Unipa-GPT: Large Language Models for university-oriented QA in Italian

This paper illustrates the architecture and training of Unipa-GPT, a chatbot relying on a Large Language Model, developed for assisting students in choosing a bachelor/master degree course at the University of Palermo. Unipa-GPT relies on gpt-3.5-turbo, it was presented in the context of the European Researchers' Night (SHARPER night). In our experiments we adopted both the Retrieval Augmented Generation (RAG) approach and fine-tuning to develop the system. The whole architecture of Unipa-GPT is presented, both the RAG and the fine-tuned systems are compared, and a brief discussion on their performance is reported. Further comparison with other Large Language Models and the experimental results during the SHARPER night are illustrated.

Conditioning Chat-GPT for information retrieval: the Unipa-GPT case study

This paper illustrates the architecture and training of Unipa-GPT, a chatbot relying on a Large Language Model, developed for assisting students in choosing a bachelor/master degree course at the University of Palermo. Unipa-GPT relies on gpt-3.5-turbo, it was presented in the context of the European Researchers' Night (SHARPER night). In our experiments we adopted both the Retrieval Augmented Generation (RAG) approach and fine-tuning to develop the system. The whole architecture of Unipa-GPT is presented, both the RAG and the fine-tuned systems are compared, and a brief discussion on their performance is reported. Further comparison with other Large Language Models and the experimental results during the SHARPER night are illustrated.

MedPix 2.0: A Comprehensive Multimodal Biomedical Dataset for Advanced AI Applications

The increasing interest in developing Artificial Intelligence applications in the medical domain, suffers from the lack of high-quality dataset, mainly due to privacy-related issues. Moreover, the recent rising of Multimodal Large Language Models (MLLM) leads to a need for multimodal medical datasets, where clinical reports and findings are attached to the corresponding CT or MR scans. This paper illustrates the entire workflow for building the data set MedPix 2.0. Starting from the well-known multimodal dataset MedPix\textsuperscript{\textregistered}, mainly used by physicians, nurses and healthcare students for Continuing Medical Education purposes, a semi-automatic pipeline was developed to extract visual and textual data followed by a manual curing procedure where noisy samples were removed, thus creating a MongoDB database. Along with the dataset, we developed a GUI aimed at navigating efficiently the MongoDB instance, and obtaining the raw data that can be easily used for training and/or fine-tuning MLLMs. To enforce this point, we also propose a CLIP-based model trained on MedPix 2.0 for scan classification tasks.

Unveiling Molecular Moieties through Hierarchical Graph Explainability

Background: Graph Neural Networks (GNN) have emerged in very recent years as a powerful tool for supporting in silico Virtual Screening. In this work we present a GNN which uses Graph Convolutional architectures to achieve very accurate multi-target screening. We also devised a hierarchical Explainable Artificial Intelligence (XAI) technique to catch information directly at atom, ring, and whole molecule level by leveraging the message passing mechanism. In this way, we find the most relevant moieties involved in bioactivity prediction. Results: We report a state-of-the-art GNN classifier on twenty Cyclin-dependent Kinase targets in support of VS. Our classifier outperforms previous SOTA approaches proposed by the authors. Moreover, a CDK1-only high-sensitivity version of the GNN has been designed to use our explainer in order to avoid the inherent bias of multi-class models. The hierarchical explainer has been validated by an expert chemist on 19 approved drugs on CDK1. Our explainer provided information in accordance to the docking analysis for 17 out of the 19 test drugs. Conclusion: Our approach is a valid support for shortening both the screening and the hit-to-lead phase. Detailed knowledge about the molecular substructures that play a role in the inhibitory action, can help the computational chemist to gain insights into the pharmacophoric function of the molecule also for repurposing purposes.

IODeep: an IOD for the introduction of deep learning in the DICOM standard

In recent years, Artificial Intelligence (AI) and in particular Deep Neural Networks (DNN) became a relevant research topic in biomedical image segmentation due to the availability of more and more data sets along with the establishment of well known competitions. Despite the popularity of DNN based segmentation on the research side, these techniques are almost unused in the daily clinical practice even if they could support effectively the physician during the diagnostic process. Apart from the issues related to the explainability of the predictions of a neural model, such systems are not integrated in the diagnostic workflow, and a standardization of their use is needed to achieve this goal. This paper presents \textit{IODeep} a new DICOM Information Object Definition (IOD) aimed at storing both the weights and the architecture of a DNN already trained on a particular image dataset that is labeled as regards the acquisition modality, the anatomical region, and the disease under investigation. The IOD architecture is presented along with a DNN selection algorithm from the PACS server based on the labels outlined above, and a simple PACS viewer purposely designed for demonstrating the effectiveness of the DICOM integration, while no modifications are required on the PACS server side. The source code are freely available at https://github.com/CHILab1/IODeep.git

Linear density-based clustering with a discrete density model

Density-based clustering techniques are used in a wide range of data mining applications. One of their most attractive features con- sists in not making use of prior knowledge of the number of clusters that a dataset contains along with their shape. In this paper we propose a new algorithm named Linear DBSCAN (Lin-DBSCAN), a simple approach to clustering inspired by the density model introduced with the well known algorithm DBSCAN. Designed to minimize the computational cost of density based clustering on geospatial data, Lin-DBSCAN features a linear time complexity that makes it suitable for real-time applications on low-resource devices. Lin-DBSCAN uses a discrete version of the density model of DBSCAN that takes ad- vantage of a grid-based scan and merge approach. The name of the algorithm stems exactly from its main features outlined above. The algorithm was tested with well known data sets. Experimental results prove the efficiency and the validity of this approach over DBSCAN in the context of spatial data clustering, enabling the use of a density-based clustering technique on large datasets with low computational cost.