How big is wearable tech, just in running research? Enough for 372 references in a literature review... But, is it too data or just noise? We are quite clearly not at the point of getting NEW actionable data in an easy, consistent and reliable way...

Google tuned a AI chatbot to answer "consumer medical questions", and doctors judged 92.6% of its answers right … compared to 92.9% of answers given by other doctors.
In particular, they show that comprehension, recall of knowledge, and medical reasoning improve with model scale and instruction prompt tuning, suggesting a bleak vision for the future: more burden on the patients to learn how to ask questions so to receive useful answers, and growingly more responsibility washing through AI in the healthcare system.

Una riforma che deve essere accompagnata anche dalla possibilità di una maggiore circolazione dei dati sanitari oggi stoppata dall’Autorità del garante della Privacy in base a norme di 20 anni fa che non possono più considerarsi in linea con l’Ue (qui il documento). Per dire: oggi degli scienziati se vogliono studiare l’appropriatezza e la sicurezza a lungo termine nella pratica clinica di un farmaco su categorie di pazienti che si sono curati 10 anni fa e che hanno dato il consenso non solo al trattamento terapeutico ma anche all’uso dei propri dati, devono fare ogni sforzo possibile per ottenere per ogni singolo nuovo studio l’autorizzazione specifica del singolo partecipante e da parte dell’Autorità Garante con una trafila infinita che fa perdere la voglia a chiunque. E se il Garante non si esprime entro 45 giorni vale la regola del silenzio-dissenso. È urgente trovare un equilibrio tra la necessità di garantire la privacy e quella di avere risposte rapide indispensabili per la Salute pubblica.

The review finds that #DigitalHealth tools are not accessible to everyone equally and that the access to and level of engagement with digital health technologies in the European Region tend to be disproportionally lower among: 🔴Residents in rural areas 🔴Ethnic minorities and non-English speakers 🔴Individuals with a lower level of education or economic status 🔴Older adults Digital progress and innovation have the potential to revolutionize healthcare, but inequalities in access must be identified and addressed to ensure #HealthForAll. 

Check out MUSYC, a free online tool to quantify drug synergy: we played with it for the past few weeks and it looks very interesting

Heavy advertising and other strategies from Silicon Valley’s playbook boost providers’ growth but not the quality of care

We are represented, thanks to our host SCImPULSE Foundation, at the assembly of CoARA (Coalition for Advancing Research Assessment)!

Please, do share with us ideas and feedbacks about how to improve research assessment according to your experience: we want to make this ecosystem more fair and welcoming, and more effective of course! Get in touch ;)

Su quali fasi della presa in carico dei pazienti con patologie croniche reumatologiche e dermatologiche è maggiormente auspicabile l'adozione di pratiche della medicina digitale? 

Per quale "profilo" di pazienti?

Quali requisiti devono rispettare queste soluzioni? 

Con un board scientifico di esperti del settore, clinici specialisti, MMG, farmacisti, infermieri e associazioni dei pazienti il Centro per la Ricerca Economica in Sanità ha provato a dare delle risposte a questi quesiti, provando a sintetizzare i diversi punti di vista...

I risultati verranno presentati il 2 Dicembre alle 10.30 

Regulators encourage stakeholders to have a patient focus but actionable steps for early and continuous engagement are not well defined. Without intentionally pursuing patient-focused measurement, clinicians, researchers, and digital product vendors risk entrenching digital versions of poor traditional assessments and proliferating low-value tools that are ineffective, burdensome, and reduce quality and efficiency in clinical research and care.
To evaluate the meaningfulness of signals derived from digital sensors, the following four-level framework is useful:

1. Meaningful Aspect of Health (MAH)

2. Concept of Interest (COI)

3. Outcome to be measured 

4. Endpoint (exclusive to research)

In order for you to deploy this framework effectively, you must follow levels 1->4 in partnership with patients at every step.

In November 2022, FDA published a new report, Successes and Opportunities in Modeling & Simulation for FDA. The report:

    Elucidates how and where M&S is used across FDA, and the type and purpose of M&S used

    Presents a selection of M&S case studies from across nearly all FDA centers, which demonstrate how M&S is playing a tangible role in FDA fulfilling its mission

    Identifies opportunities for FDA to better harness M&S in upcoming years by embracing computational advances and new (and big) data streams to develop improved public health solutions

[cit. Eugenio Santoro]
Secondo il NEJM Group (New England Journal of Medicine) in USA le prestazioni di telemedicina dovrebbero essere rimborsate solo se

-Riducono i costi dell'assistenza

-Migliorano gli esisti clinici dei pazienti

-Garantiscono una maggiore equità

Occorre + ricerca clinica per rispondere a queste domande, che spesso non viene condotta.

E in Italia? La scelta di prescrivere e rimborsare una prestazione di telemedicina non è basata su alcuna evidenza scientifica. Sarebbe il caso di promuovere questo genere di ricerca.

The modern abundance and prominence of data have led to the development of “data science” as a new field of enquiry, along with a body of epistemological reflections upon its foundations, methods, and consequences. This article provides a systematic analysis and critical review of significant open problems and debates in the epistemology of data science.

„Independent living and quality of life cannot be defined by bodily care alone. Individual user characteristics, the needs and resources of persons, should be considered when implementing AI-based AAL. One strategy could be to diversify the training data of the algorithms involved. […]. Another approach is to actively engage stakeholders and community in the design process […]. The key is user participation in the design process and feedback loops after the technology has been implanted. This approach allows to tailor AAL-technologies to the specific needs of a certain group or context of use. In addition, age scripts that are used for designing and/or have been implemented have to be critically reflected upon in order to prevent standardization of user preferences.

In G.U. n. 256 del 2 novembre 2022 è pubblicato il Decreto 21 settembre 2022 del Ministero della salute: Approvazione delle linee guida per i servizi di telemedicina - Requisiti funzionali e livelli di servizio. 

Scientists regularly use models, which help make their thinking definite, specific, and easier to challenge, test and refine. Scientists know that models always have shortcomings. There’s no such thing as “modelling the real thing,” only various approximations thereof, which may be more or less useful.

Firmato l’accordo Iss-Enpam per formazione su telemedicina e sanità digitale: La piattaforma Tech2Doc realizzata da Enpam sarà utilizzata dall’Istituto Superiore di Sanità per fare formazione sui temi della telemedicina e la sanità digitale.

Regulatory pathway for the use of in silico trials - A team of researchers of the In Silico World project, led by Prof. Francesco Pappalardo (University of Catania), investigated the current state of the art towards the regulatory system for in silico trials applied to medical devices while exploring the European regulatory system toward this topic. The results of this research have been published in the article “Toward a regulatory pathway for the use of in silico trials in the CE marking of medical devices” where the authors suggest that the European regulatory system should establish a clear pathway for the use of in silico methodologies in the development, derisking, and regulatory assessment of new medical devices

“Electronic Medical Records Are Strangling American Medicine”. Might also have something to do with the complicated billing practices that drive EHR design

Do read this critical perspective about where much of our infrastructure and services, including healthcare, seems to be directed... and then read (or re-read) the 2007 piece we shared on the side about open source in Machine Learning (thinking about how this should hold for healthIT!)

"The Need for Open Source Software in Machine Learning" by Sonnenburg et al.  - (amazing foresight - published in *2007*! By a group of authors that not by chance included for seniors of the FAIR movement)

La procedura riguarda l’affidamento di un contratto avente ad oggetto la progettazione di dettaglio, la realizzazione, la messa in esercizio e la gestione di una piattaforma di AI.

La piattaforma dovrà essere una infrastruttura altamente innovativa per abilitare servizi a supporto dei professionisti sanitari delle Case di Comunità, migliorando la capacità di presa in carico proattiva e lo stato di salute della popolazione, utilizzando processi assistenziali innovativi. Un ulteriore obiettivo è il miglioramento del governo della domanda, anche per finalità di programmazione.

AI in Medicine is seriously overhyped and claims of efficacy are poorly checked. The scientific community could start working by introducing some requirements to gatekeep publication of wild claims: (1) Lockbox sets that the ML team does not see until publication time. (2) Bigger datasets. (3) if dataset not big enough, remove unnecessary degrees of freedom (e.g. by using autoML)

The Children's Brain Tumor Network (CBTN) - Accelerating Research in Pediatric Central Nervous System Tumors through Collaboration and Open Science - The CBTN's 32 member institutions utilize a shared regulatory governance architecture at the Children's Hospital of Philadelphia to accelerate and maximize the use of biospecimens and data. As of August 2022, CBTN has enrolled over 4,700 subjects, over 1,500 parents, and collected over 65,000 biospecimen aliquots for research. Additionally, over 80 preclinical models have been developed from collected tumors. Multi-omic data for over 1,000 tumors and germline material is currently available with data generation for > 5,000 samples underway. To our knowledge, CBTN provides the largest open-access pediatric brain tumor multi-omic dataset annotated with longitudinal clinical and outcome data, imaging, associated biospecimens, child-parent genomic pedigrees, and in vivo and in vitro preclinical models.

Investing in Health for All - Between acute crises like the COVID-19 pandemic and longer-running problems like the failure to manage preventable infectious and chronic diseases, global public health has suffered from shortcomings in both the public and private sectors, and in governance at all levels. While there are no one-size-fits-all solutions, investment and policy leadership are key to setting the right priorities and then delivering on them. What will it take to build more resilient health systems, drive financing and access to new drugs, and foster cross-border and -sectoral collaboration toward shared health goals?

Great policy document from the FDA for software functions that relate to devices.

While providers tried to enable the hospital at home during the pandemic, they struggled to automate the collection of physiological readings from  commoditized personal health devices such as BP monitors, BG monitors, weight scales, thermometers or pulse oximeters. 

The software functions of the manufacturers stood in the way and created an interoperability nightmare because they operate in silos designed for self-monitoring and generate massive privacy, security and inclusion issues for health IT systems in virtual care.

The authors of this paper describe a system, which they call "Holistic AI in Medicine" (HAIM) providing a generalized framework to use multiple sources of inputs to make decisions... from the patient record, to notes and even direct observations in front of the patient... Would this be progress, or would this accentuate the limits of AI in healthcare (black-box, biais, overfitting, etc)?

 Uncertainty quantification and mathematical modeling is a very useful tool for clinical decision making. This piece covered various sources for uncertain data, and uncertainty quantification (UQ) as a means to handle them in a mathematical correct and also clinically relevant way... It closes by recommending a few simple points to make uncertain quantification as precise and clinically relevant as possible.

Conferimento del Dottorato di ricerca honoris causa in “Clinical and Experimental Medicine (CEM) Medicina clinica e sperimentale” al Prof. Gregory Yoke Hong Lip

INTELLIGENZA ARTIFICIALE E RESPONSABILITA’ CIVILE PUBBLICATA IL 28 SETTEMBRE LA PROPOSTA DI DIRETTIVA

Multicenter, Randomized Trial of a Bionic Pancreas in Type 1 Diabetes - In this 13-week, randomized trial involving adults and children with type 1 diabetes, use of a bionic pancreas was associated with a greater reduction than standard care in the glycated hemoglobin level

Not-so-smart electronics... or when design only focuses on first order context, but sells "big solutions" 

The attached document provides an overview of the core pieces of legislation and ethical principles that are relevant to the In Silico project, and thus, in silico trials.

The analysis finds the following areas of legislation are of relevance:

- Privacy and Data Protection

- Data Governance

- Clinical Trials

- Medical Devices and Medicinal Products

- Artificial Intelligence (AI)

In addition to the above pieces of legislation, the study outlines the framework for forthcoming Ethics guidelines. Based on a principlist approach, these guidelines propose the analysis of four main Biomedical Ethics principles – i.e. autonomy, justice, beneficence, non-maleficence – and the secondary principle of responsibility.

I decreti legislativi 137 e 138 del 5 agosto 2022 pubblicati nella Gazzetta Ufficiale del 13 settembre 2022 hanno lo scopo di assicurare l’adeguamento, da parte di tutti i soggetti coinvolti (organismi notificati, operatori economici e operatori sanitari), alle disposizioni dettate dai Regolamenti (UE) 2017/745 e 2017/746.  

“Physician-scientists have epitomized the blending of deep, rigorous impactful curiosity with broad attention to human health for centuries.” - This Review Article takes a global view of the aspiring physician-scientist... but where are we with the education of telemedicine and digital care experts?!

In silico medicine - a call to action and partnership

HTA, SANITA' DIGITALE, RIMBORSABILITA', NUOVI DECRETI MD E IVD

Il Garante privacy boccia lo schema di decreto per l’Ecosistema dei Dati Sanitari del nuovo FSE - Il parere negativo del Garante privacy è stato rilasciato al Ministero della salute e al Ministero per l’innovazione tecnologica e la transizione digitale sullo schema di decreto che prevede la realizzazione della nuova banca dati denominata Ecosistema Dati Sanitari (EDS), prevista dalla riforma del Fascicolo sanitario elettronico. Il Garante ha chiesto anche di correggere un secondo schema di decreto, per favorire e migliorare l’implementazione a livello nazionale del Fascicolo sanitario elettronico (FSE).

Reproducibility of real-world evidence studies using clinical practice data to inform regulatory and coverage decisions - In 150 RealWorldEvidence studies we showed that the resulting effect sizes were strongly correlated between the original publications and study reproductions in the same data sources and developed insights on where there was room for improvement

March, 10-11 2023 - Bologna (Italy)

Smartphone-based screening for atrial fibrillation: a pragmatic randomized clinical trial - A digital, site-less, smartphone app, randomized clinical trial detected more than 2-fold atrial fibrillation warranting blood thinners compared to usual care

Here the authors present FedDis to collaboratively train an unsupervised deep convolutional autoencoder on 1,532 healthy magnetic resonance scans from four different institutions, and evaluate its performance in identifying pathologies such as multiple sclerosis, vascular lesions, and low- and high-grade tumours/glioblastoma on a total of 538 volumes from six different institutions

Patient flow is hampered by factors including woefully outdated IT systems, whose impacts are both direct — on the speed of moving patients through the system — and indirect — keeping doctors from treating patients 

Il merito di questo articolo è di estendere il quadro tradizionale dei determinanti di salute alle cure digitali in modo da garantirne l’equità nell’accesso e nell’efficacia. Spesso l’approccio che si utilizza è di escludere da queste cure i pazienti che non possiedono i requisiti necessari piuttosto che cercare di ridurli al minimo così da ampliare la platea dei soggetti che possono usufruirne. A partire dalla progettazione degli strumenti digitali (telemedicina, terapie digitali, dispositivi indossabili, etc..) e dai criteri poi di acquisto degli stessi è necessario assumere un atteggiamento pro-attivo e inclusivo in modo da ridurre al minimo il digital gap per le fasce di pazienti meno avanzati.

L’intelligence artificielle dans les soins : entre craintes et opportunités 

Le domaine des soins est directement impacté par l’essor inexorable de l’intelligence artificielle (IA) et voit ses pratiques se transformer. Au-delà des évolutions technologiques et médicales qu’elle permet, l’IA a de nombreuses répercussions sur l’humain (personnel soignant, patients, développeurs,…) qui doit intégrer et s’approprier ces changements.

Manuscript submission is now open for a new article collection we are co-editing called “Data Driven and Model Based Computational Futures in Cardiovascular Practice” with Frontiers in Cardiovascular Medicine (Impact Factor 5.846, CiteScore 8.2) - and we wanted you to be among the first to know.

To date, most applications of Artificial Intelligence (AI) to healthcare have been applied to clinical questions about diseases and patient care. Now that many hospitals have electronic health records (EHRs), there is potential to use AI for operational purposes. Patient flow in hospitals is difficult to plan for, because hospitals are highly connected systems in which capacity constraints in one area (for example, lack of ward beds) block the flow of patients from other locations. 

Don't be fooled by the hype based largely on easy confirmatory tests: it is way to easy to produce counterfactual examples highlighting the actual limitations of current machine learning, even the largest linguistic models touted to "reason like humans"

Charter of Trieste - Accessibility to services for all persons.

A follow up to the ESOF2020 launch meeting.

Fascicolo sanitario elettronico. I medici contro le nuove linee guida. Fnomceo: “L’analisi del contesto è errata, lontana dalla realtà professionale”

Il 22 Giugno 2022 si presenteranno le linee guida del progetto "Digital Medicine e Strumenti terapeutici innovativi per un cambio di paradigma nell’assistenza sanitaria digitale", progetto della Fondazione Lilly con il supporto di LS CUBE Studio Legale e di un Comitato Scientifico composto da prestigiosi esperti del settore.

Chronic diseases are the leading driver in US healthcare expenditure, making up approximately half of US healthcare spent or 10% of GDP. These conditions are generally ongoing and cause functional restriction, but can oftentimes be prevented. Managing a chronic illness is a lifelong and complex task that includes medication regimes, regular doctor visits, and surgical procedures. This burden of chronic diseases often shatters productive lives with fragmented patient journeys and expensive medical bills. However, we know that this burden can be lessened with early detection, lifestyle changes, medication management, and therapy. Digital health has great potential to deliver these integrated services and greatly improve chronic care management. 

Any single digital health tool is unlikely to be useful when implemented as a one-size-fits-all solution. See how UNICEF and partners are making sure digital health tools are the right fit, one assessment at a time.

The ‘PhenotypicFeature’ is the central element of the Phenopacket schema. A ‘PhenotypicFeature’ can be used to describe any phenotypic characteristic, including signs and symptoms, laboratory findings, histopathology findings, and imaging and electrophysiological results, along with modifier and qualifier concepts. Each phenotypic feature is described using an ontology term.

“DiMe and its partners will create an open access, interactive, regulatory science support tool that is based on existing FDA and other agency guidance documents”.

In recent years, the use of artificial intelligence (AI) in medicine and healthcare has been praised for the great promise it offers, but has also been at the centre of heated controversy. This study offers an overview of how AI can benefit future healthcare, in particular increasing the efficiency of clinicians, improving medical diagnosis and treatment, and optimising the allocation of human and technical resources. The report identifies and clarifies the main clinical, social and ethical risks posed by AI in healthcare, more specifically: potential errors and patient harm; risk of bias and increased health inequalities; lack of transparency and trust; and vulnerability to hacking and data privacy breaches. The study proposes mitigation measures and policy options to minimise these risks and maximise the benefits of medical AI, including multi-stakeholder engagement through the AI production lifetime, increased transparency and traceability, in-depth clinical validation of AI tools, and AI training and education for both clinicians and citizens.

1st Symposium on Geriatronics: Regulation through AI and Robotics, on June 20-21, 2022, in GarmischPartenkirchen, organized by Munich Institute of Robotics and Machine Intelligence (MIRMI) at the Technical University of Munich and Robotics & AI Law Society - RAILS, powered by VolkswagenStiftung

Registration is still possible at https://www.eventbrite.com/e/geriatronics-summit-tickets-346812053187

Tarnanas et al. about their research on Parkinson’s Disease. “...further progress in this field hinges on the more effective application of digital biomarker and artificial intelligence...”

We like a bold claim but we hope technical validation (the one presented in this paper) is always clearly separated from clinical validation (absent, yet again) ...we don't want to keep hearing trite arguments about how technology is different and has other ways to be regulated and evolve... CLINICAL VALIDATION TO DOCUMENT BENEFITS AND RISK SHOULD BE MANDATORY TO JUSTIFY ANY HEALTH CLAIM FOR TECH TOO!

The EIC Overseas Trade Fairs (OTF) Programme 2.0 carries on its travel to the first European destination and invites you to submit your application for MEDICA 2022. The spotlight event in the medical and healthcare sector will be held on 14-17 November 2022 in Düsseldorf, Germany. 

PerMedCoE: HPC/Exascale Centre of Excellence on Personalised Medicine - A demonstration of large scale, modular/composable agent based models to the study of covid19

The Dutch Ministry of Health, Welfare and Sport (VWS) commissioned Hooghiemstra & Partners and Axon Lawyers to analyze the AI Act and any potential overlap between the act and the already-in-effect Medical Device Regulation (MDR) and the soon-to-be in effect In-vitro Diagnostic Medical Device Regulation (IVDR). Since the commissioning of this report, there have been changes proposed to the act and these have been made available to the public in a compromise text. The main changes in the compromise text have been in regards to refinement of the definitions and further clarification of the scope of the act (with regard to MDR and IVDR). Negotiations with other member states are still ongoing and feedback from the healthcare sector will continue to be included in these talks.

The report identifies and clarifies the main clinical, social and ethical risks posed by AI in healthcare, more specifically: potential errors and patient harm; risk of bias and increased health inequalities; lack of transparency and trust; and vulnerability to hacking and data privacy breaches. The study proposes mitigation measures and policy options to minimise these risks and maximise the benefits of medical AI, including multi-stakeholder engagement through the AI production lifetime, increased transparency and traceability, in-depth clinical validation of AI tools, and AI training and education for both clinicians and citizens.

Check out the 2022 edition of ‘AI Watch. National strategies on Artificial Intelligence: A European perspective’. This report compares national AI strategies and assesses how they contribute to achieve the goals of the 2021 review of the Coordinated Plan on AI

Improving current practices is not just a matter of "doing the obvious" - Among a large population with AF with a guideline indication for OACs for stroke prevention who were randomized to a mailed educational intervention or to usual care, there was no clinically meaningful, numerical, or statistically significant difference in rates of OAC initiation. More-intensive interventions are needed to try and address the public health issue of underuse of anticoagulation for stroke prevention among patients with AF.

“Nine intervention hospitals implemented an Epic sepsis prediction tool, education, and standardized order sets (six control hospitals did not).

The intervention to facilitate timely sepsis care did not improve patient outcomes.”

More bad news for precision medicine. It's not trivial to improve clinical reality.

Since its establishment in 2010, the adverse outcome pathway (AOP) concept has become a prominent tool for the risk assessment community. AOPs are a chain of biological processes, called Key Events (KEs), starting from a molecular perturbation with a stressor toward an Adverse Outcome (AO), connected by Key Event Relationships (KERs). AOPs exist to capture all mechanistic toxicological knowledge from literature and data, to direct future studies to fill gaps of existing knowledge, and to drive Integrated Approaches to Testing and Assessment (IATA) development

Fascicolo sanitario elettronico e telemedicina al servizio delle malattie rare - il tema della telemedicina alla luce del PNRR, che prevede una vera e propria rivoluzione nel sistema della sanità territoriale. “In parte il Piano prevede la creazione di edifici, che andranno popolati con personale qualificato, ma nell’ottica di erogare prestazioni a domicilio è importante concentrarci su cosa c’è nelle freccette: non sono persone che si spostano, ma telecomunicazioni – ha affermato Gabbrielli -. Questo comporta un impatto notevole sulla revisione dei processi di lavoro. Il punto più difficile dei prossimi tre o quattro anni, infatti, non saranno le tecnologie digitali e nemmeno l’evoluzione della mentalità delle persone, perché l’emergenza Covid ci ha insegnato come questa può svilupparsi rapidamente laddove le cose funzionano; ma la revisione dei processi di lavoro nella nostra sanità digitale”.

Notwithstanding ongoing confusion over the concepts of “real-world data” and “real-world evidence” in 2022, more than 5 years after the passage of the 21st Century Cures Act, the FDA continues to evaluate such data and evidence as it considers regulatory decisions. - Interview with Kushal Kadakia on the 21st Century Cures Act, Cures 2.0, and lessons from Covid-19 for medical product development.

A quick look shows that shares of 8 of the largest telehealth companies are ALL down more than 75% in 12 months; for several of them their current market cap is LESS than the amount of cash they have in the bank - how long before we'll see some of these being acquired or taken private?

Quantumacy project just made a breakthrough in quantum key distribution and post quantum encryption for data protection by design..! Their QKD enabled sensitive data sharing protocol (GDPM) will be demonstrated live for general public coming 23rd of June! Stay tuned…

Proposta redatta dal Consiglio Superiore di Sanità per la razionalizzazione e interoperabilità dei molteplici sistemi informativi sulla salute che costituiscono i Sistemi Informativi Sanitari. Viene affrontato anche l’argomento più delicato e cioè il tema del Trattamento dei dati personali sotto il profilo giuridico, rifacendosi alla principale fonte ovvero il Regolamento Comunitario (UE) 2016/679 relativo alla protezione delle persone fisiche con riguardo al trattamento dei dati personali, nonché alla libera circolazione di tali dati.

The European Commission launched the European Health Data Space (EHDS), one of the central building blocks of a strong European Health Union. The EHDS will help the EU to achieve a quantum leap forward in the way healthcare is provided to people across Europe. It will empower people to control and utilise their health data in their home country or in other Member States. It fosters a genuine single market for digital health services and products

Telehealth encounters for chronic conditions had similar rates of follow-up to in-person encounters for these conditions, whereas telehealth encounters for acute conditions seemed to be more likely than in-person encounters to require follow-up.

I sistemi di intelligenza artificiale come strumento di supporto alla diagnostica - Nel documento sono elencate una serie di proposte operative finalizzate sia all’introduzione in modo sicuro dei software di AI nella pratica clinica sia alla implementazione di infrastrutture e di modalità di governance che possano rendere il nostro Paese competitivo a livello internazionale nella programmazione e nello sviluppo dei sistemi di AI.

A cura di Consiglio Superiore di Sanità - Sezione V

Oxipit announced the first the first "regulatory approved" (they have been granted a CE mark) AI medical imaging application to perform diagnostics autonomously.

What is a CE mark for medical devices anyway? They do not share scientific publications proving their claims prospectively in clinical trials, nor any clinically relevant data/reports for us to evaluate...

Some really interesting findings about Health Wearables and Apps in this article.

- A 2020 study called the Digital Health Generation, found that health tracking devices and apps can cause anxiety in young people.

- A study published in the Journal of the American Medical Informatics Association in September 2020 suggested that smartwatch readings could be prompting people to seek medical care when they might not need it. Only 11.4% of 264 patients in the study who received an abnormal pulse alert from their Apple Watch received a "clinically actionable cardiovascular diagnosis of interest." 

Analisi dell’infrastruttura che sarà realizzata attraverso Public-Private Partnership. - Per consentire l’interoperabilità con i diversi sistemi in uso si prevede l’utilizzo, sia a livello nazionale che a livello regionale, di API Gateway basati su standard HL7 FHIR per la gestione e lo scambio di dati e documenti strutturati attraverso l’esposizione di API REST, con framework di sicurezza OAuth 2.0 e l’utilizzo di un canale criptato.

Il termine per la presentazione delle proposte è fissato entro le 12.00 del giorno 18.05.2022 - Avviso per manifestazione di interesse per la presentazione di proposte di Partnership Pubblico Privato per l’affidamento in concessione per la “Progettazione, realizzazione e gestione dei Servizi abilitanti della Piattaforma nazionale di Telemedicina PNRR" - Missione 6 Componente 1 sub-investimento 1.2.3 "Telemedicina”

Although the role of artificial intelligence (AI) in medicine is increasingly studied, most patients do not benefit because the majority of AI models remain in the testing and prototyping environment. The development and implementation trajectory of clinical AI models are complex and a structured overview is missing. The authors therefore propose a step-by-step overview to enhance clinicians’ understanding and to promote quality of medical AI research

Sapere cogliere le opportunità che provengono dai trial clinici decentralizzati è importante dal momento che i tradizionali metodi per l’acquisizione di evidenze scientifiche, anche se tutt’ora consolidati e certamente necessari, mostrano alcuni limiti nella loro possibilità di generalizzazione dei risultati, in particolare considerando la variabilità del mondo reale che oggi possiamo studiare non solo a livello tissutale e cellulare ma anche molecolare e genomico così come a livello sociale e ambientale.

GatorTron: A Large Clinical Language Model to Unlock Patient Information from Unstructured Electronic Health Records. - There are limited studies exploring large language models in the clinical domain; the current largest clinical NLP model was trained with 110 million parameters (compared with 175 billion parameters in the general domain). It is not clear how large-size NLP models can help machines understand patients' clinical information from unstructured EHRs. In this study, we developed a large clinical transformer model - GatorTron - using >90 billion words of text and evaluated it on 5 clinical NLP tasks including clinical concept extraction, relation extraction, semantic textual similarity, natural language inference, and medical question answering. 

A retraction often marks the end of a dispute over published scientific research. But in the case of a $300 digital fertility tracker marketed by Valley Electronics, it was only the beginning.

After a study on the effectiveness of its Daysy thermometer was retracted, Valley sued Chelsea Polis, a researcher who had publicly lambasted its findings, for defamation. When a federal judge threw out the case, Valley appealed, arguing that Polis went too far by calling the company “unethical” and labeling its study “junk science.” The appeal — set for March 22 — underscores the growing tensions between new, loosely regulated technologies and watchdogs insisting that published claims be supported by rigorous science.

La minaccia che una malattia X infetti la popolazione umana e si diffonda rapidamente in tutto il mondo è più grande che mai, e lasciare che la pandemia da covid19 sia una ferita nelle nostre memorie e non un’occasione per accelerare e reindirizzare la crescita delle nostre strategie e strumenti di gestione della sanità pubblica, come invece ha fatto per lo R&I biotecnologico, sarebbe una grave colpa verso il nostro futuro.

If not for the war in Europe, the story below might’ve been one of world’s biggest today. The largest-ever study comparing those who got COVID-19 and recovered and those who never got infected reveals major lasting health problems for the former group.

Different types of interactions between various biological elements (e.g., proteins, drugs and diseases) can be modeled using networks for various applications, including drug discovery and finding novel use cases of known drugs. Nevertheless, we are far from having a complete picture of all possible biological interactions that can occur in humans, and so, current networks modeling human biology remain incomplete. To try and compensate for this shortcoming, researchers are beginning to use both knowledge of biological interactions, alongside experimental data. In this work, we show how we can deduce which drugs may be good candidates for treatments by using networks to estimate how a drug can affect a disease, and overlaying elements in our network with those in experimental datasets.

Call for papers: This Research Topic solicits brief, forward-looking contributions from the editorial board members that describe the state of the art, outlining, recent developments and major accomplishments that have been achieved and that need to occur to move the field forward. Authors are encouraged to identify the greatest challenges in the sub-disciplines, and how to address those challenges.

Mental health disorders are common and effective treatments are often available, however affected individuals often struggle to access care. Mobile health (mHealth) resources are increasingly significant tools for mental health care and support, and can overcome barriers to access due to geography, overall health, transportation access, and more. Mobile technologies and applications are accessible and available to users twenty-four hours a day, and can be readily adaptable to individual needs or incorporated into a larger care plan. This PLOS ONE collection brings together a wide range of interdisciplinary research into mobile health solutions for mental health. 

The Digital Medicine Society and the Health Innovation Hub of the German Federal Ministry of Health convened a set of roundtable discussions to bring together international experts in evidence generation for digital medicine products. This Viewpoint highlights findings from these discussions with the aims of (1) accelerating and stimulating innovative approaches to digital medical product evaluation, and (2) promoting international harmonisation of best evidentiary practices. Advancing these topics and fostering international agreement on evaluation approaches will be vital to the safe, effective, and evidence-based deployment and acceptance of digital health applications globally.

If you don't do already, you should consider reading "IEEE Transactions on Medical Robotics and Bionics" regularly... plenty of terrific content like the Special Section on Surgical Vision, Navigation, and Robotics

Continuing advances in digital health technology offer a powerful new toolset to transform how health care is delivered and experienced. However, the digitization of health care is not a simple task; technologies should be implemented only if they will improve care for individuals and the systems that care for them. Thoughtful hybrid approaches will define digitization efforts, and humans—patients, care partners, communities, and clinicians—must be included in these endeavors. Human factors engineering is critical for all technologies and even more so for health-related interventions.

Pronta la bozza di schema dei contratti istituzionali di sviluppo che ogni regione dovrà sottoscrivere entro il 31 maggio 2022 col Ministero della Salute. Nello schema in sostanza ogni regione dovrà inserire il proprio Piano operativo (che in ogni caso prima della stesura del contratto vero e proprio va presentato entro il 28 febbraio) con il cronoprogramma degli interventi

AI and machine learning are top of mind for health care professionals today, but using these technologies for actual patient care has proved difficult. In this FREE, virtual event, “AI & Machine Learning for Health Care Delivery,” NEJM brings to life the articles and case studies from a special issue of NEJM Catalyst Innovations in Care Delivery. The panel of clinical leaders, physicians, and researchers will share their successful experiences with AI in this dynamic, fast-paced event. They will also discuss the barriers to adoption of AI and how to manage risk, bias, and transparency. 

Many advances in artificial intelligence (AI) for health care using deep neural networks have been commercialised. But few AI tools have been implemented in health systems. Why has this chasm occurred? Transparency, suitability, and adaptability are key reasons. The deployment of any new technology is usually managed centrally in hospitals and health systems.

The COVID-19 pandemic has underscored the potential for improved care quality, clinical outcomes, and superior patient experience. Care at Home is evolving as an indispensable component of health systems’ efforts to position themselves for success now and in the future.

È stata rilasciata sul sito del Fascicolo Sanitario Elettronico la versione 2.2 della documentazione per l’interoperabilità tra i sistemi regionali dei fascicoli sanitari.

Le specifiche, frutto del lavoro congiunto dell’Agenzia per l’Italia Digitale e del Consiglio Nazionale delle Ricerche (CNR), in accordo con il Dipartimento per la Trasformazione digitale e Sogei, sono state redatte in coerenza con il Decreto-Legge del 19 maggio 2020, n. 34 e la successiva normativa in merito al Fascicolo Sanitario Elettronico.

Eliminata con un colpo di spugna la HTA ( Health Technology Assessment ) nella scelta dei nuovi test da inserire nello Screening neonatale esteso (SNE) - La commissione Bilancio del Senato ha approvato l’emendamento n. 102.0.14, a firma dei senatori Parente e Conzatti, che va a modificare la legge 167 del 2016 sullo screening neonatale, abrogando il comma 2, articolo 4 e quindi eliminando gli studi di Hta ad opera dell’Agenas.

Improving breast cancer diagnostics with artificial intelligence for MRI - This demonstrated the generalizability of the AI system using multiple data sets from Poland and the US. In subgroup analysis, we observed consistent results across different cancer subtypes and patient demographics. Using the decision curve analysis, we showed that the AI system can reduce unnecessary biopsies in the range of clinically relevant risk thresholds. This would lead to avoiding benign biopsies in up to 20% of all BI-RADS category 4 patients. Finally, we performed an error analysis, investigating situations where AI predictions were mostly incorrect. This exploratory work creates a foundation for deployment and prospective analysis of AI-based models for breast MRI.

Potentially useful for in silico trials - a likelihood-free approach for population calibration is presented, which involves finding distributions of model parameters when fed through the model produces a set of outputs that matches available population data. Unlike most other approaches to population calibration, this method produces uncertainty quantification on the estimated distribution.

Process towards an agreement on reforming research assessment - The Commission has called for organisations to express their interest in being part of a coalition on reforming research assessment.

The coalition will bring together research funding organisations, research performing organisations, national/regional assessment authorities or agencies, associations of research funders, of research performers, of researchers, as well as, learned societies and other relevant organisations, all willing and committed to implement reforms to the current research assessment system.

Our new member, Professor Manuel Mazzara, has recently been appointed a member of the editorial board of Software

Ethical Issues of Digital Twins for Personalized Health Care Service: Preliminary Mapping Study - This paper provides a working definition of digital twins for personalized health care services by identifying 3 features that distinguish the new application from other eHealth services. On the basis of the working definition, this paper further layouts 10 major operational problems and the corresponding ethical risks.

Model Evaluation, Model Selection, and Algorithm Selection in Machine Learning - This article reviews different techniques that can be used for each of these three subtasks and discusses the main advantages and disadvantages of each technique with references to theoretical and empirical studies. Further, recommendations are given to encourage best yet feasible practices in research and applications of machine learning.

The opportunity size for a reimbursed digital health service in a country measured by the addressable statutory health insured or total population size differs a lot.

Digital health companies that want to enter the European market via the reimbursement channel will have to factor in the current implementation status of a standardized reimbursement process as well as the market opportunity size...

 The summary report of public consultation on the proposed initiative for a European Health Data Space are published! - nothing surprising to those that are invested in the topic, but a useful document to have at hand - The European Commission has consulted stakeholders on the proposed initiative for a European Health Data Space. There were 382 valid responses and 64 respondents provided additional documentation (the consultation was open from 3 May until 26 July 2021.)

A bit of optimism from tele-epidemiology and medical informatics?
The WHO has already convened a committee to determine when to update COVID-19 vaccines and with which strain of the virus. Maybe getting an annual COVID-19 shot will become as routine as getting an annual flu shot. We may have to face the sobering fact that COVID-19 isn’t going away. But if we read the antigenic maps carefully, we can stay on top of the virus’s evolution—and our strategies for combatting it can evolve with it.

Very important study that illustrates the gap in knowledge of ML in medical education. The findings are damning given how ML tools/models are steadily becoming commonplace in medicine.

Autonomous anastomosis is a challenging soft-tissue surgery task because it requires intricate imaging, tissue tracking, and surgical planning techniques, as well as a precise execution via highly adaptable control strategies often in unstructured and deformable environments. Here the authors describe an enhanced autonomous strategy for laparoscopic soft tissue surgery and demonstrate robotic laparoscopic small bowel anastomosis in phantom and in vivo intestinal tissues. This enhanced autonomous strategy allows the operator to select among autonomously generated surgical plans and the robot executes a wide range of tasks independently.

Audrey Azoulay, Director-General of UNESCO presented Thursday the first ever global standard on the ethics of artificial intelligence adopted by the member states of UNESCO at the General Conference.  This historical text defines the common values and principles which will guide the construction of the necessary legal infrastructure to ensure the healthy development of AI.

There are rich opportunities for novel AI research involving non-image data types and unconventional problem formulations, which open a broader array of possible datasets. Opportunities also exist in AI–human collaboration, an alternative to the AI-versus-human competitions common in research; we would like to see collaborative setups receive more study, as they may provide better results than either AI or humans alone and are more likely to reflect real medical practice. Despite the potential of the field, major technical and ethical questions remain for medical AI.

A good doctor sees the patient, not the symptoms. Watson saw the symptoms of inefficiency and lack of capability. It did not see the process of care and making whole, where doctors, not data, were what needed to be understood. Ask the right questions, such as "why do some patients get readmitted" and can you spot them early?, and you get the right answers. AI in healthcare is being adopted, and it shows every sign of deserving its place.

Very nice idea: MD++ Datathon! Over 200 medical student participants looking to approach healthcare in a data-driven way...

Feb. 16 | Addressing Bias in the Evolving Landscape of AI in Health Care

This session of the will explore the role of bias related to the use of data for AI in health care solutions and how clinicians, tech companies and other stakeholders are learning mitigation approaches as the AI landscape evolve.

Feb. 23 | How’s Your Digital Transformation Going?

This session will focus on the opportunities and challenges tech companies are experiencing as they navigate the complex healthcare system. 

Though the startup in question, Vainu, sees the partnership as a kind of prison reform that teaches valuable skills, other experts say the claim of job training is more evidence of hype around the promises of AI.

Health care is at a breaking point that could leave behind the people who need it most and further disenfranchise the clinicians who have borne the brunt of this crisis. The scourge of the pandemic has not only rendered an opportunity to act but a shared responsibility to create new, sustainable structures that allow both the patients the system exists to serve and the professionals it depends on to function to live healthier lives.

We build on examples obtained from the academic literature and author experience of data curation practices across a diverse range of sectors to describe a standardized RWD life cycle containing key steps in production of useful data for analysis and insights. We delineate best practices that will add value to current data pipelines. Seven themes are highlighted that ensure sustainability and scalability for RWD life cycles: data standards adherence, tailored quality assurance, data entry incentivization, deploying natural language processing, data platform solutions, RWD governance, and ensuring equity and representation in data.

Regional collaboration among scientists embedded in or working closely with local response agencies, but supported by a global network of peers, will accelerate the use of these data to help our communities. Without this type of sustained, broad-based and equitably distributed capacity investment, no amount of additional data or improved methodology is likely to result in substantial gains for disaster-affected communities.

Development and validation of a radiopathomics model to predict pathological complete response to neoadjuvant chemoradiotherapy in locally advanced rectal cancer: a multicentre observational study