Daniel聽De聽Castro
Senior Researcher at Imperial College London
Cambridge, England, United Kingdom
Summary
馃懁
Senior馃帗
Top SchoolDaniel De Castro is a Senior Researcher and final-year PhD candidate at Imperial College London specialising in machine learning for medical imaging, with a decade of experience bridging research and applied engineering. Based in Cambridge, he combines probabilistic modelling and causality to integrate clinical and demographic data into generative image models, producing population- and individual-level causal insights. He currently collaborates part-time with Kheiron Medical and is embedded at Microsoft Research, where he has contributed to notable open-source tooling such as InnerEye-DeepLearning鈥攁dding cross-validation, custom loss support and evaluation reporting for 3D medical segmentation and classification. His background spans industry internships and academic teaching, and he holds degrees from PUC-Rio, 脡cole Centrale Paris and Imperial College, reflecting a rare mix of rigorous engineering, deployed ML systems, and translational healthcare focus.
10 years of coding experience
Dipl么me d'Ing茅nieur, Engineering, Dipl么me d'Ing茅nieur, Engineering at Ecole Centrale ParisMaster of Research (MRes), Advanced Computing, Distinction, Master of Research (MRes), Advanced Computing, Distinction at Imperial College LondonPontifical Catholic University of Rio de Janeiro
Portuguese, English, French, German, Spanish
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (13)
scikit-learn10
scikit10
medical-imaging10
pytorch10
machine-learning10
azure10
deep-learning10
microsoft-azure10
python10
mask-rcnn9
faster-rcnn9
data-science8
tensorflow7
Programming languages (3)
Github contributions (5)
microsoft/InnerEye-DeepLearning
Nov 2020 - Mar 2022
Medical Imaging Deep Learning library to train and deploy 3D segmentation models on Azure Machine Learning
Role in this project:
ML Engineer
ML EngineerContributions:117 reviews, 25 commits, 23 PRs in 1 year 4 months
Contributions summary:Daniel primarily contributed to the development and improvement of a medical imaging deep learning library. Their work included implementing grouped dataset splits and cross-validation, crucial for model training and evaluation. They also added cross-validation reporting for classification models, enhancing the tools for analyzing model performance. Furthermore, the user made significant changes to enable custom loss functions, prediction targets, and reporting, indicating their involvement in expanding the library's flexibility and capabilities.
deep-learningmachine-learningazure-machine-learningdeep-learning-libraryimaging
dccastro/Morpho-MNIST
Apr 2018 - Nov 2021
Morpho-MNIST: Quantitative Assessment and Diagnostics for Representation Learning (http://jmlr.org/papers/v20/19-033.html)
Contributions:1 review, 115 commits, 5 PRs in 3 years 8 months
pytorchrepresentationjmlrrepresentation-learningquantitative
Find and Hire Top DevelopersWe鈥檝e analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.