=Paper=
{{Paper
|id=Vol-3762/560
|storemode=property
|title=AI in Medicine: Activities of the CINI-AIIS Lab at University of Naples Federico II
|pdfUrl=https://ceur-ws.org/Vol-3762/560.pdf
|volume=Vol-3762
|authors=Domenico Benfenati,Salvatore Capuozzo,Giovanni Maria De Filippis,Adriano De Simone,Michela Gravina,Lidia Marassi,Stefano Marrone,Elio Masciari,Enea Vincenzo Napolitano,Giuseppe Pontillo,Marco Postiglione,Cristiano Russo,Cristian Tommasino,Antonio Maria Rinaldi,Vincenzo Moscato,Carlo Sansone
|dblpUrl=https://dblp.org/rec/conf/ital-ia/BenfenatiCFSGM024
}}
==AI in Medicine: Activities of the CINI-AIIS Lab at University of Naples Federico II==
https://ceur-ws.org/Vol-3762/560.pdf
AI in Medicine: Activities of the CINI-AIIS Lab at University
of Naples Federico II
Domenico Benfenati1 , Salvatore Capuozzo1 , Giovanni Maria De Filippis1 , Adriano De Simone1 ,
Michela Gravina1 , Lidia Marassi1 , Stefano Marrone1 , Elio Masciari1 ,
Enea Vincenzo Napolitano1 , Giuseppe Pontillo1,4,5 , Marco Postiglione1,3 , Cristiano Russo1 ,
Cristian Tommasino1,2 , Antonio M. Rinaldi1 , Vincenzo Moscato1 and Carlo Sansone1,*
1
Department of Electrical Engineering and Information Technology (DIETI), University of Naples Federico II, Naples, Italy
2
Interdepartmental Center for Research on Management and Innovation in Healthcare (CIRMIS), University of Naples Federico II, Naples, Italy
3
Northwestern University, Department of Computer Science, McCormick School of Engineering and Applied Science, 2233 Tech Dr, Evanston, IL
60208, United States
4
Department of Radiology and Nuclear Medicine, MS Center Amsterdam, Amsterdam Neuroscience, Amsterdam UMC, Vrije Universiteit
Amsterdam, Amsterdam, The Netherlands
5
Department of Advanced Biomedical Sciences, University of Naples Federico II, Naples, Italy
Abstract
Artificial Intelligence (AI) encompasses a variety of methods and algorithms that have found applications across numerous
domains over time. The increasing complexity and abundance of data in healthcare have spurred investigations into utilizing
AI techniques within the medical field, leading to promising avenues for fostering innovation, facilitating early diagnosis, and
aiding in treatments. In this study, we provide a concise overview of several initiatives conducted within this realm at the
University of Naples Federico II node of the CINI-AIIS Lab, highlighting their primary objectives and contributions.
Keywords
Artificial Intelligence, Healthcare, Deep Learning, Machine Learning
1. Introduction intricate medical histories or multiple conditions.
AI applications encompass a variety of technologies
Artificial intelligence (AI) mimics human intelligence in rather than being confined to a single one. Among these,
machines to carry out tasks involving abstraction and Machine Learning (ML) stands out as a subset of AI, com-
problem-solving. Among the various sectors influenced prising algorithms that enable systems to autonomously
by AI, healthcare stands out as a highly promising field learn and enhance their performance through experi-
for its application. Indeed, the integration of AI in health- ence. In the realm of healthcare, traditional ML finds
care has the potential to aid both patients and healthcare widespread use, notably in precision medicine, where it
professionals, revolutionizing patient care and admin- predicts optimal treatment protocols based on patient
istrative operations. Additionally, AI-driven platforms attributes and contextual factors.
possess the capability to analyze patient data, highlight Deep Learning (DL) is a class of ML algorithms charac-
potential health issues, and enhance diagnostic preci- terized by the use of Artificial Neural Networks (ANNs)
sion for physicians, particularly in scenarios involving that simulate the structure of the human brain. DL ap-
proaches have gained popularity in pattern recognition
Ital-IA 2024: 4th National Conference on Artificial Intelligence, orga- tasks, particularly in image processing, improving medi-
nized by CINI, May 29-30, 2024, Naples, Italy
*
Corresponding author.
cal image analysis. Moreover, in recent years, there has
†
These authors contributed equally.
been a notable surge in interest surrounding the applica-
$ carlo.sansone@unina.it (C. Sansone) tion of Large Language Models (LLMs) within the medical
0009-0008-5825-8043 (D. Benfenati); 0000-0002-2578-9349 domain. LLMs are advanced natural language processing
(S. Capuozzo); 0009-0002-8395-0724 (G. M. D. Filippis); (NLP) models trained on massive amounts of text data,
0009-0001-3740-416X (A. D. Simone); 0000-0001-5033-9617 capable of understanding, generating, and processing
(M. Gravina); 0009-0006-8134-5466 (L. Marassi);
human language with remarkable accuracy and fluency.
0000-0001-6852-0377 (S. Marrone); 0000-0002-1778-5321
(E. Masciari); 0000-0002-6384-9891 (E. V. Napolitano); In this paper, we will illustrate some of the projects ex-
0000-0001-5425-1890 (G. Pontillo); 0000-0003-1470-8053 ploiting AI techniques in the medical field carried out at
(M. Postiglione); 0000-0002-8732-1733 (C. Russo); the University of Naples Federico II node of the CINI-AIIS
0000-0001-9763-8745 (C. Tommasino); 0000-0001-7003-4781 Lab, highlighting their innovative aspects and contribu-
(A. M. Rinaldi); 0000-0002-0754-7696 (V. Moscato);
tions.
0000-0002-8176-6950 (C. Sansone)
© 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License
Attribution 4.0 International (CC BY 4.0).
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
2. PIE: a novel paradigm for sideration of their initial perspectives, thus facilitating
the generation of revised reports. Ultimately, the internist
medical recommendations physician consolidates all reports to produce a unified
Existing works in the biomedical domain employing summary for the human specialist utilizing the system.
LLMs often overlook the phenomenon of LLM hallu-
cinations [1], which consists in the system generating
responses that are either factually incorrect, nonsensi-
3. AI in Patient Support:
cal, or disconnected from the input prompt. To address Opportunities and Risks
this concern, we introduce a novel paradigm, denoted
as Predict→Interpret→Explain (PIE). This paradigm en- The capacity of LLMs to interpret and generate natural
tails employing a model trained on high-quality and con- language with exceptional understanding and contextual
trolled data for predictions, interpreting its internal mech- awareness has paved the way for innovative approaches
anisms using state-of-the-art eXplainable Artificial Intel- to enhancing patient support. One possible application
ligence (XAI) techniques, and subsequently utilizing a of LLMs is represented by chatbots, virtual assistants
pool of LLM-based agents as reasoning engines to con- capable of creating a welcoming communication environ-
vey medical recommendations and the interpretation to ment for patients by providing both emotional support
medical professionals. Below, we outline each step in the and informative responses, in a way that is similar to
PIE pipeline, which we have experimented with in the a human operator. The natural conversational element
context of a medication recommendation downstream and the ability to understand and respond to patients’
task. needs can contribute to making patients feel comfortable
The Predict module exploits the wealth of informa- communicating with a virtual model, even when aware
tion embedded within Electronic Health Records (EHRs) of its non-human nature. The potential of a model to
derived from the MIMIC-III dataset [2]. We system- emulate authentic human interaction had already been
atically analyze both the structured and unstructured explored in the past, in the 1960s, when research in nat-
data present in these records to construct heterogeneous ural language processing led to the development of the
graphs, taking into account the semantic interrelations ELIZA system, capable of emulating a Rogerian thera-
among various medical concepts and harnessing intrinsic pist [5]. The Naples’ CINI AI-IS node had also pioneered
correlations within EHR data. Subsequently, a suite of this approach, designing and implementing a chatbot
Graph Neural Networks (GNNs) has been trained for a architecture intended to support patients in performing
link prediction task. This task entails discerning forth- pre-screening procedures [6, 7].
coming associations between patients and medications. With the advent of Large Language Models (LLMs),
The Interpret module is pivotal for elucidating the this emulation capability is coupled with a remarkable
decision-making process of the GNN model. It consists accuracy of responses. The fact that some studies are
in understanding the underlying rationales behind a de- beginning to show a preference for responses provided
cision. Interpretation serves as an indispensable tool by AI models like ChatGPT, in terms of accuracy, nuance,
bridging the intricate nature of AI models with the hu- and even empathy, opens up many interesting possibili-
man requisites for transparency and trustworthiness in ties for the future of medicine and healthcare. In a recent
decision-making processes. In our study, we evaluated study, some responses from ChatGPT in the medical field
the efficacy of the Integrated Gradients [3] and GNNEx- were evaluated as significantly higher quality compared
plainer [4] methodologies, both furnishing node weights to those of doctors and more empathetic [8]. This latter
that delineate their significance within the predictive data appears particularly interesting when considering
context. the potential integration of AI models into healthcare sys-
In the Explain module, the outcomes derived from the tems to improve doctors’ responses to patient inquiries
previous phases undergo refinement through a pool of and lighten their workload.
LLM-based collaborative agents, operating in accordance However, this progress also presents significant risks.
with a specific protocol. Initially, an internist physician Considering only the effect on patients, there is concern
enlists a panel of specialists (e.g., cardiologists, nephrolo- that the widespread use of AI to provide psychological
gists), tailored to the individual patient’s characteristics. support may contribute to exacerbating stigma towards
Each specialist evaluates the patient’s characteristics, the certain categories of patients, especially in the psychiatric
predictions from the GNN-based model, and the findings context, relegating them to "non-human" interactions.
of the Interpret phase to generate a comprehensive as- Additionally, there is a risk of eroding trust in traditional
sessment, elucidating the justification behind the model’s medicine and perceiving healthcare as impersonal, as
prediction. Then, specialists review the reports of their well as potentially reducing the responsibility of human
peers and engage in a discussion that may lead to recon- operators in expressing empathy and managing interac-
tions with patients, prompting various reflections on the
areas of overlap between chatbots and human operators. clinical applications, providing deep insights into patient-
If the new possibilities offered by AI in the medical field specific brain characteristics. Indeed, the network’s abil-
are manifold, it seems crucial to recognize its limitations ity to predict the subject’s age allows for the evaluation
and ensure that these technologies are used ethically and of brain-PAD as a biomarker for a range of neurological
under the supervision of qualified professionals. Technol- diseases, including Multiple Sclerosis, Alzheimer’s dis-
ogy, as a tool and given its current state of development, ease, and schizophrenia. Moreover, the decomposition of
should integrate rather than replace human interaction brain MRI into its main components of shape and appear-
in the medical context, ensuring that values such as em- ance enables the assessment of how a specific subject
pathy, clinical judgment, and professional ethics remain deviates from an age-specific standard template. When
at the centre of healthcare. While the combination of hu- applied to a population of patients with neurological dis-
man expertise and technological advancements promises eases, we expect that the deformation field will capture
to significantly improve healthcare, its success will de- disease-specific characteristics such as atrophy, while the
pend on the ability to effectively balance the skills of texture component may be useful for lesion detection.
both. Finally, the ability to disentangle age-related features
provides the architecture with generative capacity, allow-
ing for longitudinal evaluation obtained by varying the
4. ASAD Project age-specific latent representation
In neuroimaging, Deep Learning (DL) has been widely
used to model chronological age as a function of brain 5. Dementia Severity Assessment
Magnetic Resonance Imaging (MRI) scans in healthy in-
dividuals with excellent results [9]. The extent to which with Incomplete Multimodal
a person deviates from healthy brain-ageing trajecto- Data
ries, expressed as the difference between predicted and
chronological age (brain-predicted age difference, brain- Alzheimer’s disease (AD) is the most common cause of
PAD), has been proposed as an index of structural brain dementia, affecting millions of elderly people around the
health, sensitive to brain pathology in a wide spectrum world. AD is a neurodegenerative disorder, and early
of neurological disorders [10]. However, the presence of detection is a key element to improve the quality of life
subject-specific characteristics in different acquisitions of affected patients and their families. In clinical trials,
necessitates capturing factors of variation within the tar- Magnetic Resonance Imaging (MRI) and Positron Emis-
get population. In this project, our aim is to propose sion Tomography (PET) are mostly used for the early
an innovative DL-based model for age, shape, and ap- diagnosis of neurodegenerative disorders since they pro-
pearance disentanglement (ASAD) in brain MRI. This vide volumetric and metabolic function information of
model will enable a more precise quantification of the the brain, respectively.
impact of pathologies on the brain, significantly enhanc- There is the need of combining information from hetero-
ing the analysis of the brain-PAD. We will employ an geneous and complementary sources, such as MRI and
Autoencoder architecture consisting of an Encoder to PET, to evaluate the structural and metabolic character-
define three latent representations for age, shape, and istics of the brain. This makes the Multimodal Learning
appearance, respectively. Additionally, a Regressor will well suited in the case of dementia assessment. Tech-
be utilized to compute the predicted age, along with a set niques for multimodal data fusion can be categorized
of three decoders: Da, Ds, and Dt. Specifically, Da and into early, intermediate, and late fusion. Early fusion in-
Ds will extract the texture (appearance) and the deforma- tegrates multiple sources of data into a single features vec-
tion field (shape component), respectively, while Dt will tor, before being used as input to a learning model. Late
consider the age-specific latent representation to provide fusion, also referred as decision-level fusion, combines
an age-specific template. Following a similar approach as according to a given rule the decisions from multiple clas-
proposed in [11], the ASAD architecture will be trained sifiers, each trained on separate modalities. Intermediate
to reconstruct the input from the disentangled compo- fusion, also named as joint learning, exploits the deep
nents, using only data from healthy controls with varying neural networks to transform raw inputs into higher-
age ranges. This enables the network to model a healthy level and shared representations, which are constructed,
brain-aging trajectory. Subsequently, the architecture for instance, by merging into a single layer, units coming
will be applied to heterogeneous patient populations en- from multiple modality-specific paths. However, when
compassing a wide spectrum of neurological disorders. working with a multimodal dataset in the medical field, it
This application aims to detect disease-specific character- is not easy to have images of all the involved modalities,
istics within the disentangled components. The disentan- belonging to the same patient. For each subject, paired
glement of age, shape, and appearance may have several acquisition consists of images coming from all the differ-
ent sources and collected at the same time or in a specific
range. In a real scenario, patients may have incomplete
acquisitions, in which some modalities are missed.
In the work proposed in [12], we conducted a systematic
analysis of early, late and joint approaches in fusion for
dementia severity assessment on the publicly available
OASIS-3 dataset [13]. We focused on 3D Convolutional
Neural Network (CNN) to exploit the volumetric features
of the involved images, including in the training step
strategies to handle a high imbalance and incomplete
dataset. In particular, we analyzed the effects of the in- Figure 1: Use case: Data-centric approach for digital health
complete dataset in each multimodal fusion technique, transformation
and in the case of intermediate fusion, we proposed a
Multiple Input - Multi Output 3D CNN whose training
iteration changes according to the characteristics of the when making decisions, leading to better outcomes for
input volumes. To further assess the generalization abil- patients. This requires three fundamental steps: inte-
ity of the implemented methodology, we are including grate, open, innovate: use interoperability standards to
the ADNI dataset [14], a study consisting of about 2500 integrate existing systems and data. Storing data in an
subjects and focusing on the progression of mild cogni- open, vendor-neutral format will then enable ecosystems
tive impairment and early AD [15]. of vendors to innovate. Real use cases of data-centric
architectures for healthcare, such as the one proposed
6. Data-Centric AI for Healthcare at Karolinska University Hospital, have been already de-
veloped (see Figure 1). The adoption of standards like
In the age of digital transformation, healthcare is rapidly OpenEHR, FHIR, HL7, and ontologies like Snomed CT
evolving into a data-driven ecosystem. Imagine a health- represent the technical foundations upon which this vi-
care system where patient records are seamlessly inter- sion can be realized to achieve semantic and structural
connected, diagnosis is made with unprecedented pre- interoperability in personal health data, that is to ensure
cision, and treatments are tailored in real time. Data- high data quality.
centric architectures are the key to unlocking this vision-
ary healthcare landscape. The transition from a model-
centric approach to AI to a data-centric one signifies a
7. UNet-based multi-class nuclei
shift in emphasis when it comes to creating and imple- segmentation
menting AI systems. Model-centric AI aims at producing
the best model for a given dataset, whereas data-centric In recent years, the application of AI in Healthcare is
AI aims at systematically and algorithmically producing increasingly stimulating researchers interests [16, 17].
the best dataset to feed a given ML model. Nuclei panoptic segmentation, i.e., the simultaneous de-
Current challenges and limitations in health data gov- tection, segmentation, and classification of nuclear in-
ernance have demonstrated the need for a real digital stances, is at the core of the automation of several tasks in
transformation of healthcare, where decisions are made digital pathology, particularly in the analysis of routine
based on data, whether it is patient history, laboratory Hematoxylin and Eosin (H&E) stained histology slides.
results, or imaging data. AI algorithms can assist in iden- Distillation Framework. In our framework, we adopt
tifying high-risk patients, predicting treatment outcomes, an offline technique using HoVerNet as a pre-trained
and enabling personalized medicine. Nevertheless, ensur- teacher network. Given that HoVerNet performs nuclei
ing patient data security, AI algorithm accountability, and instance segmentation and classification through three
transparency are crucial to address privacy, security, and branches, our distillation strategy is based on the idea of
bias concerns. Emerging technologies, such as Extended combining all output branches of HoVerNet into a single
reality (XR) and blockchains, are being already used for branch network. Note that we aim to train a student that
improving patient care and guarantee the security and can replace only the HoVerNet backbone, not its post-
privacy of the data. In this context, the data-centric man- processing steps, which we left unvaried. We employ
ifesto serves as a beacon for the healthcare community. a single-branch UNet as our student model and join all
Collaboration among clinicians, healthcare organizations, HoVerNet branch outputs into a single branch with a
and technology vendors is indispensable in implement- number of output channels equal to the total number
ing a data-centric approach to coalesce around a common of HoVerNet’s branches. In particular, we used a Mix
vision, and to ensure that all relevant data is considered Vision Transformer (MixViT) as the backbone for UNet,
resulting in the best combination based on our experi- fined by the absence of identifiable biochemical or struc-
ments. Our loss is a linear combination of the student tural anomalies. These conditions affect almost 50% of
loss between the student and the ground truth and the infants in their first year of life [20]
distillation loss between the student and the teacher reg- This study examines the considerable impact of FGIDs
ulated by 𝛼 parameter. on children, their families and healthcare systems, and
In this work we used two datasets, namely PanNuke [18] highlights the historic challenge of identifying children at
for training HoVer-UNet and CoNSeP [19] for validating risk due to unclear pathophysiology. The research aims
results on external data. In the case of PanNuke dataset, to identify early-life risk factors for FGIDs [21], within
our solution achieved comparable performance to HoV- the first year of life. Using a prospective observational
erNet, demonstrating a significant advantage in terms of cohort design, the study enrolled term and preterm in-
processing speed. When the CoNSeP dataset is consid- fants from a tertiary care university hospital in Foggia,
ered, the results showed that our solution outperforms Italy, between 1 January 2020 and 31 December 2022,
HoVerNet in terms of panoptic quality, though it falls excluding infants with severe disease or major neonatal
short in terms of F-score detection. Regarding classifica- complications. By using conventional statistical methods
tion metrics, our solution outperforms HoVerNet across and Machine Learning (ML), this study identified birth
neoplastic and epithelial nuclei; it is practically equal for weight, cord blood pH, and maternal age as significant
miscellaneous and worse for inflammatory. Lastly, the predictors for FGIDs. A logistic regression predictive
inference time is about three times lower. model also established an inverse relationship between
Figure 2 shows visual examples of the results of HoV- these variables and the occurrence of FGIDs. Using these
erNet and HoVer-UNet compared with the CoNSeP refer- findings, the study created a ML-based predictive model
ence standard. Overall, the similarity between the results and a practical, user-friendly web interface for risk as-
supports the practical effectiveness of our approach. sessment. This enables clinicians to identify infants at a
higher risk for FGIDs. The approach marks a pioneering
Ground Truth HoVerNet Ours Ground Truth HoVerNet Ours step in FGID risk prediction.
Acknowledgments
This work was supported in part by the Piano Nazionale
Ripresa Resilienza (PNRR) Ministero dell’Università e
della Ricerca (MUR) Project under Grant PE0000013-FAIR
References
[1] M. A. Ahmad, I. Yaramis, T. D. Roy, Creating trust-
worthy llms: Dealing with hallucinations in health-
care AI, CoRR abs/2311.01463 (2023). URL: https://
Neoplastic Inflammatory Epithelial Miscellaneous
doi.org/10.48550/arXiv.2311.01463. doi:10.48550/
ARXIV.2311.01463. arXiv:2311.01463.
Figure 2: Nuclei segmentation and classification comparison [2] A. Johnson, T. Pollard, L. Shen, L. Lehman, M. Feng,
between CoNSeP ground truth, HoVerNet, and our predic- M. Ghassemi, B. Moody, P. Szolovits, L. Celi,
tions. R. Mark, MIMIC-III, a freely accessible critical care
database, Scientific Data 3 (2016).
[3] M. Sundararajan, A. Taly, Q. Yan, Ax-
iomatic Attribution for Deep Networks, CoRR
8. Infantile Predictors of abs/1703.01365 (2017). URL: http://arxiv.org/abs/
Functional Gastrointestinal 1703.01365. arXiv:1703.01365.
[4] Z. Ying, D. Bourgeois, J. You, M. Zitnik, J. Leskovec,
Disorders GNNExplainer: Generating Explanations for
Functional Gastrointestinal Disorders (FGIDs) are a sig- Graph Neural Networks, in: H. M. Wallach,
nificant challenge in pediatric healthcare due to their H. Larochelle, A. Beygelzimer, F. d’Alché-Buc,
prevalence and impact on infants. FGIDs refer to a range E. B. Fox, R. Garnett (Eds.), Advances in Neural
of conditions, including infant colic, regurgitation, func- Information Processing Systems 32: Annual
tional diarrhea, and functional constipation, that are de- Conference on Neural Information Processing
Systems 2019, NeurIPS 2019, December 8-14, 2019,
Vancouver, BC, Canada, 2019, pp. 9240–9251. URL: [15] A. De Simone, C. Sansone, A multimodal deep
https://proceedings.neurips.cc/paper/2019/hash/ learning based approach for alzheimer’s disease
d80b7040b773199015de6d3b4293c8ff-Abstract. diagnosis, in: International Conference on Image
html. Analysis and Processing, Springer, 2023, pp. 131–
[5] R. Bowman, O. Cooney, J. W. Newbold, A. Thieme, 139.
L. Clark, G. Doherty, B. Cowan, Exploring how [16] C. Tommasino, F. Merolla, C. Russo, S. Staibano,
politeness impacts the user experience of chatbots A. M. Rinaldi, Histopathological image deep feature
for mental health support, International Journal of representation for cbir in smart pacs, Journal of
Human-Computer Studies 184 (2024) 103181. Digital Imaging 36 (2023) 2194–2209.
[6] F. Amato, S. Marrone, V. Moscato, G. Piantadosi, [17] A. M. Rinaldi, C. Russo, C. Tommasino, Effects of
A. Picariello, C. Sansone, et al., Chatbots meet color stain normalization in histopathology image
ehealth: Automatizing healthcare., in: WAIAH@ retrieval using deep learning, in: 2022 IEEE Interna-
AI* IA, 2017, pp. 40–49. tional Symposium on Multimedia (ISM), IEEE, 2022,
[7] F. Amato, S. Marrone, V. Moscato, G. Piantadosi, pp. 26–33.
A. Picariello, C. Sansone, Holmes: Ehealth in the [18] J. Gamper, N. A. Koohbanani, K. Benes, S. Graham,
big data and deep learning era, Information 10 M. Jahanifar, S. A. Khurram, A. Azam, K. Hewitt,
(2019) 34. N. Rajpoot, Pannuke dataset extension, insights and
[8] J. W. Ayers, A. Poliak, M. Dredze, E. C. Leas, Z. Zhu, baselines, arXiv preprint arXiv:2003.10778 (2020).
J. B. Kelley, D. J. Faix, A. M. Goodman, C. A. [19] S. Graham, Q. D. Vu, S. E. A. Raza, A. Azam, Y. W.
Longhurst, M. Hogarth, et al., Comparing physi- Tsang, J. T. Kwak, N. Rajpoot, Hover-net: Simul-
cian and artificial intelligence chatbot responses to taneous segmentation and classification of nuclei
patient questions posted to a public social media in multi-tissue histology images, Medical image
forum, JAMA internal medicine 183 (2023) 589–596. analysis 58 (2019) 101563.
[9] M. Tanveer, M. Ganaie, I. Beheshti, T. Goel, N. Ah- [20] A. Chogle, C. A. Velasco-Benitez, I. J. Koppen, J. E.
mad, K.-T. Lai, K. Huang, Y.-D. Zhang, J. Del Ser, Moreno, C. R. R. Hernández, M. Saps, A population-
C.-T. Lin, Deep learning for brain age estimation: based study on the epidemiology of functional gas-
A systematic review, Information Fusion (2023). trointestinal disorders in young children, The Jour-
[10] J. H. Cole, R. P. Poudel, D. Tsagkrasoulis, M. W. nal of pediatrics 179 (2016) 139–143.
Caan, C. Steves, T. D. Spector, G. Montana, Predict- [21] D. Bi, H. Jiang, K. Yang, T. Guan, L. Hou, G. Shu,
ing brain age with deep learning from raw imaging Neonatal risk factors for functional gastrointestinal
data results in a reliable and heritable biomarker, disorders in preterm infants in the first year of life
NeuroImage 163 (2017) 115–124. (2022).
[11] Z. Shu, M. Sahasrabudhe, R. A. Guler, D. Samaras,
N. Paragios, I. Kokkinos, Deforming autoencoders:
Unsupervised disentangling of shape and appear-
ance, in: Proceedings of the European conference
on computer vision (ECCV), 2018, pp. 650–665.
[12] M. Gravina, A. García-Pedrero, C. Gonzalo-Martín,
C. Sansone, P. Soda, Multi input–multi output 3d
cnn for dementia severity assessment with incom-
plete multimodal data, Artificial Intelligence in
Medicine 149 (2024) 102774.
[13] P. J. LaMontagne, T. L. Benzinger, J. C. Morris,
S. Keefe, R. Hornbeck, C. Xiong, E. Grant, J. Has-
senstab, K. Moulder, A. G. Vlassenko, et al., Oasis-
3: longitudinal neuroimaging, clinical, and cogni-
tive dataset for normal aging and alzheimer disease,
MedRxiv (2019).
[14] C. R. Jack Jr, M. A. Bernstein, N. C. Fox, P. Thomp-
son, G. Alexander, D. Harvey, B. Borowski, P. J. Brit-
son, J. L. Whitwell, C. Ward, et al., The alzheimer’s
disease neuroimaging initiative (adni): Mri meth-
ods, Journal of Magnetic Resonance Imaging: An
Official Journal of the International Society for Mag-
netic Resonance in Medicine 27 (2008) 685–691.