Towards the non-destructive analysis of multilayered samples: A novel XRF-VNIR-SWIR hyperspectral imaging system combined with multiblock data processing.

The new challenge in the investigation of cultural heritage is the possibility to obtain stratigraphical information about the distribution of the different organic and inorganic components without sampling. In this paper recently commercialized analytical set-up, which is able to co-register VNIR, SWIR, and XRF spectral data simultaneously, is exploited in combination with an innovative multivariate and multiblock high-throughput data processing for the analysis of multilayered paintings. The instrument allows to obtain elemental and molecular information from superficial to subsurface layers across the investigated area. The chemometric strategy proved to be highly efficient in data reduction and for the extraction and integration of the most useful information coming from the three different spectroscopies, also filling the gap between data acquisition and data understanding through the combination of principal component analysis (PCA), brushing, correlation diagrams and maps (within and between spectral blocks) on the low-level fused. In particular, correlation diagrams and maps provide useful information for the reconstruction of a stratigraphic structure without the need to take any sample, thanks to the effective account for inter-correlation among data (variables), which is able to effectively characterize the possible combinations of components located in the same depth level. The highly innovative technology and the data processing strategy are applied for the multi-level characterization of a complex painting reproduction as an illustrative pilot study.

Hybrid transformer UNet for thyroid segmentation from ultrasound scans.

Deep learning based medical image segmentation methods have been widely used for thyroid gland segmentation from ultrasound images, which is of great importance for the diagnosis of thyroid disease since it can provide various valuable sonography features. However, existing thyroid gland segmentation models suffer from: (1) low-level features that are significant in depicting thyroid boundaries are gradually lost during the feature encoding process, (2) contextual features reflecting the changes of difference between thyroid and other anatomies in the ultrasound diagnosis process are either omitted by 2D convolutions or weakly represented by 3D convolutions due to high redundancy. In this work, we propose a novel hybrid transformer UNet (H-TUNet) to segment thyroid glands in ultrasound sequences, which consists of two parts: (1) a 2D Transformer UNet is proposed by utilizing a designed multi-scale cross-attention transformer (MSCAT) module on every skipped connection of the UNet, so that the low-level features from different encoding layers are integrated and refined according to the high-level features in the decoding scheme, leading to better representation of differences between anatomies in one ultrasound frame; (2) a 3D Transformer UNet is proposed by applying a 3D self-attention transformer (SAT) module to the very bottom layer of 3D UNet, so that the contextual features representing visual differences between regions and consistencies within regions could be strengthened from successive frames in the video. The learning process of the H-TUNet is formulated as a unified end-to-end network, so the intra-frame feature extraction and inter-frame feature aggregation can be learned and optimized jointly. The proposed method was evaluated on Thyroid Segmentation in Ultrasonography Dataset (TSUD) and TG3k Dataset. Experimental results have demonstrated that our method outperformed other state-of-the-art methods with respect to the certain benchmarks for thyroid gland segmentation.