A5018488419- Medical Imaging and Analysis
- Scoliosis diagnosis and treatment
- Spine and Intervertebral Disc Pathology
- Spinal Fractures and Fixation Techniques
- Pelvic and Acetabular Injuries
- Musculoskeletal pain and rehabilitation
- Shoulder Injury and Treatment
- Artificial Intelligence in Healthcare and Education
- Radiomics and Machine Learning in Medical Imaging
- AI in cancer detection
- Clinical practice guidelines implementation
- Reliability and Agreement in Measurement
- Health Systems, Economic Evaluations, Quality of Life
- Radiology practices and education
- Orthopedic Surgery and Rehabilitation
- Bioinformatics and Genomic Networks
- Aortic Disease and Treatment Approaches
- Coronary Artery Anomalies
- Single-cell and spatial transcriptomics
- Biomedical Text Mining and Ontologies
- Topic Modeling
- Delphi Technique in Research
- Aortic aneurysm repair treatments
- Advanced X-ray and CT Imaging
- Gene expression and cancer classification
Schulthess-Klinik
2023-2025
ETH Zurich
2023-2024
Istituto Ortopedico Galeazzi
2020-2022
Istituti di Ricovero e Cura a Carattere Scientifico
2021-2022
Italian Institute of Technology
2022
Istituto Ortopedico Gaetano Pini
2021
University of Milano-Bicocca
2019
Polytechnic University of Turin
2019
"Garbage in, garbage out" summarises well the importance of high-quality data in machine learning and artificial intelligence. All used to train validate models should indeed be consistent, standardised, traceable, correctly annotated, de-identified, considering local regulations. This narrative review presents a summary techniques that are ensure all these requirements fulfilled, with special emphasis on radiological imaging freely available software solutions can directly employed by...
Problem: Clinical practice requires the production of a time- and resource-consuming great amount notes. They contain relevant information, but their secondary use is almost impossible, due to unstructured nature. Researchers are trying address this problems, with traditional promising novel techniques. Application in real hospital settings seems not be possible yet, though, both because relatively small dirty dataset, for lack language-specific pre-trained models. Aim: Our aim demonstrate...
Abstract In this work we propose to use Deep Learning automatically calculate the coordinates of vertebral corners in sagittal x-rays images thoracolumbar spine and, from those landmarks, relevant radiological parameters such as L1–L5 and L1–S1 lordosis sacral slope. For purpose, used 10,193 annotated with landmarks ground truth. We realized a model that consists 2 steps. step 1, trained Convolutional Neural Networks identify each vertebra image respectively. 2, refined localization using...
Abstract Background Minimizing radiation exposure is crucial in monitoring adolescent idiopathic scoliosis (AIS). Generative adversarial networks (GANs) have emerged as valuable tools being able to generate high-quality synthetic images. This study explores the use of GANs sagittal radiographs from coronal views AIS patients. Methods A dataset 3,935 patients who underwent spine and pelvis radiographic examinations using EOS system, which simultaneously acquires images, was analyzed. The...
Chronic low back pain (CLBP) is a prevalent condition significantly reducing quality of life. Lumbar steroid injections are widely used conservative treatment option, but their effectiveness varies among patients. This study aimed to develop predictive framework that integrates clinical variables and patient demographics evaluate post-treatment satisfaction in CLBP patients undergoing lumbar injection therapy. We performed retrospective analysis 212 the intensity changes using Numerical...
Background and Objectives: Commonly being the first step in trauma routine imaging, up to 67% fractures are missed on plain radiographs of thoracolumbar (TL) spine. The aim this study was develop a deep learning model that detects traumatic sagittal TL Identifying vertebral simple radiographic projections would have significant clinical financial impact, especially for low- middle-income countries where computed tomography (CT) magnetic resonance imaging (MRI) not readily available could...
Machine learning (ML), a subset of artificial intelligence, is crucial for spine care and research due to its ability improve treatment selection outcomes, leveraging the vast amounts data generated in health more accurate diagnoses decision support. ML's potential particularly notable radiological image analysis, including localization labeling anatomical structures, detection classification findings, prediction clinical thereby paving way personalized medicine. The manuscript discusses...
The study aims to create a preoperative model from baseline demographic and health-related quality of life scores (HRQOL) predict good excellent early clinical outcome using machine learning (ML) approach. A single spine surgery center retrospective review prospectively collected data January 2016 December 2020 the institutional registry (SpineREG) was performed. inclusion criteria were age ≥ 18 years, both sexes, lumbar arthrodesis procedure, complete follow up assessment (Oswestry...
Abstract Purpose The study aims to assess if the angle of trunk rotation (ATR) in combination with other readily measurable clinical parameters allows for effective non-invasive scoliosis screening. Methods We analysed 10,813 patients (4–18 years old) who underwent and radiological evaluation a tertiary clinic specialised spinal deformities. considered as predictors ATR, Prominence (mm), visible asymmetry waist, scapulae shoulders, familiarity, sex, BMI, age, menarche, localisation curve....
Study Design: Retrospective study. Objectives: Huge amounts of images and medical reports are being generated in radiology departments. While these datasets can potentially be employed to train artificial intelligence tools detect findings on radiological images, the unstructured nature limits accessibility information. In this study, we tested if natural language processing (NLP) useful generate training data for deep learning models analyzing planar radiographs lumbar spine. Methods: NLP...
We developed and used a deep learning tool to process biplanar radiographs of 9,832 non-surgical patients suffering from spinal deformities, with the aim reporting statistical distribution radiological parameters describing shape correlations interdependencies between them. An existing able automatically perform three-dimensional reconstruction thoracolumbar spine has been improved analyze large set trunk. For all patients, following were calculated: spinopelvic parameters; lumbar lordosis;...
Background: Aortic dilatation is common in hypertensive patients and associated with higher risk of cardiovascular events. Parameters predicting further during lifetime are poorly understood. Aim: To predict the midterm aortic diameter evolution a cohort known at Sinus Valsalva (SOV) level. Methods: We prospectively analyzed essential outpatients without any other factor for dilatation. They underwent serial echocardiographic evaluations from 2003 to 2016. Results: Two hundred forty-two...
Abstract Adolescent idiopathic scoliosis is a three-dimensional deformity of the spine which frequently corrected with implantation instrumentation generally good or excellent clinical results; mechanical post-operative complications such as implant loosening and breakage are however relatively frequent. The rate associated lack consensus about surgical decision-making process; choices length, anchoring implants degree correction indeed mostly based on personal views previous experience...
Abstract Aims To create, using a machine learning (ML) approach, preoperative model from baseline demographic and health-related quality of life scores (HRQOL) to predict good excellent early clinical outcome. Patients Methods A single spine surgery center retrospective review prospectively collected data January 2016 December 2020 the institutional registry (SpineREG) was performed. The inclusion criteria were age ≥ 18 years, both sexes, lumbar arthrodesis procedure, complete follow up...
Retrospective data analysis.This study aims to develop a deep learning model for the automatic calculation of some important spine parameters from lateral cervical radiographs.We collected two datasets different institutions. The first dataset 1498 images was used train and optimize find best hyperparameters while second 79 as an external validation set evaluate robustness generalizability our model. performance assessed by calculating median absolute errors between prediction ground truth...
Advances in next-generation sequencing have provided high-dimensional RNA-seq datasets, allowing the stratification of some tumor patients based on their transcriptomic profiles. Machine learning methods been used to reduce and cluster data. Recently, uniform manifold approximation projection (UMAP) was applied project genomic datasets low-dimensional Euclidean latent space. Here, we evaluated how different representations UMAP embedding can impact analysis breast cancer (BC) stratification....