A5016189819- Colorectal Cancer Screening and Detection
- Logic, Reasoning, and Knowledge
- Surgical Simulation and Training
- Colorectal Cancer Surgical Treatments
- Machine Learning in Healthcare
- Radiomics and Machine Learning in Medical Imaging
- Medical Image Segmentation Techniques
- Liver Disease Diagnosis and Treatment
- Advanced Neural Network Applications
- Cerebrovascular and Carotid Artery Diseases
- Cell Image Analysis Techniques
- COVID-19 diagnosis using AI
- Semantic Web and Ontologies
- Retinal Imaging and Analysis
- Artificial Intelligence in Healthcare
- AI-based Problem Solving and Planning
- Explainable Artificial Intelligence (XAI)
- Multimodal Machine Learning Applications
- Cardiovascular Health and Risk Factors
- AI in cancer detection
- Gene expression and cancer classification
- Peripheral Artery Disease Management
- Medical Imaging Techniques and Applications
- Time Series Analysis and Forecasting
- Artificial Intelligence in Healthcare and Education
University of Calabria
2018-2025
German Cancer Research Center
2020-2023
Heidelberg University
2020-2023
Abstract Image-based tracking of medical instruments is an integral part surgical data science applications. Previous research has addressed the tasks detecting, segmenting and based on laparoscopic video data. However, proposed methods still tend to fail when applied challenging images do not generalize well they have been trained on. This paper introduces Heidelberg Colorectal (HeiCo) set - first publicly available enabling comprehensive benchmarking instrument detection segmentation...
Abstract Continual Learning (CL) is a novel AI paradigm in which tasks and data are made available over time; thus, the trained model computed on basis of stream data. CL-based approaches able to learn new skills knowledge without forgetting previous ones, with no guaranteed access previously encountered data, mitigating so-called “catastrophic forgetting” phenomenon. Interestingly, by making systems improve time need for large amounts or computational resources, CL can help at reducing...
Intraoperative tracking of laparoscopic instruments is often a prerequisite for computer and robotic-assisted interventions. While numerous methods detecting, segmenting medical based on endoscopic video images have been proposed in the literature, key limitations remain to be addressed: Firstly, robustness, that is, reliable performance state-of-the-art when run challenging (e.g. presence blood, smoke or motion artifacts). Secondly, generalization; algorithms trained specific intervention...
Intraoperative tracking of laparoscopic instruments is often a prerequisite for computer and robotic-assisted interventions. While numerous methods detecting, segmenting medical based on endoscopic video images have been proposed in the literature, key limitations remain to be addressed: Firstly, robustness, that is, reliable performance state-of-the-art when run challenging (e.g. presence blood, smoke or motion artifacts). Secondly, generalization; algorithms trained specific intervention...
Challenges have become the state-of-the-art approach to benchmark image analysis algorithms in a comparative manner. While validation on identical data sets was great step forward, results is often restricted pure ranking tables, leaving relevant questions unanswered. Specifically, little effort has been put into systematic investigation what characterizes images which fail. To address this gap literature, we (1) present statistical framework for learning from challenges and (2) instantiate...
We present a novel framework for disease classification from high-dimensional gene expression data or several characteristic of patients. take advantage Principle Component Analysis to perform dimensionality reduction and heatmaps embedding the complex information in 2-D image, we make use convolutional neural network different tumor types. Experimental analyses show that proposed method achieves good performance, encourages its application other genomic pathological context.
Abstract The assessment of vascular complexity in the lower limbs provides relevant information about peripheral artery occlusive diseases (PAOD), thus fostering improvements both therapeutic decisions and prognostic estimation. current clinical practice consists visually inspecting evaluating cine-angiograms interested region, which is largely operator-dependent. We present here an automatic method for segmenting vessel tree compute a quantitative measure, terms fractal dimension (FD),...
Providing accurate diagnosis of diseases generally requires complex analyses many clinical, biological and pathological variables. In this context, solutions based on machine learning techniques achieved relevant results in specific disease detection classification, can hence provide significant clinical decision support. However, such approaches suffer from the lack proper means for interpreting choices made by models, especially case deep-learning ones. order to improve interpretability...
X-ray computed microtomography ({\mu}-CT) is a non-destructive technique that can generate high-resolution 3D images of the internal anatomy medical and biological samples. These enable clinicians to examine gain insights into disease or anatomical morphology. However, extracting relevant information from requires semantic segmentation regions interest, which usually done manually results time-consuming tedious. In this work, we propose novel framework uses convolutional neural network (CNN)...
Challenges have become the state-of-the-art approach to benchmark image analysis algorithms in a comparative manner. While validation on identical data sets was great step forward, results is often restricted pure ranking tables, leaving relevant questions unanswered. Specifically, little effort has been put into systematic investigation what characterizes images which fail. To address this gap literature, we (1) present statistical framework for learning from challenges and (2) instantiate...