This work aims to implement an automated data-driven model for breast cancer detection in mammograms to support physicians’ decision process within a breast cancer screening or detection program. The public available CBIS-DDSM and the INbreast datasets were used as sources to implement the transfer learning technique on full-field digital mammography proprietary dataset. The proprietary dataset reflects a real heterogeneous case study, consisting of 190 masses, 46 asymmetries, and 71 distortions. Several Yolo architectures were compared, including YoloV3, YoloV5, and YoloV5-Transformer. In addition, Eigen-CAM was implemented for model introspection and outputs explanation by highlighting all the suspicious regions of interest within the mammogram. The small YoloV5 model resulted in the best developed solution obtaining an mAP of 0.621 on proprietary dataset. The saliency maps computed via Eigen-CAM have proven capable solution reporting all regions of interest also on incorrect prediction scenarios. In particular, Eigen-CAM produces a substantial reduction in the incidence of false negatives, although accompanied by an increase in false positives. Despite the presence of hard-to-recognize anomalies such as asymmetries and distortions on the proprietary dataset, the trained model showed encouraging detection capabilities. The combination of Yolo predictions and the generated saliency maps represent two complementary outputs for the reduction of false negatives. Nevertheless, it is imperative to regard these outputs as qualitative tools that invariably necessitate clinical radiologic evaluation. In this view, the model represents a trusted predictive system to support cognitive and decision-making, encouraging its integration into real clinical practice.

An open challenge in translational bioinformatics is the analysis of sequenced metagenomes from various environmental samples. Of course, several studies demonstrated the 16S ribosomal RNA could be considered as a barcode for bacteria classification at the genus level, but till now it is hard to identify the correct composition of metagenomic data from RNA-seq short-read data. 16S short-read data are generated using two next generation sequencing technologies, i.e. whole genome shotgun (WGS) and amplicon (AMP); typically, the former is filtered to obtain short-reads belonging to a 16S shotgun (SG), whereas the latter take into account only some specific 16S hypervariable regions. The above mentioned two sequencing technologies, SG and AMP, are used alternatively, for this reason in this work we propose a deep learning approach for taxonomic classification of metagenomic data, that can be employed for both of them. To test the proposed pipeline, we simulated both SG and AMP short-reads, from 1000 16S full-length sequences. Then, we adopted a k-mer representation to map sequences as vectors into a numerical space. Finally, we trained two different deep learning architecture, i.e., convolutional neural network (CNN) and deep belief network (DBN), obtaining a trained model for each taxon. We tested our proposed methodology to find the best parameters configuration, and we compared our results against the classification performances provided by a reference classifier for bacteria identification, known as RDP classifier. We outperformed the RDP classifier at each taxonomic level with both architectures. For instance, at the genus level, both CNN and DBN reached 91.3% of accuracy with AMP short-reads, whereas RDP classifier obtained 83.8% with the same data. In this work, we proposed a 16S short-read sequences classification technique based on k-mer representation and deep learning architecture, in which each taxon (from phylum to genus) generates a classification model. Experimental results confirm the proposed pipeline as a valid approach for classifying bacteria sequences; for this reason, our approach could be integrated into the most common tools for metagenomic analysis. According to obtained results, it can be successfully used for classifying both SG and AMP data.

The SARS-CoV-2 virus pandemic had devastating effects on various aspects of life: clinical cases, ranging from mild to severe, can lead to lung failure and to death. Due to the high incidence, data-driven models can support physicians in patient management. The explainability and interpretability of machine-learning models are mandatory in clinical scenarios. In this work, clinical, laboratory and radiomic features were used to train machine-learning models for COVID-19 prognosis prediction. Using Explainable AI algorithms, a multi-level explainable method was proposed taking into account the developer and the involved stakeholder (physician, and patient) perspectives. A total of 1023 radiomic features were extracted from 1589 Chest X-Ray images (CXR), combined with 38 clinical/laboratory features. After the pre-processing and selection phases, 40 CXR radiomic features and 23 clinical/laboratory features were used to train Support Vector Machine and Random Forest classifiers exploring three feature selection strategies. The combination of both radiomic, and clinical/laboratory features enabled higher performance in the resulting models. The intelligibility of the used features allowed us to validate the models’ clinical findings. According to the medical literature, LDH, PaO2 and CRP were the most predictive laboratory features. Instead, ZoneEntropy and HighGrayLevelZoneEmphasis - indicative of the heterogeneity/uniformity of lung texture - were the most discriminating radiomic features. Our best predictive model, exploiting the Random Forest classifier and a signature composed of clinical, laboratory and radiomic features, achieved AUC=0.819, accuracy=0.733, specificity=0.705, and sensitivity=0.761 in the test set. The model, including a multi-level explainability, allows us to make strong clinical assumptions, confirmed by the literature insights.