (NLR), lactate dehydrogenase (LDH) and Lung Immune Prognostic Index (LIPI) score in the independent TEST set. The clinical usability study showed that lung expert and nonexpert physicians improved their prediction with the explainable AI (XAI) ML CB-only based tool. Although multimodal integration with MLEF (CB+CT+DP) was associated with higher performance, its incremental benefit remains uncertain, not translated in TEST and EXVAL. The I 3 LUNG project is a pioneering framework showing the clinical usefulness of AI tools. A prospective validation of the decision support system (both CB and multimodal) is currently undergoing in more than 2,000 patients.

21 October 2025 in medRxiv
Verified
Preprint
Open access
PDF file
FCFederica Corso
Federica Corso
VPVittoria Peppoloni
Vittoria Peppoloni
LMLaura Mazzeo
Laura Mazzeo
GLGiuseppe Leone
Giuseppe Leone
LPLuana Passos
Luana Passos
APARSELA PRELAJ
ARSELA PRELAJ
Real-world data (RWD), largely stored in unstructured electronic health records (EHRs), are critical for understanding complex diseases like cancer. However, extracting structured information from these narratives is challenging due to linguistic variability, semantic complexity, and privacy concerns. This study evaluates the performance of four locally deployable and small language models (SLMs), LLaMA, Mistral, BioMistral, and MedLLaMA, for information extraction (IE) from Italian EHRs within the APOLLO 11 trial on non-small cell lung cancer (NSCLC). We examined three prompting strategies (zero-shot, few-shot, and annotated few-shot) across English and Italian, involving clinicians with varying expertise to assess prompt design’s impact on accuracy. Results show that general-purpose models (e.g., LLaMA 3.1 8B) outperform biomedical models in most tasks, particularly in extracting binary features. Multiclass variables such as TNM staging, PD-L1, and ECOG were more difficult due to implicit language and lack of standardization. Few-shot prompting and native-language inputs significantly improved performance and reduced hallucinations. Clinical expertise enhanced consistency in annotation, particularly among students using annotated examples. The study confirms that privacy-preserving SLMs can be deployed locally for efficient and secure cancer data extraction. Findings highlight the need for hybrid systems combining SLMs with expert input and underline the importance of aligning clinical documentation practices with SLM capabilities. This is the first study to benchmark SLMs on Italian EHRs and investigate the role of clinical expertise in prompt engineering, offering valuable insights for the future integration of SLMs into real-world clinical workflows.

16 July 2026 in Communications Medicine
Verified
Journal Article
Open access
PDF file
FCFederica Corso
Federica Corso
VPVittoria Peppoloni
Vittoria Peppoloni
LMLaura Mazzeo
Laura Mazzeo
GLGiuseppe Leone
Giuseppe Leone
LPLuana Passos
Luana Passos
APArsela Prelaj
Arsela Prelaj
Background Real-world data (RWD) in unstructured electronic health records (EHRs) is crucial for understanding complex diseases like cancer, but extracting structured information is challenging due to linguistic variability, semantic complexity, and privacy concerns. This study evaluates the performance of four small, locally deployable language models for information extraction from Italian EHRs. Methods We examine three prompting strategies (zero-shot, few-shot, and annotated few-shot) across English and Italian, involving clinicians with varying expertise to assess the impact of prompt design on accuracy. We evaluate the performance of four open-source small language models (SLMs) for clinical information extraction from Italian electronic health records (EHRs) in the APOLLO 11 trial on non-small cell lung cancer (NSCLC). The extraction protocol involves four steps: problem definition, data preprocessing, Large Language Model (LLM)-based information extraction, and output evaluation. Results We show that general-purpose models (e.g., LLaMA 3.1 8B) outperform biomedical models in most tasks, particularly in extracting binary features. Multiclass variables such as TNM (Tumor, Node, Metastasis) staging, PD-L1 (Programmed death-ligand 1), and ECOG-PS (Eastern Cooperative Oncology Group-Performance Status) are more difficult due to implicit language and lack of standardization. Few-shot prompting and native-language inputs significantly improve performance and reduced hallucinations. Clinical expertise enhances consistency in the extraction, particularly among students using annotated examples. Conclusions The study confirms that privacy-preserving SLMs can be deployed locally for efficient and secure cancer data extraction. Findings highlight the need for hybrid systems combining SLMs with expert input and underline the importance of aligning clinical documentation practices with SLM capabilities. This is the first study to benchmark SLMs on Italian EHRs and investigate the role of clinical expertise in prompt engineering, offering valuable insights for the future integration of SLMs into real-world clinical workflows.