Autori : Nicola Rieke Jonny Hancox Wenqi Li Fausto Milletarì Holger R. Roth Shadi Albarqouni Spyridon Bakas Mathieu N. Galtier Bennett A. Landman Klaus Maier-Heinová Sébastien Ourselin Micah Sheller Ronald M. Summers Andrew Trask Daguang Xu Maximilian Baust M. Jorge Cardoso Authors: Nikolaj Rieke Jonny Hancox Ján Li Fausto Milletarí Holger R. Rothová Shadi Albarqouni Spyridon Bakas Mathieu N. Galtier Bennett A. Landman Klaus Maier-Heinová Sébastien Ourselin Micah Sheller Režisér Ronald M. Summers Andrew Trask Daguang Xu Maximilián Baust M. Jorge Cardoso Abstract Údaje riadené strojovým učením (ML) sa objavili ako sľubný prístup k budovaniu presných a robustných štatistických modelov z lekárskych údajov, ktoré sú zhromažďované v obrovských množstvách modernými systémami zdravotnej starostlivosti. Existujúce lekárske údaje nie sú plne využívané ML predovšetkým preto, že sedí v dátových silách a obavy o súkromie obmedzujú prístup k týmto údajom. Avšak bez prístupu k dostatočným údajom bude ML zabrániť dosiahnutiu svojho plného potenciálu a v konečnom dôsledku urobiť prechod z výskumu na klinickú prax. Tento dokument zvažuje kľúčové faktory prispievajúce k tomuto problému, skúma, ako federované učenie (FL) môže poskytnúť riešenie pre budúc Úvodná Research on artificial intelligence (AI), and particularly the advances in machine learning (ML) and deep learning (DL) have led to disruptive innovations in radiology, pathology, genomics and other fields. Modern DL models feature millions of parameters that need to be learned from sufficiently large curated data sets in order to achieve clinical-grade accuracy, while being safe, fair, equitable and generalising well to unseen data , , , . 1 2 3 4 5 Napríklad školenie detektora nádorov založeného na AI si vyžaduje veľkú databázu zahŕňajúcu celé spektrum možných anatómie, patológií a typov vstupných údajov. . Even if data anonymisation could bypass these limitations, it is now well understood that removing metadata such as patient name or date of birth is often not enough to preserve privacy Je možné napríklad rekonštruovať tvár pacienta z údajov počítačovej tomografie (CT) alebo magnetickej rezonancie (MRI). Ďalším dôvodom, prečo zdieľanie údajov nie je v zdravotníctve systematické, je to, že zhromažďovanie, spravovanie a udržiavanie vysokokvalitného súboru údajov si vyžaduje značný čas, úsilie a náklady.V dôsledku toho môžu mať takéto súbory údajov významnú obchodnú hodnotu, takže je menej pravdepodobné, že budú slobodne zdieľané. 6 7 8 Federalizované vzdelávanie (FL) , , is a learning paradigm seeking to address the problem of data governance and privacy by training algorithms collaboratively without exchanging the data itself. Originally developed for different domains, such as mobile and edge device use cases , nedávno získala trakciu pre zdravotnícke aplikácie , , , , , , , FL umožňuje získavať poznatky v spolupráci, napr. vo forme modelu konsenzu, bez toho, aby sa údaje pacientov presunuli za hranice firewallov inštitúcií, v ktorých sa nachádzajú.Namiesto toho sa proces ML vyskytuje lokálne v každej zúčastnenej inštitúcii a prenášajú sa iba charakteristiky modelu (napr. parametre, gradienty) ako je znázornené na obrázku. . Recent research has shown that models trained by FL can achieve performance levels comparable to ones trained on centrally hosted data sets and superior to models that only see isolated single-institutional data , . 9 10 11 12 13 14 15 16 17 18 19 20 1 16 17 FL agregácia servera – typický FL pracovný postup, v ktorom federácia tréningových uzlov dostane globálny model, predloží svoje čiastočne vyškolené modely na centrálny server intermitentne na agregáciu a potom pokračuje v tréningu na konsenzusovom modeli, ktorý server vráti. FL peer to peer – alternatívna formulácia FL, v ktorej každý tréningový uzol vymieňa svoje čiastočne vyškolené modely s niektorými alebo všetkými svojimi rovesníkmi a každý robí svoju vlastnú agregáciu. Centralizované školenie – všeobecný pracovný postup mimo FL školenia, v ktorom lokality získavajúce dáta darujú svoje dáta do centrálneho jazera údajov, z ktorého oni a iní môžu extrahovať dáta pre miestne, nezávislé školenie. a b c A successful implementation of FL could thus hold a significant potential for enabling precision medicine at large-scale, leading to models that yield unbiased decisions, optimally reflect an individual’s physiology, and are sensitive to rare diseases while respecting governance and privacy concerns. However, FL still requires rigorous technical consideration to ensure that the algorithm is proceeding optimally without compromising safety or patient privacy. Nevertheless, it has the potential to overcome the limitations of approaches that require a single pool of centralised data. We envision a federated future for digital health and with this perspective paper, we share our consensus view with the aim of providing context and detail for the community regarding the benefits and impact of FL for medical applications (section “Data-driven medicine requires federated efforts”), as well as highlighting key considerations and challenges of implementing FL for digital health (section “Technical considerations”). Data-driven medicine requires federated efforts ML and especially DL is becoming the de facto knowledge discovery approach in many industries, but successfully implementing data-driven applications requires large and diverse data sets. However, medical data sets are difficult to obtain (subsection “The reliance on data”). FL addresses this issue by enabling collaborative learning without centralising data (subsection “The promise of federated efforts”) and has already found its way to digital health applications (subsection “Current FL efforts for digital health”). This new learning paradigm requires consideration from, but also offers benefits to, various healthcare stakeholders (section “Impact on stakeholders”). The reliance on data Hoci je to dobre známa požiadavka, najmodernejšie algoritmy sa zvyčajne vyhodnocujú na starostlivo vyhodnotených dátových súboroch, často pochádzajúcich len z niekoľkých zdrojov. To môže zaviesť predsudky, v ktorých demografia (napr. pohlavie, vek) alebo technická nerovnováha (napr. akvizitný protokol, výrobca zariadení) skresľuje predpovede a nepriaznivo ovplyvňuje presnosť pre určité skupiny alebo lokality. Potreba veľkých databáz pre výcvik AI vyvolala mnohé iniciatívy, ktoré sa snažia spojiť údaje z viacerých inštitúcií. Tieto údaje sa často zhromažďujú do takzvaných dátových jazier. , alebo ako zdroj pre hospodársky rast a vedecký pokrok, napr. Národný bezpečný prístav NHS Škótsko , French Health Data Hub , and Health Data Research UK . 21 22 23 24 Substantial, albeit smaller, initiatives include the Human Connectome , the UK Biobank , the Cancer Imaging Archive (TCIA) , NIH CXR8 NIH hlboké poškodenie , the Cancer Genome Atlas (TCGA) , the Alzheimer’s Disease Neuroimaging Initiative (ADNI) Ako aj lekárske výzvy ako napríklad výzva CAMELYON Medzinárodná výzva pre multimodálnu segmentáciu nádorov mozgu (BraTS) , , or the Medical Segmentation Decathlon . Public medical data is usually task- or disease-specific and often released with varying degrees of license restrictions, sometimes limiting its exploitation. 25 26 27 28 29 30 31 32 33 34 35 36 37 Centralising or releasing data, however, poses not only regulatory, ethical and legal challenges, related to privacy and data protection, but also technical ones. Anonymising, controlling access and safely transferring healthcare data is a non-trivial, and sometimes impossible task. Anonymised data from the electronic health record can appear innocuous and GDPR/PHI compliant, but just a few data elements may allow for patient reidentification . The same applies to genomic data and medical images making them as unique as a fingerprint Preto, pokiaľ proces anonymizácie nezničí vernosť údajov, pravdepodobne zneužívajúcich, opätovnú identifikáciu pacienta alebo únik informácií nemožno vylúčiť.Gated access pre schválených používateľov je často navrhnutý ako hypotetické riešenie tohto problému. Avšak okrem obmedzenia dostupnosti údajov je to praktické len v prípadoch, keď je súhlas udelený vlastníkom údajov bezpodmienečný, pretože stiahnutie údajov od tých, ktorí mohli mať prístup k údajom, je prakticky nevymáhateľné. 7 38 Sľub federálnych snah Sľub spoločnosti FL je jednoduchý – riešiť výzvy týkajúce sa ochrany osobných údajov a riadenia údajov tým, že umožní ML z údajov, ktoré nie sú umiestnené spoločne. V nastavení FL každý správca údajov nielen definuje svoje vlastné procesy riadenia a súvisiace politiky ochrany osobných údajov, ale tiež kontroluje prístup k údajom a má schopnosť ich zrušiť. To zahŕňa výcvik aj fázu validácie. Týmto spôsobom by FL mohol vytvoriť nové príležitosti, napríklad umožnením veľkého rozsahu validácie v rámci inštitúcií alebo umožnením nového výskumu o zriedkavých ochoreniach, kde sú miery incidentov nízke a dátové súbory v každej inštitúcii sú príliš malé. Presun modelu na údaje a nie naopak má ďalšiu veľkú výhodu: vysokorýchlostné As depicted in Fig. , a FL workflow can be realised with different topologies and compute plans. The two most common ones for healthcare applications are via an aggregation server , , a peer to peer prístupy , Vo všetkých prípadoch FL implicitne ponúka určitý stupeň súkromia, pretože účastníci FL nikdy nemajú priamy prístup k údajom z iných inštitúcií a dostávajú len modelové parametre, ktoré sú agregované na viacerých účastníkoch.V pracovnom postupe FL so serverom agregácie môžu zúčastnené inštitúcie dokonca zostať navzájom neznáme. , , , Preto mechanizmy, ako je diferenciálna súkromie , or learning from encrypted data have been proposed to further enhance privacy in a FL setting (c.f. section “Technical considerations”). Overall, the potential of FL for healthcare applications has sparked interest in the community and FL techniques are a growing area of research , . 2 16 17 18 15 39 40 41 42 43 44 45 46 12 20 FL topológie – komunikačná architektúra federácie. Centralizovaný: agregácia server koordinuje tréningové iterácie a zhromažďuje, agreguje a distribuuje modely do a z tréningových uzlov (Hub & Spoke). Decentralizované: Každý tréningový uzol je pripojený k jednému alebo viacerým rovesníkom a agregácia prebieha na každom uzle paralelne. Hierarchical: federated networks can be composed from several sub-federations, which can be built from a mix of Peer to Peer and Aggregation Server federations ( )). FL compute plans—trajectory of a model across several partners. Sekvenčné vzdelávanie / cyklické transferové učenie. agregácia serverov, Peer to Peer. a b c d e f g Súčasné úsilie FL o digitálne zdravie Vzhľadom k tomu, že FL je všeobecná paradigma učenia, ktorá odstraňuje požiadavku na zhromažďovanie údajov pre vývoj modelov umelej inteligencie, rozsah aplikácií FL pokrýva celý rozsah AI pre zdravotnú starostlivosť. Poskytnutím príležitosti zachytiť väčšiu variabilitu údajov a analyzovať pacientov v rôznych demografických oblastiach môže FL umožniť rušivé inovácie pre budúcnosť, ale je tiež zamestnaná práve teraz. V kontexte elektronických zdravotných záznamov (EHR) napríklad FL pomáha reprezentovať a nájsť klinicky podobných pacientov. , , as well as predicting hospitalisations due to cardiac events , úmrtnosť a doba pobytu ICU Aplikovateľnosť a výhody FL boli preukázané aj v oblasti lekárskeho zobrazovania, pre segmentáciu celého mozgu v MRI , rovnako ako segmentácia mozgového nádoru , . Recently, the technique has been employed for fMRI classification to find reliable disease-related biomarkers and suggested as a promising approach in the context of COVID-19 . 13 47 14 19 15 16 17 18 48 It is worth noting that FL efforts require agreements to define the scope, aim and technologies used which, since it is still novel, can be difficult to pin down. In this context, today’s large-scale initiatives really are the pioneers of tomorrow’s standards for safe, fair and innovative collaboration in healthcare applications. These include consortia that aim to advance Výskum, ako je napríklad projekt Trustworthy Federated Data Analytics (TFDA) Spoločná obrazová platforma Nemeckého konzorcia pre rakovinu , which enable decentralised research across German medical imaging research institutions. Another example is an international research collaboration that uses FL for the development of AI models for the assessment of mammograms Štúdia ukázala, že FL generované modely prekonali tie, ktoré boli vyškolené na údajoch jedného inštitútu a boli generalizovateľnejšie, takže stále fungovali dobre na údajoch iných inštitútov. academic 49 50 51 Spojením zdravotníckych zariadení, ktoré nie sú obmedzené na výskumné centrá, môže FL priamo Vplyv. prebiehajúci projekt HealthChain , for example, aims to develop and deploy a FL framework across four hospitals in France. This solution generates common models that can predict treatment response for breast cancer and melanoma patients. It helps oncologists to determine the most effective treatment for each patient from their histology slides or dermoscopy images. Another large-scale effort is the Federated Tumour Segmentation (FeTS) initiative , which is an international federation of 30 committed healthcare institutions using an open-source FL framework with a graphical user interface. The aim is to improve tumour boundary detection, including brain glioma, breast tumours, liver tumours and bone lesions from multiple myeloma patients. clinical 52 53 Another area of impact is within výskum a preklad. FL umožňuje spolupracujúci výskum aj pre konkurenčné spoločnosti.V tomto kontexte je jednou z najväčších iniciatív projekt Melloddy Je to projekt zameraný na nasadenie multi-task FL v dátových súboroch 10 farmaceutických spoločností.Trenovaním spoločného prediktívneho modelu, ktorý odvodzuje spôsob, akým sa chemické zlúčeniny viažu na proteíny, partneri majú v úmysle optimalizovať proces objavovania liekov bez toho, aby odhalili svoje vysoko cenné interné údaje. priemyselný 54 Impact on stakeholders FL zahŕňa zmenu paradigmy z centralizovaných dátových jazier a je dôležité pochopiť jej vplyv na rôzne zainteresované strany v ekosystéme FL. Clinicians Clinicians are usually exposed to a sub-group of the population based on their location and demographic environment, which may cause biased assumptions about the probability of certain diseases or their interconnection. By using ML-based systems, e.g., as a second reader, they can augment their own expertise with expert knowledge from other institutions, ensuring a consistency of diagnosis not attainable today. While this applies to ML-based system in general, systems trained in a federated fashion are potentially able to yield even less biased decisions and higher sensitivity to rare cases as they were likely exposed to a more complete data distribution. However, this demands some up-front effort such as compliance with agreements, e.g., regarding the data structure, annotation and report protocol, which is necessary to ensure that the information is presented to collaborators in a commonly understood format. Patients Pacienti sa zvyčajne liečia lokálne. Zriadenie FL v globálnom meradle by mohlo zabezpečiť vysokú kvalitu klinických rozhodnutí bez ohľadu na umiestnenie liečby. Najmä pacienti, ktorí potrebujú lekársku starostlivosť v odľahlých oblastiach, by mohli mať prospech z rovnakých vysokokvalitných diagnóz s pomocou ML, ktoré sú k dispozícii v nemocniciach s veľkým počtom prípadov. To isté platí pre zriedkavé alebo geograficky nezvyčajné ochorenia, ktoré pravdepodobne budú mať miernejšie dôsledky, ak sa môžu urobiť rýchlejšie a presnejšie diagnózy. FL môže tiež znížiť prekážku stať sa darcom údajov, pretože pacienti môžu byť uistení, že údaje zostanú s ich vlastnou inštitúciou a prístup k údajom môže byť Hospitals and practices Hospitals and practices can remain in full control and possession of their patient data with complete traceability of data access, limiting the risk of misuse by third parties. However, this will require investment in on-premise computing infrastructure or private-cloud service provision and adherence to standardised and synoptic data formats so that ML models can be trained and evaluated seamlessly. The amount of necessary compute capability depends of course on whether a site is only participating in evaluation and testing efforts or also in training efforts. Even relatively small institutions can participate and they will still benefit from collective models generated. Researchers and AI developers Researchers and AI developers stand to benefit from access to a potentially vast collection of real-world data, which will particularly impact smaller research labs and start-ups. Thus, resources can be directed towards solving clinical needs and associated technical problems rather than relying on the limited supply of open data sets. At the same time, it will be necessary to conduct research on algorithmic strategies for federated training, e.g., how to combine models or updates efficiently, how to be robust to distribution shifts , , . FL-based development implies also that the researcher or AI developer cannot investigate or visualise all of the data on which the model is trained, e.g., it is not possible to look at an individual failure case to understand why the current model performs poorly on it. 11 12 20 Healthcare providers Healthcare providers in many countries are affected by the on-going paradigm shift from volume-based, i.e., fee-for-service-based, to value-based healthcare, which is in turn strongly connected to the successful establishment of precision medicine. This is not about promoting more expensive individualised therapies but instead about achieving better outcomes sooner through more focused treatment, thereby reducing the cost. FL has the potential to increase the accuracy and robustness of healthcare AI, while reducing costs and improving patient outcomes, and may therefore be vital to precision medicine. Manufacturers Výrobcovia zdravotníckeho softvéru a hardvéru by mohli profitovať aj z FL, pretože kombinácia učenia z mnohých zariadení a aplikácií, bez zverejnenia informácií špecifických pre pacienta, môže uľahčiť nepretržité overovanie alebo zlepšovanie ich systémov založených na ML. Avšak realizácia takejto schopnosti môže vyžadovať významné upgrade na lokálne výpočty, ukladanie dát, sieťové schopnosti a súvisiaci softvér. Technical considerations FL is perhaps best-known from the work of Konečnỳ et al. , but various other definitions have been proposed in the literature , , , . A FL workflow (Fig. ) možno realizovať prostredníctvom rôznych topológií a výpočtových plánov (obr. V tejto časti budeme podrobnejšie diskutovať o tom, čo je FL, ako aj zdôrazňovať kľúčové výzvy a technické úvahy, ktoré vznikajú pri aplikácii FL v digitálnom zdraví. 55 9 11 12 20 1 2 Federated learning definition FL is a learning paradigm in which multiple parties train collaboratively without the need to exchange or centralise data sets. A general formulation of FL reads as follows: Let denote a global loss function obtained via a weighted combination of local losses , computed from private data , ktorý má bydlisko na jednotlivých zúčastnených stranách a nikdy sa medzi nimi nezdieľal: K xk kde > 0 označuje príslušné hmotnostné koeficienty. wk In practice, each participant typically obtains and refines a global consensus model by conducting a few rounds of optimisation locally and before sharing updates, either directly or via a parameter server. The more rounds of local training are performed, the less it is guaranteed that the overall procedure is minimising (Eq. ) , . The actual process for aggregating parameters depends on the network topology, as nodes might be segregated into sub-networks due to geographical or legal constraints (see Fig. Agresívne stratégie sa môžu spoliehať na jeden agregujúci uzol (modely hubov a hovorcov) alebo na viaceré uzly bez akejkoľvek centralizácie. Príkladom je peer-to-peer FL, kde existujú spojenia medzi všetkými alebo podskupinou účastníkov a aktualizácie modelu sú zdieľané iba medzi priamo prepojenými lokalitami. , Vezmite na vedomie, že stratégie agregácie nevyhnutne nevyžadujú informácie o úplnej aktualizácii modelu; klienti by sa mohli rozhodnúť zdieľať iba podskupinu parametrov modelu, aby sa znížila komunikácia a zabezpečilo lepšie zachovanie súkromia. or to produce multi-task learning algorithms having only part of their parameters learned in a federated manner. 1 9 12 2 15 56 10 A unifying framework enabling various training schemes may disentangle compute resources (data and servers) from the , as depicted in Fig. . The latter defines the trajectory of a model across several partners, to be trained and evaluated on specific data sets. Počítačový plán 2 Challenges and considerations Despite the advantages of FL, it does not solve all issues that are inherent to learning on medical data. A successful model training still depends on factors like data quality, bias and standardisation . These issues have to be solved for both federated and non-federated learning efforts via appropriate measures, such as careful study design, common protocols for data acquisition, structured reporting and sophisticated methodologies for discovering bias and hidden stratification. In the following, we touch upon the key aspects of FL that are of particular relevance when applied to digital health and need to be taken into account when establishing FL. For technical details and in-depth discussion, we refer the reader to recent surveys , , . 2 11 12 20 Data heterogeneity Medical data is particularly diverse—not only because of the variety of modalities, dimensionality and characteristics in general, but even within a specific protocol due to factors such as acquisition differences, brand of the medical device or local demographics. FL may help address certain sources of bias through potentially increased diversity of data sources, but inhomogeneous data distribution poses a challenge for FL algorithms and strategies, as many are assuming independently and identically distributed (IID) data across the participants. In general, strategies such as are prone to fail under these conditions , , , čiastočne porazil samotný účel spoločných vzdelávacích stratégií. nedávne výsledky však naznačujú, že tréning FL je stále uskutočniteľný , even if medical data is not uniformly distributed across the institutions , alebo zahŕňa miestny bias . Research addressing this problem includes, for example, , part-data-sharing strategy and FL with domain-adaptation Ďalšou výzvou je, že heterogénnosť údajov môže viesť k situácii, v ktorej globálne optimálne riešenie nemusí byť optimálne pre jednotlivých miestnych účastníkov. FedAvg 9 9 57 58 59 16 17 51 FedProx 57 58 18 Súkromie a bezpečnosť Údaje o zdravotnej starostlivosti sú vysoko citlivé a musia byť zodpovedajúcim spôsobom chránené v súlade s príslušnými postupmi dôvernosti.Niektoré z kľúčových úvah sú preto kompromisy, stratégie a zostávajúce riziká týkajúce sa potenciálu ochrany súkromia spoločnosti FL. Privacy vs. performance: It is important to note that FL does not solve all potential privacy issues and—similar to ML algorithms in general—will always carry some risks. Privacy-preserving techniques for FL offer levels of protection that exceed today’s current commercially available ML models . However, there is a trade-off in terms of performance and these techniques may affect, for example, the accuracy of the final model . Furthermore, future techniques and/or ancillary data could be used to compromise a model previously considered to be low-risk. 12 10 Level of trust: Broadly speaking, participating parties can enter two types of FL collaboration: —for FL consortia in which all parties are considered trustworthy and are bound by an enforceable collaboration agreement, we can eliminate many of the more nefarious motivations, such as deliberate attempts to extract sensitive information or to intentionally corrupt the model. This reduces the need for sophisticated counter-measures, falling back to the principles of standard collaborative research. dôveryhodný —V systémoch FL, ktoré fungujú vo väčších rozmeroch, môže byť nepraktické vytvoriť vykonateľnú dohodu o spolupráci. Niektorí klienti sa môžu zámerne pokúsiť degradovať výkon, znižovať systém alebo extrahovať informácie od iných strán. Preto budú potrebné bezpečnostné stratégie na zmiernenie týchto rizík, ako je pokročilé šifrovanie odovzdávania modelov, bezpečná autentifikácia všetkých strán, vysledovateľnosť akcií, diferenciálne súkromie, overovacie systémy, integrita vykonávania, dôvernosť modelu a ochrana pred útokmi súperov. Non-trusted Information leakage: By definition, FL systems avoid sharing healthcare data among participating institutions. However, the shared information may still indirectly expose private data used for local training, e.g., by model inversion of the model updates, the gradients themselves alebo nepriateľské útoky , . FL is different from traditional training insofar as the training process is exposed to multiple parties, thereby increasing the risk of leakage via reverse-engineering if adversaries can observe model changes over time, observe specific model updates (i.e., a single institution’s update), or manipulate the model (e.g., induce additional memorisation by others through gradient-ascent-style attacks). Developing counter-measures, such as limiting the granularity of the updates and adding noise , and ensuring adequate differential privacy , may be needed and is still an active area of research . 60 61 62 63 16 18 44 12 Sledovateľnosť a zodpovednosť Rovnako ako pri všetkých bezpečnostne kritických aplikáciách, reprodukovateľnosť systému je pre FL v zdravotníctve dôležitá. Na rozdiel od centralizovaného školenia, FL vyžaduje multi-party výpočty v prostrediach, ktoré vykazujú značnú rozmanitosť z hľadiska hardvéru, softvéru a sietí. Sledovateľnosť všetkých systémových aktív vrátane histórie prístupu k údajom, tréningových konfigurácií a hyperparametrického nastavenia počas tréningových procesov je preto povinná. Najmä v nedôveryhodných federáciách, sledovateľnosť a procesy zodpovednosti vyžadujú integritu vykonávania. Po procese školenia dosiahne vzájomne dohodnuté kritériá optimalizácie modelu, môže byť tiež užitočné merať výšku príspevku od každého účastníka, ako napríklad výpočtové zdroje spotrebované Jedným z dôsledkov FL je, že výskumníci nie sú schopní preskúmať údaje, na ktorých sú modely vyškolené, aby mali zmysel pre neočakávané výsledky. Okrem toho, prijímanie štatistických meraní ich tréningových údajov ako súčasť pracovného postupu vývoja modelu bude musieť byť schválené spolupracujúcimi stranami ako neporušujúce súkromie. Hoci každá lokalita bude mať prístup k svojim vlastným surovým údajom, federácie sa môžu rozhodnúť poskytnúť nejaký druh bezpečného vnútrouzlového zobrazovacieho zariadenia na uspokojenie tejto potreby alebo môže poskytnúť iný spôsob, ako zvýšiť vysvetľovateľnosť a interpretovateľnosť globálneho modelu. 64 System architecture Na rozdiel od bežiaceho veľkého rozsahu FL medzi spotrebiteľskými zariadeniami, ako je McMahan et al. , healthcare institutional participants are equipped with relatively powerful computational resources and reliable, higher-throughput networks enabling training of larger models with many more local training steps, and sharing more model information between nodes. These unique characteristics of FL in healthcare also bring challenges such as ensuring data integrity when communicating by use of redundant nodes, designing secure encryption methods to prevent data leakage, or designing appropriate node schedulers to make best-use of the distributed computational devices and reduce idle time. 9 The administration of such a federation can be realised in different ways. In situations requiring the most stringent data privacy between parties, training may operate via some sort of “honest broker” system, in which a trusted third party acts as the intermediary and facilitates access to data. This setup requires an independent entity controlling the overall system, which may not always be desirable, since it could involve additional cost and procedural viscosity. However, it has the advantage that the precise internal mechanisms can be abstracted away from the clients, making the system more agile and simpler to update. In a peer-to-peer system each site interacts directly with some or all of the other participants. In other words, there is no gatekeeper function, all protocols must be agreed up-front, which requires significant agreement efforts, and changes must be made in a synchronised fashion by all parties to avoid problems. Additionally, in a trustless-based architecture the platform operator may be cryptographically locked into being honest by means of a secure protocol, but this may introduce significant computational overheads. záver ML, and particularly DL, has led to a wide range of innovations in the area of digital healthcare. As all ML methods benefit greatly from the ability to access data that approximates the true global distribution, FL is a promising approach to obtain powerful, accurate, safe, robust and unbiased models. By enabling multiple parties to train collaboratively without the need to exchange or centralise data sets, FL neatly addresses issues related to egress of sensitive medical data. As a consequence, it may open novel research and business avenues and has the potential to improve patient care globally. However, already today, FL has an impact on nearly all stakeholders and the entire treatment cycle, ranging from improved medical image analysis providing clinicians with better diagnostic tools, over true precision medicine by helping to find similar patients, to collaborative and accelerated drug discovery decreasing cost and time-to-market for pharma companies. Not all technical questions have been answered yet and FL will certainly be an active research area throughout the next decade . Despite this, we truly believe that its potential impact on precision medicine and ultimately improving medical care is very promising. 12 Reporting summary Further information on research design is available in the linked to this article. Nature Research Reporting Summary References LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. , 436 (2015). Nature 521 Wang, F., Casalino, L. P. & Khullar, D. Deep learning in medicine—promise, progress, and challenges. , 293–294 (2019). JAMA Intern. Med. 179 Chartrand, G. et al. Deep learning: a primer for radiologists. , 2113–2131 (2017). Radiographics 37 De Fauw, J. et al. Clinically applicable deep learning for diagnosis and referral in retinal disease. , 1342 (2018). Nat. Med. 24 Sun, C., Shrivastava, A., Singh, S. & Gupta, A. Revisiting unreasonable effectiveness of data in deep learning era. In , 843–852 ( , 2017). Proceedings of the IEEE international conference on computer vision IEEE Van Panhuis, W. G. et al. A systematic review of barriers to data sharing in public health. , 1144 (2014). BMC Public Health 14 Rocher, L., Hendrickx, J. M. & De Montjoye, Y.-A. Estimating the success of re-identifications in incomplete datasets using generative models. , 1–9 (2019). Nat. Commun. 10 Schwarz, C. G. et al. Identification of anonymous mri research participants with face-recognition software. , 1684–1686 (2019). N. Engl. J. Med. 381 McMahan, B., Moore, E., Ramage, D., Hampson, S. & y Arcas, B. A. Communication-efficient learning of deep networks from decentralized data. In , 1273–1282. (2017). Artificial Intelligence and Statistics https://scholar.google.de/scholar?hl=de&as_sdt=0%2C5&q=Communicationefficient+learning+of+deep+networks+from+decentralized+data&btnG= Li, T., Sahu, A. K., Talwalkar, A. & Smith, V. Federated learning: Challenges, methods, and future directions. , 50–60 (IEEE, 2020). IEEE Signal Processing Magazine 37 Yang, Q., Liu, Y., Chen, T. & Tong, Y. Federated machine learning: concept and applications. , 12 (2019). ACM Trans. Intell. Syst. Technol. (TIST) 10 Kairouz, P. et al. Advances and open problems in federated learning. (2019). arXiv preprint arXiv:1912.04977 Lee, J. et al. Privacy-preserving patient similarity learning in a federated environment: development and analysis. , e20 (2018). JMIR Med. Inform. 6 Brisimi, T. S. et al. Federated learning of predictive models from federated electronic health records. , 59–67 (2018). Int. J. Med. Inform. 112 Roy, A. G., Siddiqui, S., Pölsterl, S., Navab, N. & Wachinger, C. Braintorrent: a peer-to-peer environment for decentralized federated learning. (2019). arXiv preprint arXiv:1905.06731 Li, W. et al. Privacy-preserving federated brain tumour segmentation. In , 133–141 (Springer, 2019). International Workshop on Machine Learning in Medical Imaging Sheller, M. J., Reina, G. A., Edwards, B., Martin, J. & Bakas, S. Multi-institutional deep learning modeling without sharing patient data: a feasibility study on brain tumor segmentation. In , 92–104 (Springer, 2018). International MICCAI Brainlesion Workshop Li, X. et al. Multi-site fmri analysis using privacy-preserving federated learning and domain adaptation: abide results. (2020). arXiv preprint arXiv:2001.05647 Huang, L. et al. Patient clustering improves efficiency of federated machine learning to predict mortality and hospital stay time using distributed electronic medical records. , 103291 (2019). J. Biomed. Inform. 99 Xu, J. & Wang, F. Federated learning for healthcare informatics. (2019). arXiv preprint arXiv:1911.06270 Roy, A. & Banerjee, A. Ibm’s merge healthcare acquisitio . (2015) (Accessed 10 February 2020). n https://www.reuters.com/article/us-merge-healthcare-m-a-ibm/ibm-to-buy-merge-healthcare-in-1-billion-deal-idUSKCN0QB1ML20150806 Nhs scotland’s national safe haven. (2015) (Accessed 10 February 2020). https://www.gov.scot/publications/charter-safe-havens-scotland-handling-unconsented-data-national-health-service-patient-records-support-research-statistics/pages/4/ Cuggia, M. & Combes, S. The french health data hub and the german medical informatics initiatives: Two national projects to promote data sharing in healthcare. , 195–202 (2019). Yearbook Med. Informat. 28 Health Data Research UK. (Health Data Research UK, 2020) (Accessed 10 Feb 2020). https://www.hdruk.ac.uk/ Sporns, O., Tononi, G. & Kötter, R. The human connectome: a structural description of the human brain. . , e42, (2005). PLoS Comput. Biol 1 https://doi.org/10.1371/journal.pcbi.0010042 Sudlow, C. et al. Uk biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. . , e1001779. (2015). PLoS Med 12 https://doi.org/10.1371/journal.pmed.1001779 Clark, K. et al. The cancer imaging archive (tcia): maintaining and operating a public information repository. , 1045–1057 (2013). J. Digit. Imaging. 26 Wang, X. et al. Chestx-ray8: Hospital-scale chest X-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases. In , 2097–2106 ( , 2017). Proceedings of the IEEE conference on computer vision and pattern recognition IEEE Yan, K., Wang, X., Lu, L. & Summers, R. M. Deeplesion: automated mining of large-scale lesion annotations and universal lesion detection with deep learning. , 036501 (2018). J Med. Imaging. 5 Tomczak, K., Czerwińska, P. & Wiznerowicz, M. The cancer genome atlas (tcga): an immeasurable source of knowledge. , A68 (2015). Contemp. Oncol. 19 Jack Jr., C. R. et al. The alzheimer’s disease neuroimaging initiative (adni): Mri methods. , 685–691 (2008). J. Magn. Reson. Imaging 27 . (2020) (Accessed 24 July 2020). Grand Challenge-a Platform for End-to-end Development of Machine Learning Solutions in Biomedical Imaging https://grand-challenge.org/ Litjens, G. et al. 1399 h&e-stained sentinel lymph node sections of breast cancer patients: the camelyon dataset. , giy065 (2018). GigaScience 7 Menze, B. H. et al. The multimodal brain tumor image segmentation benchmark (brats). , 1993–2024 (2014). IEEE Trans. Med. Imaging 34 Bakas, S. et al. Identifying the best machine learning algorithms for brain tumor segmentation, progression assessment, and overall survival prediction in the brats challenge. (2018). arXiv preprint arXiv:1811.02629 Bakas, S. et al. Advancing the cancer genome atlas glioma MRI collections with expert segmentation labels and radiomic features. , 170117 (2017). Sci. Data 4 Simpson, A. L. et al. A large annotated medical image dataset for the development and evaluation of segmentation algorithms. (2019). arXiv preprint arXiv:1902.09063 Yeh, F.-C. et al. Quantifying differences and similarities in whole-brain white matter architecture using local connectome fingerprints. , e1005203 (2016). PLoS Comput. Biol. 12 Chang, K. et al. Distributed deep learning networks among institutions for medical imaging. , 945–954 (2018). J. Am. Med. Inform. Assoc. 25 Shokri, R., Stronati, M., Song, C. & Shmatikov, V. Membership inference attacks against machine learning models. In , 3-18 (IEEE, 2017). 2017 IEEE Symposium on Security and Privacy (SP) Sablayrolles, A., Douze, M., Ollivier, Y., Schmid, C. & Jégou, H. White-box vs black-box: Bayes optimal strategies for membership inference. In Chaudhuri, K. & Salakhutdinov, R. (eds) , 5558–5567. (PMLR, 2019). Proceedings of the 36th International Conference on Machine Learning, {ICML} 97 http://proceedings.mlr.press/v97/sablayrolles19a.html Zhang, C., Bengio, S., Hardt, M., Recht, B. & Vinyals, O. Understanding deep learning requires rethinking generalization. In , (OpenReview.net, 2017). 5th International Conference on Learning Representations, {ICLR}. https://openreview.net/forum?id=Sy8gdB9xx Carlini, N., Liu, C., Erlingsson, Ú., Kos, J. & Song, D. The secret sharer: evaluating and testing unintended memorization in neural networks. In Heninger, N. & Traynor, P. (eds) { } ({ } , 267–284. ({USENIX} Association, Santa Clara, CA, USA, 2019). 28th USENIX Security Symposium USENIX Security 19 https://www.usenix.org/conference/usenixsecurity19/presentation/carlini Abadi, M. et al. Deep learning with differential privacy. In , 308–318 (ACM, 2016). Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security Shokri, R. & Shmatikov, V. Privacy-preserving deep learning. In , 1310–1321 (ACM, 2015). Proceedings of the 22nd ACM SIGSAC conference on computer and communications security Langlotz, C. P. et al. A roadmap for foundational research on artificial intelligence in medical imaging: from the 2018 nih/rsna/acr/the academy workshop. , 781–791 (2019). Radiology 291 Kim, Y., Sun, J., Yu, H. & Jiang, X. Federated Tensor Factorization for Computational Phenotyping. In . 887–895. (ACM, Halifax, NS, Canada, 2017). Proceedings of the 23rd {ACM} {SIGKDD} International Conference on Knowledge Discoveryand Data Mining https://doi.org/10.1145/3097983.3098118 He, C., Annavaram, M. & Avestimehr, S. Fednas: Federated deep learning via neural architecture search. (2020). https://sites.google.com/view/cvpr20-nas/ Trustworthy federated data analytics (tfda). (2020) (Accessed 28 May 2020). https://tfda.hmsp.center/ Joint Imaging Platform (Jip). (2020) (Accessed 28 May 2020). https://jip.dktk.dkfz.de/jiphomepage/ Medical institutions collaborate to improve mammogram assessment ai. (2020) (Accessed 28 May 2020). https://blogs.nvidia.com/blog/2020/04/15/federated-learning-mammogram-assessment/ Healthchain consortium. (2020) (Accessed 28 May 2020). https://www.substra.ai/en/healthchain-project The federated tumor segmentation (fets) initiative. (2020) (Accessed 28 May 2020). https://www.fets.ai Machine learning ledger orchestration for drug discovery. (2020). Accessed 28 May 2020. https://cordis.europa.eu/project/id/831472 Konečny`, J., McMahan, H. B., Ramage, D. & Richtárik, P. Federated optimization: Distributed machine learning for on-device intelligence. (2016). arXiv preprint arXiv:1610.02527 Lalitha, A., Kilinc, O. C., Javidi, T. & Koushanfar, F. Peer-to-peer federated learning on graphs. (2019). arXiv preprint arXiv:1901.11173 Li, T., Sahu, A. K., Zaheer, M., Sanjabi, M., Talwalkar, A. & Smith, V. Federated optimization in heterogeneous networks. (2018). arXiv preprint arXiv:1812.06127 Zhao, Y. et al. Federated learning with non-iid data. (2018). arxivabs/1806.00582 Li, X., Huang, K., Yang, W., Wang, S. & Zhang, Z. On the convergence of fedavg on non-IID data. (2020). https://openreview.net/forum?id=HJxNAnVtDS Wu, B. et al. P3sgd: patient privacy preserving SGD for regularizing deep CNNs in pathological image classification. In (pp. 2099–2108) (2019). Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Zhu, L., Liu, Z. & Han, S. Deep leakage from gradients. In Wallach, H. M. et al. (eds) , 14747–14756. (2019). Advances in Neural Information Processing Systems 32: Annual Conference on Neural Information Processing Systems http://papers.nips.cc/paper/9617-deep-leakage-from-gradients Wang, Z. et al. Beyond inferring class representatives: user-level privacy leakage from federated learning. In 2512–2520. (IEEE, Paris, France, 2019). 2019 {IEEE} Conferenceon Computer Communications, {INFOCOM} https://doi.org/10.1109/INFOCOM.2019.8737416 Hitaj, B., Ateniese, G. & Perez-Cruz, F. Deep models under the gan: information leakage from collaborative deep learning. In , CCS’17, 603–618 (Association for Computing Machinery, New York, NY, USA, 2017). Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security Ghorbani, A. & Zou, J. Data shapley: Equitable valuation of data for machine learning. In (pp. 2242-2251) (2019). International Conference on Machine Learning Acknowledgements Túto prácu podporilo UK Research and Innovation London Medical Imaging & Artificial Intelligence Centre for Value-Based Healthcare, Wellcome/EPSRC Centre for Medical Engineering (WT203148/Z/16/Z), Wellcome Flagship Programme (WT213038/Z/18/Z), Intramural Research Programme of the National Institutes of Health (NIH) Clinical Center, National Cancer Institute of the NIH under award number U01CA242871, National Institute of Neurological Disorders and Stroke of the NIH under award number R01NS042645, ako aj Helmholtz Initiative and Networking Fund (projekt “Trustworthy Federated Data Analytics”) a PRIME program nemeckej akademickej výmenné služby (DAAD) s finančnými prostriedkami z nemeckého federálneho ministerstva školstva a výskumu (B This paper is under CC by 4.0 Deed (Attribution 4.0 International) license. available on nature Tento papier je under CC by 4.0 Deed (Attribution 4.0 International) license. available on nature
