A5013466066- Radiomics and Machine Learning in Medical Imaging
- Cell Image Analysis Techniques
- AI in cancer detection
- Advanced Fluorescence Microscopy Techniques
- Breast Cancer Treatment Studies
- Image Processing Techniques and Applications
- MRI in cancer diagnosis
- Lung Cancer Diagnosis and Treatment
- Medical Imaging Techniques and Applications
- Emotion and Mood Recognition
- Microfluidic and Bio-sensing Technologies
- Single-cell and spatial transcriptomics
- Photoacoustic and Ultrasonic Imaging
- 3D Printing in Biomedical Research
- Head and Neck Cancer Studies
- Colorectal Cancer Screening and Detection
- Gene expression and cancer classification
- Cellular Mechanics and Interactions
- Cutaneous Melanoma Detection and Management
- Breast Implant and Reconstruction
- Image and Signal Denoising Methods
- Sparse and Compressive Sensing Techniques
- Lung Cancer Treatments and Mutations
- Advanced Image Processing Techniques
- Reconstructive Surgery and Microvascular Techniques
University of Rome Tor Vergata
2018-2025
Istituto Tumori Bari
2021-2024
Istituti di Ricovero e Cura a Carattere Scientifico
2021-2024
Recently, accurate machine learning and deep approaches have been dedicated to the investigation of breast cancer invasive disease events (IDEs), such as recurrence, contralateral second cancers. However, are poorly interpretable.
Abstract The dynamic contrast-enhanced MR imaging plays a crucial role in evaluating the effectiveness of neoadjuvant chemotherapy (NAC) even since its early stage through prediction final pathological complete response (pCR). In this study, we proposed transfer learning approach to predict if patient achieved pCR (pCR) or did not (non-pCR) by exploiting, separately combination, pre-treatment and early-treatment exams from I-SPY1 TRIAL public database. First, low-level features, i.e.,...
Contrast-enhanced spectral mammography (CESM) is an advanced instrument for breast care that still operator dependent. The aim of this paper the proposal automated system able to discriminate benign and malignant lesions based on radiomic analysis. We selected a set 58 regions interest (ROIs) extracted from 53 patients referred Istituto Tumori “Giovanni Paolo II” Bari (Italy) cancer screening phase between March 2017 June 2018. 464 features different kinds, such as points corners interest,...
Abstract We describe a novel method to achieve universal, massive, and fully automated analysis of cell motility behaviours, starting from time-lapse microscopy images. The approach was inspired by the recent successes in application machine learning for style recognition paintings artistic transfer. originality relies i) on generation atlas collection single-cell trajectories order visually encode multiple descriptors motility, ii) pre-trained Deep Learning Convolutional Neural Network...
Cancer treatment planning benefits from an accurate early prediction of the efficacy. The goal this study is to give three-year Breast Recurrence (BCR) for patients who underwent neoadjuvant chemotherapy. We addressed task a new perspective based on transfer learning applied pre-treatment and early-treatment DCE-MRI scans. Firstly, low-level features were automatically extracted MR images using pre-trained Convolutional Neural Network (CNN) architecture without human intervention....
In the last decades, unsupervised deep learning based methods have caught researchers' attention, since in many real applications, such as medical imaging, collecting a large amount of training examples is not always feasible. Moreover, construction good set time consuming and hard because selected data to be enough representative for task. this paper, we focus on Deep Image Prior (DIP) framework propose combine it with space-variant Total Variation regularizer an automatic estimation local...
Abstract In breast cancer patients, an accurate detection of the axillary lymph node metastasis status is essential for reducing distant occurrence probabilities. case patients resulted negative at both clinical and instrumental examination, nodal commonly evaluated performing sentinel lymph-node biopsy, that a time-consuming expensive intraoperative procedure (SLN) assessment. The aim this study was to predict 142 clinically by means radiomic features extracted from primary tumor ultrasound...
High-throughput phenotyping is becoming increasingly available thanks to analytical and bioinformatics approaches that enable the use of very high-dimensional data availability dynamic models link phenomena across levels: from genes cells, cells organs, through whole organism. The combination phenomics, deep learning, machine learning represents a strong potential for phenotypical investigation, leading way more embracing approach, called phenomics (MLP). In particular, in this work we...
To date, some artificial intelligence (AI) methods have exploited Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) to identify finer tumor properties as potential earlier indicators of pathological Complete Response (pCR) in breast cancer patients undergoing neoadjuvant chemotherapy (NAC). However, they work either for sagittal or axial MRI protocols. More flexible AI tools, be used easily clinical practice across various institutions accordance with its own imaging acquisition...
The application of deep learning on whole-slide histological images (WSIs) can reveal insights for clinical and basic tumor science investigations. Finding quantitative imaging biomarkers from WSIs directly the prediction disease-free survival (DFS) in stage I-III melanoma patients is crucial to optimize patient management. In this study, we designed a learning-based model with aim prognostic predict 1-year DFS cutaneous patients. First, referred cohort 43 (31 DF cases, 12 non-DF cases)...
Live-cell microscopy routinely provides massive amounts of time-lapse images complex cellular systems under various physiological or therapeutic conditions. However, this wealth data remains difficult to interpret in terms causal effects. Here, we describe CausalXtract, a flexible computational pipeline that discovers and possibly time-lagged effects from morphodynamic features cell–cell interactions live-cell imaging data. CausalXtract methodology combines network-based information-based...
Non-small cell lung cancer (NSCLC) represents 85% of all new diagnoses and presents a high recurrence rate after surgery. Thus, an accurate prediction risk in NSCLC patients at diagnosis could be essential to designate more aggressive medical treatments. In this manuscript, we apply transfer learning approach predict patients, exploiting only data acquired during its screening phase. Particularly, used public radiogenomic dataset having primary tumor CT image clinical information. Starting...
So far, baseline Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) has played a key role for the application of sophisticated artificial intelligence-based models using Convolutional Neural Networks (CNNs) to extract quantitative imaging information as earlier indicators pathological Complete Response (pCR) achievement in breast cancer patients treated with neoadjuvant chemotherapy (NAC). However, these did not exploit DCE-MRI exams their full geometry 3D volume but analysed...
Abstract Several studies have emphasised how positive and negative human papillomavirus (HPV+ HPV−, respectively) oropharyngeal squamous cell carcinoma (OPSCC) has distinct molecular profiles, tumor characteristics, disease outcomes. Different radiomics-based prediction models been proposed, by also using innovative techniques such as Convolutional Neural Networks (CNNs). Although some of these reached encouraging predictive performances, there evidence explaining the role radiomic features...
Abstract Cell-cell interactions are an observable manifestation of underlying complex biological processes occurring in response to diversified biochemical stimuli. Recent experiments with microfluidic devices and live cell imaging show that it is possible characterize kinematics via computerized algorithms unravel the effects targeted therapies. We study influence spatial temporal resolutions time-lapse videos on motility interaction descriptors computational models mimic dynamics among...
Background: For assessing the predictability of oncology neoadjuvant therapy results, background parenchymal enhancement (BPE) parameter in breast magnetic resonance imaging (MRI) has acquired increased interest. This work aims to qualitatively evaluate BPE as a potential predictive marker for therapy. Method: Three radiologists examined, triple-blind modality, MRIs 80 patients performed before start chemotherapy, after three months from treatment, and surgery. They identified portion...
One of the major problems in bioimaging, often highly underestimated, is whether features extracted for a discrimination or regression task will remain valid broader set similar experiments presence unpredictable perturbations during image acquisition process. Such an issue even more important when it addressed context deep learning due to lack priori known relationship between black-box descriptors (deep features) and phenotypic properties biological entities under study. In this regard,...
Abstract Non-Small cell lung cancer (NSCLC) is one of the most dangerous cancers, with 85% all new diagnoses and a 30–55% recurrence rate after surgery. Thus, an accurate prediction risk in NSCLC patients during diagnosis could be essential to drive targeted therapies preventing either overtreatment or undertreatment patients. The radiomic analysis CT images has already shown great potential solving this task; specifically, Convolutional Neural Networks (CNNs) have been proposed providing...
Abstract Background Accurate characterization of newly diagnosed a solid adnexal lesion is key step in defining the most appropriate therapeutic approach. Despite guidance from International Ovarian Tumor Analyzes Panel, evaluation these lesions can be challenging. Recent studies have demonstrated how machine learning techniques applied to clinical data solve this diagnostic problem. However, ML models often consider as black‐boxes due difficulty understanding decision‐making process used by...
Background: Oncology nurses support cancer patients in meeting their self-care needs, often neglecting own emotions and needs. This study aims to investigate the variations five facets of holistic mindfulness among Italian oncology based on gender, work experience oncology, shift work. Method: A cross-sectional was carried out 2023 amongst all registered who were employed an setting working Italy. Results: There no significant differences (p ≥ 0.05) according field, Conclusion: Could be...