The proposal aims to address the issue of the expensive and inaccessible research publishing industry by leveraging modern internet-based social technology to create an open reviewing and quality-ranking web portal. The portal will facilitate manuscript submission and an automated, free, community-based open access, peer review, and quality-rating system. The project will develop a proof of concept portal using ReactJS/NextJS and TailwindCSS for the frontend, Django for the backend, and PostgreSQL as the database server. The deliverables include a functional web portal that allows for detailed comments, ratings, and reputation-based filtering, which can improve the quality of research and provide a platform for communities of reviewers to form and manage the reviewing process. The system will eventually be tested with the Aperture Neuro open access publishing platform of the Organization for Human Brain Mapping (OHBM), in consultation with that community.

Neuroptimus, an open-source parameter optimization software tailored for neuroscience applications, has been instrumental in advancing the construction and optimization of biologically detailed models, utilizing algorithms such as evolutionary algorithms and swarm intelligence. Neuroptimus includes a graphical user interface (GUI), and works on multiple platforms including PCs and supercomputers. Building upon its success, this proposal aims to enhance the GUI of Neuroptimus to further streamline the parameter optimization process.The project targets to address key areas of improvement within both the graphical user interface and command-line interface of Neuroptimus. By integrating missing functionalities and refining existing features, the goal is to enhance user experience, flexibility, and accessibility. Specific enhancements include enabling users to save and load optimization settings seamlessly within the GUI, providing real-time monitoring of optimization progress, enhancing the interpretability of optimization results through various output formats, and facilitating analysis and visualization of both intermediate and final optimization outcomes. Through these enhancements, the Neuroptimus GUI will become more informative, user-friendly, and conducive to efficient parameter optimization for biophysical models in neuroscience research. Project Deliverables: 1- Implementing Saving and Loading Optimization Settings. 2- Implementing Progress Monitoring Functionality. 3- Implementing Saving Optimization Results in an Interpretable Format. 4- Extending UI with Visualization Options for Final and Intermediate Results. 5- Researching and Proposing Additional UI/UX Enhancements.

The proposal aims to address the issue of the expensive and inaccessible research publishing industry by leveraging modern internet-based social technology to create an open reviewing and quality-ranking web portal. The portal will facilitate manuscript submission and an automated, free, community-based open access, peer review, and quality-rating system. The project will develop a proof of concept portal using ReactJS/NextJS and TailwindCSS for the frontend, Django for the backend, and PostgreSQL as the database server. The deliverables include a functional web portal that allows for detailed comments, ratings, and reputation-based filtering, which can improve the quality of research and provide a platform for communities of reviewers to form and manage the reviewing process. The system will eventually be tested with the Aperture Neuro open access publishing platform of the Organization for Human Brain Mapping (OHBM), in consultation with that community.

Ensuring healthy vision in infants requires early detection of potential problems. Current Infant Visual Function Measurement Systems built in previous GSOC 22 and GSOC 23 are efficient and robust solutions but are limited by difficulties with infant cooperation and shortcomings in eye-tracking technology. To address this, a comprehensive upgrade is proposed. A user-friendly and visually engaging interface will be built for researchers and clinicians. Additionally, a wider variety of captivating stimuli, both visual and auditory, will be incorporated to keep infants engaged during testing. The core functionality, eye-tracking, will be significantly improved by exploring advanced deep learning techniques. This will enable precise gaze estimation in all directions, including up and down, providing a more complete picture of an infant's visual attention. Furthermore, a novel test will be developed to identify potential vision issues associated with anisometropic amblyopia. Finally, the system will be designed for seamless integration with various external eye-tracking hardware devices, increasing its flexibility and applicability. This project strives to create a user-friendly and robust Infant Visual Function Measurement System, empowering researchers and clinicians with a powerful tool for early detection and intervention, ultimately leading to better visual development for young children.

The annotation of research data is essential to ensure its findability and reusability. High-quality data annotations require domain expertise, so that annotations are relevant, but also specific technical skills, such as knowing how to handle JSON/XML files. Additionally, people are often reluctant to change their workflows, and the technical affordances in the case of data annotation intensify this challenge. As a result, researchers tend to stick to their “data handling traditions” as soon as data operations become too complex. Unfortunately, this often means that even though projects like Neurobagel are working hard to make life easier for researchers, these tools are not widely adopted. My idea for contributing to the Neurobagel project is to combine a user-friendly interface with a Large Language Model (LLM) approach to make the annotation of tabular data even more effortless for researchers. For the end user, the process should be to provide a tabular file and get a first-pass annotation for review without any intermediate steps. From a technical perspective, this should be accomplished by using a Large Language Model (LLM) to categorize columns, such as participant ID or age. To improve the predictions of the LLM, already annotated data will be linked to explicit knowledge from existing ontologies such as SNOMED CT or the Cognitive Atlas and used to provide context for the LLM.

The proposal involves contributing to the LORIS codebase, a pivotal resource for neuroscience research. LORIS plays a crucial role in facilitating Open Science and advancing global research by offering an open-stack platform for the management of diverse medical data. My work will include but will not be limited to enhancing the system's functionality through new modules, increasing test coverage, and developing new API endpoints.

SciCommons aims to revolutionize academic publishing by providing an open peer review platform. The current system faces challenges such as time constraints and limited reviewer capacity. SciCommons addresses these by allowing any user to review, comment, and rate research articles, enhancing article quality and decreasing review time. Anonymity ensures ethical and unbiased reviews. Building upon its existing functionalities, the project will introduce private communities for exclusive peer review, an immersive discussion/chat sections for each article. Deliverables include modularizing the frontend, enhancing user experience, implementing small feature enhancements, and integrating private communities and discussion channels. The project will prioritize code quality and adhere to open-source standards. By democratizing scholarly discourse, SciCommons fosters collaboration and transparent peer review.

I am proposing a project for Google Summer of Code to develop an interactive SWC to NeuroML converter in Python. The converter will simplify the conversion of the widely used SWC format to the NeuroML standard for biophysically detailed neuronal models. The project will include an interactive tool for users to utilise, and it will be tested against neuronal reconstructions from NeuroMorpho.org

The Human Neocortical Neurosolver (HNN) is a computational tool that enables scientists to study brain responses at the cellular and circuit levels. However, the current version of HNN-core has a limitation in that it cannot simulate large batches of simulations. This hinders the efficient optimization of parameters, which is crucial for accurate modeling. To address this issue, I am working on improving HNN-core by adding batch simulation functionality. I am also developing tutorials on advanced optimization techniques, specifically focusing on simulation-based inference (SBI), a deep learning-based Bayesian inference method. These enhancements will significantly contribute to the HNN's codebase, enabling researchers to extract crucial insights for developing theories about human brain response origins.

Agent-based Models (ABM) use both first-person and third-person perspectives. This project aims to develop an AR/VR experiment application which is compatible with all devices such as Meta quest and Web XR. This will in turn allow learners and researchers from around the world to access and interact with the models. This application will also incorporate haptic feedback for users to make it more interactive. A well known experimental paradigm (e.g., Morris Water Maze), where the subject can interact with both the environment and stimuli, will be used. This environment should allow for participant data to be collected, and will exhibit a high degree of environmental realism.