Jorge Miguel Silva

Alexandre Cotorobai, Jorge Miguel Silva, Jose Luis Oliveira

Privacy and regulatory barriers often hinder centralized machine learning solutions, particularly in sectors like healthcare where data cannot be freely shared. Federated learning has emerged as a powerful paradigm to address these concerns; however, existing frameworks primarily support gradient-based models, leaving a gap for more interpretable, tree-based approaches. This paper introduces a federated learning framework for Random Forest classifiers that preserves data privacy and provides robust performance in distributed settings. By leveraging PySyft for secure, privacy-aware computation, our method enables multiple institutions to collaboratively train Random Forest models on locally stored data without exposing sensitive information. The framework supports weighted model averaging to account for varying data distributions, incremental learning to progressively refine models, and local evaluation to assess performance across heterogeneous datasets. Experiments on two real-world healthcare benchmarks demonstrate that the federated approach maintains competitive predictive accuracy - within a maximum 9\% margin of centralized methods - while satisfying stringent privacy requirements. These findings underscore the viability of tree-based federated learning for scenarios where data cannot be centralized due to regulatory, competitive, or technical constraints. The proposed solution addresses a notable gap in existing federated learning libraries, offering an adaptable tool for secure distributed machine learning tasks that demand both transparency and reliable performance. The tool is available at https://github.com/ieeta-pt/fed_rf.

Rui Lebre, Eduardo Pinho, Jorge Miguel Silva et al.

One of the most noticeable trends in healthcare over the last years is the continuous growth of data volume produced and its heterogeneity. In the medical imaging field, the evolution of digital systems is supported by the PACS concept and the DICOM standard. These technologies are deeply grounded in medical laboratories, supporting the production and providing healthcare practitioners with the ability to set up collaborative work environments with researchers and academia to study and improve healthcare practice. However, the complexity of those systems and protocols makes difficult and time-consuming to prototype new ideas or develop applied research, even for skilled users with training in those environments. Dicoogle emerges as a reference tool to achieve those objectives through a set of resources aggregated in the form of a learning pack. It is an open-source PACS archive that, on the one hand, provides a comprehensive view of the PACS and DICOM technologies and, on the other hand, provides the user with tools to easily expand its core functionalities. This paper describes the Dicoogle framework, with particular emphasis in its Learning Pack package, the resources available and the impact of the platform in research and academia. It starts by presenting an overview of its architectural concept, the most recent research backed up by Dicoogle, some remarks obtained from its use in teaching, and worldwide usage statistics of the software. Moreover, a comparison between the Dicoogle platform and the most popular open-source PACS in the market is presented.

Jorge Miguel Silva

Finding the shortest program that generates a sequence is uncomputable, and for six decades that fact has been mistaken for a wall around finding any generating program. It is not a wall but a price, and this paper measures it. For every algorithm that learns about a candidate program only through its score, a class spanning Levin search, evolutionary methods, simulated annealing, and the cross-entropy method, we define the coupling width of a search problem and prove an unconditional worst-case lower bound, exponential in that width with base one less than the domain size. From it follows a conservation law: structural knowledge injected into a search trades one for one against the search it removes, and their sum can never fall below the length of the program sought. Levin's 1973 upper bound and the lower bound proved here are the two ends of one conserved quantity, closing on each other as the instruction set grows. The only escape is to read a candidate's structure rather than its score, and its price, which we prove for generic targets, is incompleteness. A deterministic engine built on this theory recovers a generating program, certified by compressing its data and predicting an unseen continuation, for 2,383 of 3,914 sequences across four independent populations, including 244 of the 256 elementary cellular automata, with measured discovery cost rising along program length more than an order of magnitude inside the score-oracle worst case.

Jorge Miguel Silva, Diogo Pratas, Rui Antunes et al.

The artistic community is increasingly relying on automatic computational analysis for authentication and classification of artistic paintings. In this paper, we identify hidden patterns and relationships present in artistic paintings by analysing their complexity, a measure that quantifies the sum of characteristics of an object. Specifically, we apply Normalized Compression (NC) and the Block Decomposition Method (BDM) to a dataset of 4,266 paintings from 91 authors and examine the potential of these information-based measures as descriptors of artistic paintings. Both measures consistently described the equivalent types of paintings, authors, and artistic movements. Moreover, combining the NC with a measure of the roughness of the paintings creates an efficient stylistic descriptor. Furthermore, by quantifying the local information of each painting, we define a fingerprint that describes critical information regarding the artists' style, their artistic influences, and shared techniques. More fundamentally, this information describes how each author typically composes and distributes the elements across the canvas and, therefore, how their work is perceived. Finally, we demonstrate that regional complexity and two-point height difference correlation function are useful auxiliary features that improve current methodologies in style and author classification of artistic paintings. The whole study is supported by an extensive website (http://panther.web.ua.pt) for fast author characterization and authentication.

Jorge Miguel Silva, Tiago Marques Godinho, David Silva et al.

Digital medical imaging laboratories contain many distinct types of equipment provided by different manufacturers. Interoperability is a critical issue and the DICOM protocol is a de facto standard in those environments. However, manufacturers' implementation of the standard may have non-conformities at several levels, which will hinder systems' integration. Moreover, medical staff may be responsible for data inconsistencies when entering data. Those situations severely affect the quality of healthcare services since they can disrupt system operations. The existence of software able to confirm data quality and compliance with the DICOM standard is important for programmers, IT staff and healthcare technicians. Although there are a few solutions that try to accomplish this goal, they are unable to deal with certain situations that require user input. Furthermore, these cases usually require the setup of a working environment, which makes the sharing of validation information more difficult. This article proposes and describes the development of a Web DICOM validation service for the community. This solution requires no configuration by the user, promotes validation results share-ability in the community and preserves patient data privacy since files are de-identified on the client side.