=Paper=
{{Paper
|id=Vol-3762/573
|storemode=property
|title=From Covid-19 detection to cancer grading: how medical-AI is boosting clinical diagnostics and may improve treatment
|pdfUrl=https://ceur-ws.org/Vol-3762/573.pdf
|volume=Vol-3762
|authors=Andrea Berti,Rossana Buongiorno,Gianluca Carloni,Claudia Caudai,Francesco Conti,Giulio Del Corso,Danila Germanese,Davide Moroni,Eva Pachetti,Maria Antonietta Pascali,Sara Colantonio
|dblpUrl=https://dblp.org/rec/conf/ital-ia/BertiBCC0CGMPPC24
}}
==From Covid-19 detection to cancer grading: how medical-AI is boosting clinical diagnostics and may improve treatment==
https://ceur-ws.org/Vol-3762/573.pdf
From Covid-19 detection to cancer grading: how medical-AI
is boosting clinical diagnostics and may improve treatment
Andrea Berti1,2 , Rossana Buongiorno1,2 , Gianluca Carloni1,2 , Claudia Caudai1 ,
Francesco Conti1,3 , Giulio Del Corso1 , Danila Germanese1 , Davide Moroni1 , Eva Pachetti1,2 ,
Maria Antonietta Pascali1 and Sara Colantonio1,*,†
1
Institute of Information Science and Technologies, ISTI, National Research Council of Italy, CNR, via G. Moruzzi, 1, Pisa, 56124, Italy
2
Department of Information Engineering, University of Pisa, Via Caruso 16,56122, Pisa, Italy
3
Department of Mathematics, University of Pisa, Largo B. Pontecorvo, 56126, Pisa, Italy
Abstract
The integration of artificial intelligence (AI) into medical imaging has guided an era of transformation in healthcare. This
paper presents the research activities that a multidisciplinary research group within the Signals and Images Lab of the
Institute of Information Science and Technologies of the National Research Council of Italy is carrying out to explore the great
potential of AI in medical imaging. From the convolutional neural network-based segmentation of Covid-19 lung patterns to
the radiomic signature for benign/malignant breast nodule discrimination, to the automatic grading of prostate cancer, this
work highlights the paradigm shift that AI has brought to medical imaging, revolutionizing diagnosis and patient care.
Keywords
Visual intelligence, Medical imaging, Radiomics, Convolutional Neural Networks, Deep Neural Networks, Trustworthy AI
1. Introduction efficiency, increasing diagnostic accuracy, and fostering
greater patient satisfaction. However, it is important to
Medical imaging modalities such as computed tomogra- strike a delicate balance between promoting the benefits
phy (CT), magnetic resonance imaging (MRI), positron of AI in clinical practice, which are evident, and address-
emission tomography (PET), and ultrasound (US) play ing concerns about the transparency, trustworthiness,
a key role in providing healthcare professionals with and potential bias of AI algorithms.
detailed and exhaustive visual data of the human body. This paper summarises the ongoing activities of a mul-
These imaging techniques generate significant amounts tidisciplinary research group within the Signals and Im-
of data that require efficient analysis and interpretation. ages Lab of the Institute of Information Science and Tech-
This is where Artificial Intelligence (AI) comes in. nologies of the National Research Council of Italy. The
AI may emulate human cognitive processes in analyz- group aims to explore the potential applications of AI in
ing and understanding healthcare data. By focusing on promoting and supporting health and well-being, while
the analysis of biomedical images using computational also addressing the challenges related to algorithms’ ex-
techniques such as object detection, segmentation and plainability and transparency.
registration, AI has the potential to enhance diagnos-
tic and prognostic accuracy by identifying patterns and
correlations that may be difficult for humans to observe 2. AI for clinical diagnosis
[1].
In the past, the use of AI in medicine was constrained In the following, we provide a brief overview of the re-
by technological limitations until 1998, when the US search conducted in the field of AI supporting clinical
Food and Drug Administration (FDA) approved the first diagnostics. The primary focus is on medical imaging,
computer-aided detection (CAD) system for mammogra- given that radiology is expected to benefit most from
phy [2]. Since then, there has been exponential growth recent advancements in AI.
in the use of AI techniques in the medical field.
Today, hospitals are actively exploring AI solutions 2.1. AI for Fatty Liver Content Estimation
to support operational efforts aimed at improving cost from US Imaging
Hepatic steatosis, characterized by the accumulation of
Ital-IA 2023: 3rd National Conference on Artificial Intelligence, orga-
fat within the liver, when coupled with inflammation,
nized by CINI, May 29–31, 2023, Pisa, Italy
*
Corresponding author. can contribute to the advancement of fibrosis towards
$ sara.colantonio@isti.cnr.it (S. Colantonio) cirrhosis and hepatocellular carcinoma [3]. Therefore,
© 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License
Attribution 4.0 International (CC BY 4.0). early detection and quantification of steatosis (via fat
CEUR
Workshop
Proceedings
http://ceur-ws.org
ISSN 1613-0073
CEUR Workshop Proceedings (CEUR-WS.org)
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
fraction assessment) are crucial tasks for predicting the Clinical Management of COVID-19 Patients", funded by
disease progression. Magnetic Resonance Spectroscopy Tuscany Region).
is the gold standard for the fat fraction assessment, while We conducted a comparison between them to ascertain
US imaging is commonly used to identify liver steatosis insights into the cognitive mechanisms that can drive
during screenings. Despite being non-invasive, US is a neural model towards optimal performance for this
highly operator-dependent [4]. task, as well as to identify the optimal balance between
In collaboration with a team from the IFC-CNR and the volume of data, time, and computational resources
Pisa University Hospital, we conducted a systematic com- necessary. From the results of the analysis, it can be
parison between three Deep Learning (DL) models in concluded that Attention-UNet outperforms the other
estimating, from US images, the fat fraction [5]. The models by achieving the best performance of 81.93%, in
compared models were the following: (i) a determinis- terms of 2D Dice score on the test set.
tic Convolutional Neural Network (CNN), similar to the
one in [6], (ii) an MC Dropout CNN model, and (iii) a 2.3. AI for Alzheimer disease detection
Bayesian CNN with probabilistic output.
In comparison to [6], the multi-center dataset in- On top of the work [13], the cerebrospinal fluid of 21 sub-
creased up to 186 subjects. jects who received a clinical diagnosis of Alzheimer’s dis-
Regression results showed good prediction perfor- ease (AD) as well as of 22 pathological controls has been
mance for all architectures on the 5-fold test sets (Nor- collected and analysed by Raman Spectroscopy (RS). The
malized RMSE 5.87%, 5.35%, and 5.82% for deterministic, aim of this research is to understand if the Raman spectra
MC Dropout, and Bayesian CNN, respectively). How- could be used to distinguish AD from controls, after a pre-
ever, the introduction of uncertainty quantification (UQ), processing procedure. We applied machine learning to a
contributes to decreasing the percentage of mispredicted set of topological descriptors extracted from the spectra,
cases (from 32.4% for classical CNN to less than 9% for achieving a high classification accuracy of 86% (the best
Bayesian one). Furthermore, the possibility of having performing combination is the Ridge classifier applied
access to information about the confidence with which to the persistence landscapes vectorization).Our experi-
the network produces its outputs is a great advantage, mentation indicates that RS and topological analysis may
especially from the point of view of physicians who want be effective to confirm or disprove a clinical diagnosis of
to use neural networks as computer-aided diagnosis. Alzheimer’s disease. Also, it opens the way to possibly
increasing and/or confirming the knowledge about the
precise molecular events and biological pathways behind
2.2. AI for Covid-19 Pulmonary Patterns
the Alzheimer’s disease, e.g., by identifying the bands of
Identification the Raman spectrum relevant for AD detection.
During the Coronavirus Disease 2019 (COVID-19) pan-
demic, High-Resolution Computed Tomography (HRCT) 2.4. AI for the Diagnosis of Eosinophilic
of the chest has been adopted as a method to visually iden- Esophagitis
tify two distinct abnormal pulmonary patterns: Ground
Glass Opacity (GGO), characterized by increased attenu- Eosinophilic esophagitis (EoE) is a chronic disease charac-
ation and hazy density in lung lobes, and Consolidation, terized by esophageal symptoms and eosinophilic inflam-
indicated by bilateral areas of lung tissue filled with fluid mation of the esophagus. Among patients with dyspha-
instead of air [7]. However, these patterns appear scat- gia, EoE and non-EoE patients should receive different
tered with undefined contours and often lack contrast therapies and therefore must be timely and correctly iden-
with surrounding healthy tissue. tified from the clinical history or by using more invasive
Consequently, the segmentation and quantification of procedures (endoscopic and/or histological information).
pathological lung regions from HRCT data have proven In [14], an RDF-based ML model was trained on a multi-
to be very challenging. center international database (273 EoE and 55 non-EoE
In [8], we compared four state-of-the-art CNNs based dysphagia patients clinical and endoscopic data, collected
on the encoder-decoder paradigm for the binary seg- from Guy’s and St. Thomas’ Hospital NHS Foundation
mentation of COVID-19 infections (UNet [9], Attention- Trust (GSTT, London, United Kingdom), Pisa Univ. Hos-
UNet [10], Recurrent–Residual UNet (R2-UNet) [11], R2- pital (Pisa, Italy), and Padua Univ. Hospital (Padua, Italy))
Attention UNet [12]), after training and testing them on to provide indications for the investigation of EoE in
90 HRCT volumetric scans of COVID-19 patients. The adults reporting dysphagia or to inform point-of-care
images were collected from the database of the Pisa Uni- decision-making for performing esophageal biopsies in
versity Hospital (in the framework of the regional project adults with dysphagia.
"Optimised - An Optimised Path for the Data Flow and The model was further evaluated on an independent co-
hort of 93 consecutive patients with dysphagia, result-
ing in an AUC of 0.90 (using clinical data) and an AUC able procedure. Multi-parametric Magnetic Resonance
of 0.94 (using a combination of clinical and endoscopic Imaging (mpMRI) is frequently employed to get an initial
data) The model, re-trained on the whole dataset, has assessment of the tumor. To this end, numerous studies
been integrated into an open-access online tool (https: have explored ML/DL models for automatic PCa grading
//webapplicationing.shinyapps.io/PointOfCare-EoE/). from mpMRI images [17].
However, developing accurate and generalizable DL
models for medical imaging, where data is often scarce,
3. AI for cancer grading presents a significant challenge. Few-shot learning (FSL)
offers a promising solution, particularly since the ad-
AI algorithms are showing potential in improving the
vancements in meta-learning [18]. For this reason, we
current protocol for grading various cancers, such as
investigated FSL techniques for assessing PCa aggres-
breast and prostate cancer. In the following sections, we
siveness from mpMRI images. We proposed a two-step
provide a brief description of our research in this area.
approach: a disentangled self-supervised learning (SSL)
pre-training step for robust feature extraction, followed
3.1. AI for the discrimination between by meta-fine-tuning utilizing finer-grained classes and
benign/malignant breast nodules in the coarser-grained ones in meta-testing for enhanced
ABVS and DBT images generalization [19]. Our approach achieved a mean AU-
ROC of 0.821 for a 4-way (ISUP 2-5) 5-shot setting. We
Although imaging techniques are commonly used for further explored enhancing FSL models performance by
breast cancer screening, biopsy is the only method avail- leveraging synthetic image generation, employing a De-
able to categorize a breast lesion as benign or malignant. noising Diffusion Probabilistic Model (DDPM).
However, biopsies are invasive and costly procedures Also, we proposed a new technique to discover and
that can cause discomfort in patients [15]. exploit causality signals from images via neural networks
Radiomic analysis of biomedical images shows promise for classification purposes [20, 21]. We model how the
in addressing various clinical challenges, such as early presence of a feature in one part of the image affects
detection and classification of breast tumors. the appearance of others in different parts of the image.
In the P.I.N.K study [16], 66 women were enrolled. Our method consists of a convolutional backbone and
Their paired Automated Breast Volume Scanner (ABVS) a causality-factors extractor computing weights to en-
and Digital Breast Tomosynthesis (DBT) images, an- hance feature maps according to their causal influence
notated with cancerous lesions, populated the first in the scene. We evaluated our method on a dataset of
ABVS+DBT dataset. This allowed for radiomic analy- prostate MRI images for cancer diagnosis and studied its
sis to differentiate between malignant and benign breast effectiveness of our module both in fully-supervised and
cancer. 1-shot learning. On the binary classification of cancer
Three Machine Learning (ML) methods were em- versus no-tumor cases, our method led to a maximum
ployed: Random Decision Forests (RDF), Support Vector test accuracy of 0.72, representing a 5 % increase to the
Machines (SVM), and Logistic Regression (Logit). They baseline [21]. On distinguishing ISUP grades in 1-shot
were trained and validated using an ad hoc nested LOO learning, we obtained a 0.71 AUROC for the classification
cross-validation procedure to ensure a minimally biased ISUP 2 vs. all the others, with 13 % increasing to the base-
estimation of the model’s generalization ability, even line [20]. Our attention-inspired improved the overall
with a limited sample size. The study’s main finding classification and produced more robust XAI predictions
highlights the superior effectiveness of RDF model in focusing on relevant parts of the image.
accurately predicting tumor classification using radiomic
features in both ABVS and DBT acquisitions. It achieved
AUC-ROC values of 89.9% with a subset of 19 features.
3.3. AI for chondrosarcoma grading from
Additionally, promising outcomes were achieved using Raman Spectroscopy
solely textural radiomic features to train RDF model, withRaman Spectroscopy (RS) allows for the observation of
AUC-ROC values of 71.8% and 74.1% for ABVS and DBT, changes in biochemical constituents (such as proteins,
respectively. This suggests the potential for integrating lipid structures, DNA, and vitamins) among different
virtual biopsy into routine medical practice. tissues by obtaining their biochemical maps. Recently,
RS has been applied to chondrogenic tumor classification
3.2. AI for prostate cancer grading from with excellent results [22].
MRI acquisitions Chondrogenic tumors are the second largest group of
bone tumors worldwide. They are generally classified as
Current methods for determining Prostate cancer (PCa) primary chondrosarcomas when they occur in previously
aggressiveness rely on biopsy, an invasive and uncomfort- normal bone. Secondary chondrosarcomas result from
tracted from 134 T2-weighted Magnetic Resonance Imag-
ing (MRI) images of patients who underwent radiother-
apy. The MRI scans were obtained from ProstateNet
(https://prostatenet.eu), the repository designed within
the framework of the EU H2020 ProCAncer-I project.
Data regarding the presence and severity of rectal and
urinary side effects after treatment were also included.
The results demonstrated that radiomics-based ap-
proaches can be effective in predicting radiotherapy-
induced side effects, achieving an AUROC of 70.8%. Also,
a set of simplified model variants was used to estimate
epistemic uncertainty and provide a reliability score to
Figure 1: Representative histologic images of the tumours complement the main model’s prediction.
analyzed in this study (hematoxylin and eosin staining). EC
(Panel a); CS G1 (Panel b); CS G2 (Panel c); CS G3 (Panel d),
from [23]. 5. AI for the newborn and infant
5.1. Thermal imaging for stress and
the malignant transformation of a benign cartilaginous well-being
lesion and are classified into three grades: CS G1, CS G2 In this field, we investigated also the use of thermal imag-
and CS G3. Enchondroma (EC) is a non-cancerous tumor. ing for stress discrimination [27, 28], to the aim of detect-
Distinguishing between EC and CS G1 is a critical issue ing stress in adults under stress stimuli, and of assessing
for pathologists, as it generates many false positive and the efficacy of the hortotherapy for female adolescence af-
false negative diagnoses [24]. fected by anorexia nervosa. Notably, we are moving to a
In [23] we showed that the combination of persistent more challenging task: deepen the understanding of ther-
homology and ML techniques can support the classifica- mal profiles in the newborn (possibly pre-term) in order
tion of Raman spectra extracted from cancerous tissues to develop or improve new treatment techniques related
to achieve a reliable chondrosarcomas grading. to the maturation of the newborn thermo-regulation sys-
A total of 410 Raman spectra from 10 patients with tem. A study protocol, joint work with the lab NINA
primary chondrogenic tumors of the skeleton, treated and the NICU of Santa Chiara Hospital in Pisa, is under
at Azienda Ospedaliera Universitaria Pisana (Pisa), were review.
used to train the machine learning models. Despite the
small size of the experimental dataset, the results show
that the method not only achieved high accuracy on 5.2. AI for baby facial gestures
previously unseen data samples; also such a methos can recognition
be easily integrated into a Raman spectroscopic system as One open issue related to children’s research concerns
an automatic tool to assist clinicians in grading tumors. neonatal imitation (NI), namely the primitive ability of
infants to mirror the actions of others [29]. The question
4. AI for predicting of whether imitation is present from birth is of great
importance as it can foster a deeper understanding of how
radiotherapy-induced toxicity in it contributes to later developmental outcomes, which is
prostate cancer crucial for the preterm newborn.
Computer vision methods may unobtrusively detect
Radiotherapy is a commonly used treatment for prostate and analyze the most relevant facial features, thus provid-
cancer (PCa). In recent years, there has been a surge of in- ing clinicians (or parents, caregivers, etc.) with objective
terest in leveraging ML methods to analyze radiomic fea- data about children’s health status [30]. However, for in-
tures derived from multiparametric MRI (mpMRI) scans fants, this is a challenging task, due to significant changes
of PCa. However, little attention has been given to pre- in their facial morphology compared to adults, and to
dicting radiation-induced toxicity [25] before starting the increased complexity in data collection caused by
radiotherapy. In the work carried out in the framework unpredictable variations in their facial poses [31].
of the EU H2020 ProCAncer-I project [26], we aimed to In [32], we analyzed videos of 10 newborns (8 preterms,
predict radiotherapy-induced side effects, including both 2 at term, ≤ 4 weeks post-term equivalent age), perform-
genito-urinary and rectal toxicity. ing tasks such as tongue protrusion and mouth opening,
A RDF model was trained on radiomic features ex- to classify open/closed mouths. The videos were analyzed
research agenda, J. of ambient intell. and human-
ized comp. 14 (2023) 8459–8486.
[2] J. E. Goldberg, B. Reig, A. A. Lewin, Y. Gao, L. Hea-
cock, S. L. Heller, L. Moy, New horizons: artificial
intelligence for digital breast tomosynthesis, Radio-
Graphics 43 (2022) e220060.
[3] A. Han, M. Byra, E. Heba, M. P. Andre, J. W. Erd-
man Jr, R. Loomba, C. B. Sirlin, W. D. O’Brien Jr,
Figure 2: Data preparation procedure: the original image Noninvasive diagnosis of nonalcoholic fatty liver
is processed using Face Landmarker of Google MediaPipe disease and quantification of liver fat with radiofre-
Solutions to identify a rough contour of the mouth (a). This quency ultrasound data using one-dimensional con-
imprecise contour is used to crop/reorient the image. An volutional neural networks, Radiology 295 (2020)
adaptive brightness/contrast enhancement is applied to the 342–350.
final image (b). [4] M. Mancini, A. Prinster, G. Annuzzi, R. Liuzzi,
R. Giacco, C. Medagli, M. Cremone, G. Clemente,
S. Maurea, G. Riccardi, et al., Sonographic hepatic-
at frame-level, for a total of 41000 labeled frames. In each renal ratio as indicator of hepatic steatosis: com-
frame, we identified mouth landmarks and cropped the parison with 1h magnetic resonance spectroscopy,
images around the mouth, then we applied an image pre- Metabolism 58 (2009) 1724–1730.
processing pipeline (which included mouth orientation, [5] G. Del Corso, M. A. Pascali, C. Caudai, L. De Rosa,
resizing, brightness and contrast enhancement, see Fig- A. Salvati, M. Mancini, L. Ghiadoni, F. Bonino, M. R.
ure 2) to improve classification performance. A CNN was Brunetto, S. Colantonio, F. Faita, Ann uncertainty
trained using a ten-fold cross-validation, which resulted estimates in assessing fatty liver content from ul-
in highly reliable results: accuracy, precision, and recall trasound data, Submitted (2024).
over 92% on unseen data. [6] S. Colantonio, A. Salvati, C. Caudai, F. Bonino,
L. D. Rosa, M. A. Pascali, D. Germanese, M. R.
Brunetto, F. Faita, A deep learning approach for
6. Conclusions hepatic steatosis estimation from ultrasound imag-
ing, in: K. Wojtkiewicz, J. Treur, E. Pimenidis,
AI has a big potential to improve care and health sys- M. Maleszka (Eds.), Adv. in Comp. Collective Intel-
tems, specially for diagnostic tasks, even if facing very ligence, Springer International Publishing, Cham,
important technical issues like unbalance dataset, data 2021, pp. 703–714.
drift, heterogeneous acquisition protocols, and input data [7] T. Ai, Z. Yang, H. Hou, C. Zhan, C. Chen, W. Lv,
and annotations of variable quality. Also, future research Q. Tao, Z. Sun, L. Xia, Correlation of chest ct and
should involve healthcare professionals and caregivers as rt-pcr testing for coronavirus disease 2019 (covid-
designers and users, comply with health-related regula- 19) in china: A report of 1014 cases, Radiology 296
tions, improve transparency and privacy, integrate with (2020) E32–E40.
healthcare technological infrastructure, explain their de- [8] R. Buongiorno, G. Del Corso, D. Germanese, L. Col-
cisions to the users, and establish evaluation metrics and ligiani, L. Python, C. Romei, S. Colantonio, Enhanc-
design guidelines. ing covid-19 ct image segmentation: A comparative
study of attention and recurrence in unet models,
Acknowledgments J. of Imaging 9 (2023) 283.
[9] O. Ronneberger, P. Fischer, T. Brox, U-net: Convo-
This publication is based upon the work carried out lutional networks for biomedical image segmenta-
within the COST Action GoodBrother (CA19121), the tion, in: N. Navab, J. Hornegger, W. M. Wells, A. F.
EU H2020 projects ProCAncer-I (GA 952159) and FAITH Frangi (Eds.), Med. Image Comp. and Computer-
(GA 101135932), the PAR FAS Tuscany Region NAVIGA- Assisted Intervention – MICCAI 2015, Springer In-
TOR, PRAMA and OPTIMISED. ternational Publishing, 2015.
[10] O. Oktay, J. Schlemper, L. L. Folgoc, M. Lee, M. Hein-
rich, K. Misawa, K. Mori, S. McDonagh, N. Y. Ham-
References merla, B. Kainz, B. Glocker, D. Rueckert, Attention
u-net: Learning where to look for the pancreas,
[1] Y. Kumar, A. Koul, R. Singla, M. F. Ijaz, Artificial 2018.
intelligence in disease diagnosis: a systematic liter- [11] M. Z. Alom, M. Hasan, C. Yakopcic, T. M. Taha,
ature review, synthesizing framework and future V. K. Asari, Recurrent residual convolutional neural
network based on u-net (r2u-net) for medical im- tribution of raman spectroscopy to diagnosis and
age segmentation, arXiv preprint arXiv:1802.06955 grading of chondrogenic tumors, Scientific Reports
(2018). 10 (2020) 2155.
[12] Q. Zuo, S. Chen, Z. Wang, R2au-net: Attention [23] F. Conti, M. D’Acunto, C. Caudai, S. Colantonio,
recurrent residual convolutional neural network for R. Gaeta, D. Moroni, M. A. Pascali, Raman spec-
multimodal medical image segmentation, Security troscopy and topological machine learning for can-
and Comm. Networks 2021 (2021) 1–10. cer grading, Scientific reports 13 (2023) 7282.
[13] F. Conti, M. Banchelli, V. Bessi, C. Cecchi, F. Chiti, [24] C. D. Savci-Heijink, A. H. Cleven, J. V. Bovée,
S. Colantonio, C. D’Andrea, M. de Angelis, D. Mo- Benign and low-grade cartilaginous tumors: An
roni, B. Nacmias, et al., Alzheimer disease detection update on differential diagnosis, Diagnostic
from raman spectroscopy of the cerebrospinal fluid Histopathology 28 (2022) 501–509.
via topological machine learning, Eng. Proc. 51 [25] H. Abdollahi, S. R. Mahdavi, B. Mofid, M. Bakhshan-
(2023) 14. deh, A. Razzaghdoust, A. Saadipoor, K. Tanha, Rec-
[14] P. Visaggi, G. Del Corso, F. B. Svizzero, M. Ghisa, tal wall mri radiomics in prostate cancer patients:
S. Bardelli, A. Venturini, D. S. Donati, B. Barberio, prediction of and correlation with early rectal toxi-
E. Marciano, M. Bellini, et al., Artificial intelligence city, Int. j.l of rad. biology 94 (2018) 829–837.
tools for the diagnosis of eosinophilic esophagitis in [26] G. Del Corso, E. Pachetti, R. Buongiorno, A. C. Ro-
adults reporting dysphagia: development, external drigues, D. Germanese, M. A. Pascali, J. Almeida,
validation, and software creation for point-of-care N. Rodrigues, M. Tsiknakis, N. Papanikolaou,
use, The J. of Allergy and Clinical Immunology: In D. Regge, K. Marias, P.-I. Consortium, S. Colan-
Practice (2023). tonio, Radiomis-based reliable predictions of side
[15] J. M. Hemmer, J. C. Kelder, H. P. van Heesewijk, effects after radiotherapy for prostate cancer, in:
Stereotactic large-core needle breast biopsy: anal- Accepted to ISBI2024- the 21st Int. Symp. on Biom.
ysis of pain and discomfort related to the biopsy Imaging, 2024.
procedure, European rad. 18 (2008) 351–354. [27] F. Gioia, M. A. Pascali, A. Greco, S. Colantonio, E. P.
[16] G. Del Corso, D. Germanese, C. Caudai, G. Anastasi, Scilingo, Discriminating stress from cognitive load
P. Belli, A. Formica, A. Nicolucci, S. Palma, M. A. using contactless thermal imaging devices, in: 2021
Pascali, S. Pieroni, et al., Adaptive machine learning 43rd Ann. Int. Conf. of the IEEE Eng. in Med. and
approach for importance evaluation of multimodal Biology Soc. (EMBC), 2021, pp. 608–611.
breast cancer radiomic features, J. of Imaging Inf. [28] O. Curzio, L. Billeci, V. Belmonti, S. Colantonio,
in Med. (2024) 1–10. L. Cotrozzi, C. F. De Pasquale, M. A. Morales, C. Nali,
[17] M. He, Y. Cao, C. Chi, X. Yang, R. Ramin, S. Wang, M. A. Pascali, F. Venturi, A. Tonacci, N. Zannoni,
G. Yang, O. Mukhtorov, L. Zhang, A. Kazantsev, S. Maestro, Horticultural therapy may reduce psy-
et al., Research progress on deep learning in mag- chological and physiological stress in adolescents
netic resonance imaging based diagnosis and treat- with anorexia nervosa: A pilot study, Nutrients 14
ment of prostate cancer: a review on the current (2022).
status and perspectives, Frontiers in Oncology 13 [29] A. N. Meltzoff, M. K. Moore, Imitation of facial and
(2023) 1189370. manual gestures by human neonates, Science 198
[18] Y. Wang, Q. Yao, J. T. Kwok, L. M. Ni, Generaliz- (1977) 75–78.
ing from a few examples: A survey on few-shot [30] D. Germanese, S. Colantonio, M. Del Coco,
learning, ACM computing surveys (csur) 53 (2020) P. Carcagnì, M. Leo, Computer vision tasks for
1–34. ambient intelligence in children’s health, Informa-
[19] E. Pachetti, S. A. Tsaftaris, S. Colantonio, tion 14 (2023).
Boosting few-shot learning with disentangled [31] D. Kuefner, V. Macchi Cassia, M. Picozzi, E. Bricolo,
self-supervised learning and meta-learning for Do all kids look alike? evidence for an other-age
medical image classification, arXiv preprint effect in adults., J. of Exp. Psychology: Human
arXiv:2403.17530 (2024). Perception and Performance 34 (2008) 811.
[20] G. Carloni, E. Pachetti, S. Colantonio, Causality- [32] G. Del Corso, D. Germanese, M. A. Pascali,
driven one-shot learning for prostate cancer grad- S. Bardelli, A. Cuttano, F. Festante, A. Guzzetta,
ing from mri, in: Proc. of the IEEE/CVF Int. Conf. L. Rocchitelli, S. Colantonio, Facial landmark iden-
on Computer Vision, 2023, pp. 2616–2624. tification and data preparation can significantly im-
[21] G. Carloni, S. Colantonio, Exploiting causality sig- prove the extraction of newborns’ facial features,
nals in medical images: A pilot study with empirical in: Submitted, 2023.
results, Expert Sys. with Appl. (2024) 123433.
[22] M. D’Acunto, R. Gaeta, R. Capanna, A. Franchi, Con-