Blockchain Papers

Sean T. Manion

What is the brain that it can understand science?What is science that it can understand the brain?These two basic questions (with homage to Warren McCulloch in the framing) have guided my career, aiming to understand the brain and an effort to understand science. This journey has taken me from academic lab work to clinical research oversight and government policy to the emerging health & science technology industry and back to academia. It has now led me to co-lead, along with Dr. Jennifer Lovejoy of the Institute of Systems Biology, this section of Frontiers in Systems Biology -Systems Concepts, Theory and Policy in Biology and Medicine. Our journal Chief Editor, Dr. Yoram Vodovotz, has laid out the overarching vision for this and the other sections (Vodovotz, 2021). This Grand Challenge is an effort to add another layer of detail to the portions of that broad scope contained in our Systems Concepts section (Lovejoy, 2024).Systems biology and systems medicine have roots going back to at least World War II, when biologists and physiologists were recruited into the war effort in the United States and Britain, trained in computational approaches, and joined with engineers and mathematicians to solve complex problems with communications, radar, anti-aircraft guns and more (Churchill, 1949). This alignment led to the foundation of the field of cybernetics and the related Macy Conferences in the U.S. postwar, while in Britain, "This coalescing of biological, engineering, and mathematics frameworks would continue to great effect a few years later as the Ratio Club," (Husbands, 2008). In the decades that followed, this robust milieu of ideas would foster the development of everything from general systems theory and information theory to artificial intelligence (AI) and cognitive science (Pickering, 2010). Despite this early alignment, it would be decades before systems biology and systems medicine arose as formal fields of inquiry (Green, 2017).Science has arguably been the most effective way of generating and validating new knowledge for the past few centuries. New technologies and computing approaches now provide us with novel tools to accelerate this process. While early work is being done to explore the use of these new tools for science, these have been limited in success to real-world application to detailed aspects of biology and medicine (McCoy, 2024). A comprehensive conceptual framework may be a more effective way to realize the value of technology in accelerating science. Modern science is not a simple holistic process, but an amalgam of processes and interests that have accumulated over centuries.By analyzing this system of science, we can better synthesize a new approach to using the array of emerging technologies now available. This will require us to revisit the current human and institutional processes that govern the creation of new scientific knowledge. A human and machine hybrid approach, aligned with a governance in the classic cybernetic style (i.e. control and communication in humans and machines), may allow us to optimize our scientific efforts and advance knowledge for the betterment of all of humanity.There has been much excitement about the potential of emerging technologies applied to science in recent yearsfrom AI to applications of blockchain technologies and web3 applied as decentralized science (DeSci) (Weidener, 2024). In these nascent efforts there has often been an oversimplification of science in order to capture technical requirements to automate or simulate biomedical research.Science is not done by a single person or organization. Science embodies the contribution of multiple individualswhose brains are themselves collections of dozens of subsystems (Kirby, 2024)processed through a series of refinement and testing. The results of these are moved through a longitudinal process of validation, contextual framing against prior accumulated knowledge, and consensus determination of evidence level and confidence in the results. Only then does this new knowledge contribute to the body of generalized knowledge we applied to the real world.Creating a new technology-accelerated knowledge system for biomedical science -what I'm calling here Scientia Machinamay be best approached through first articulating the conceptual and epistemological framework of the current system of biomedical science as it moves from data to information to evidence to knowledge and its application. Along the way it passes through layers of trust and is eventually captured in the artifacts of biomedical science we have come to rely on and expect. For applications of emerging technologysuch as the automated complex information processing of AI and the automated trust and governance of blockchainto be most beneficial to science, we should use them to systematically augment and accelerate these processes and creation of the artifacts of science while maintaining or improving the basic conceptual framework of biomedical knowledge discovery and implementation. Eventually parts of the current system may be sundowned leading to an even greater acceleration of science.This Scientia Machina framework starts with identifying key layers of trust in the biomedical bench to bedside process of evidence based medicine. Here I have proposed five layers of trust along with examples of their current artifacts and processes, plus potential approaches to augmenting these with technology and related adjustment to the current workflow (Figure 1).Data Layer -Data are collected in experimental and/or clinical context, often based on specific methodology. The principal investigator (PI) and team, along with the equipment and techniques used, are trusted to produce and capture explainable and reproducible data. This layer is only sometimes made transparent and rarely validated.Future of Data -Data are verifiable through trackable provenance and alignment with related metadata (e.g. demographics, treatment delivery details, device and equipment specifications, etc.). Data can be accessed for querying and algorithm training without moving, copying or exposing the data.Information Layer -Data are combined in datasets with contextual meta-data (e.g. demographics of research participants). The PI and team are trusted to compile, store and manage this data. It is increasingly becoming requested by funders and publishers to be made available. Some programs promote dataset sharing through centralized repositories or direct PI to PI contact.Future of Information -Data confidence fabrics allow sorting combined datasets based on confidence levels for each data point related to their associated metadata, with deployable programming to temporarily convert non-standard data into a calculable or trainable standard.Evidence Layer -Analysis of the datasets and testing hypotheses produces results that interpreted as findings. These are presented as novel assertion, backed by the data and methods, and put into the context of previously identified findings in the field in the form of a manuscript submitted for peerreview. The journal editors and peer-reviewers are trusted to confirm the assertions are supported by the evidence, fit (or convincingly contradict) previously established knowledge in the field.Future of Evidence -Swarm approach, i.e. networked, auditable crowd-sourcing, to peer review with a wider array of contributors with inputs weighted based on preset governance and continuous crowd feedback for nearer to real-time review with broader, multi-discipline input.Knowledge Layer -Combined sets of published articles are reviewed by a group of experts against certain criteria to answer specific questions about the state of evidence in the field as systematic reviews and meta-analyses to provide the most up-to-date knowledge in the specific area of focus.The groups of authors along with editors and peer-reviewers of those systematic reviews and metaanalyses are trusted to have executed and validated, respectively, a thorough and sound assessment of the evidence for the area in question to provide new knowledge.Future of Knowledge -Swarm approach (see above) to systematic review with network on demand request for new or updated reviews of existing evidence along with evidence threshold signals (i.e. sufficient new evidence in a particular areas prompts new or updated systematic review).Applied Knowledge Layer -Applications of knowledge can come in various forms, including pharmaceuticals, devices and procedures. The application of knowledge is periodically assessed for incorporation into clinical practice guidelines (CPG) and similar clinical guidance documents. The CPG group is trusted to have found and appropriately graded all of the available evidence and refined knowledge on a topic area to best inform clinicians how to address the area optimally.Future of Knowledge Application -Networked clinical practice guideline wiki (collaboratively edited living document) allowing for continuous, network refereed input and update of new knowledge.Each of these layers and their future states can be augmented, enhanced, accelerated and potentially replaced with appropriate applications of an array of automated processing and trust technologies. Additional administrative areas of biomedical research such as gap analysis, funding, regulatory review and more can be similarly improved.The call to action for this Grand Challenge is to: a) Consider the core elements of what we need to maintain and continue to elevate from our past and current successful biomedical research and knowledge translation effort, along with areas where those efforts have been flawed, corrupt or unsuccessful. b) Critique (and adjust or replace as needed) the Scientia Machina framework proposed here as the backbone for the layers of trust that are the core elements to be maintained as we continue to bring new technologies into biomedical research to accelerate and improve science. c) Capture and assess those current pilots to apply emerging technology -especially within AI (complex information processing) and DeSci (automated governance, auditing and/or incentivization) as umbrella categories for these effortsand place them in the context of a broader framework of what we are trying to achieve with biomedical research. d) Conceptualize gaps in our current efforts along with bridges from the current status quo to the desired future that may give us a better chance of success at transformational change to the systems of biomedical research and knowledge translation. e) Communicate all aspects of the above areas in appropriate venues of biology, medicine, technology and policy. This includes formal submissions of manuscript on any related topics to this journal section and its partnered sections as appropriate.This proposed conceptual framework is merely a jumping off point for broader consideration of how to maintain the core elements of the trust we have imbued in biomedical research as we continue to explore applications of emerging technology to improve its quality, manage its costs, and accelerate its contribution to the health and well-being of everyone. In the not so distant future, it is conceivable that we may be able to make all available relevant data on a topic or a patient accessible to any researcher to make AI-augmented and blockchain-audited hypothesis testing to provide near realtime, peer-validated contributions to evidence-based medicine. This could allow clinicians to query and access this near real-time evidence as part of compressing the 17 years it takes to go from bench to bedside by a factor of 10,000xgiving us new, actionable evidence-based precision medicine for patients in under a day. This future is within reach. Aligning behind a shared framework like Scientia Machina can bring it into our reality even faster. Better science. Cheaper research. Faster Miracles.STM is the sole author, having conceived, written, and edited this manuscript.

Lukas Weidener, Logan Bishop-Currey, Marie-Madeleine Fürbeth

No abstract is available for this record.

Arvind Muthukrishnan

We, as oral physicians, do an extensive literature search to find out the best diagnostic investigation or best therapeutic option for a disease that we encounter in clinical practice. But do we pause for a moment and look at the literature to see whether the researchers have formulated the most specific focused research question in their works? Failure in this first step to formulate the most appropriate research question would subsequently affect the entire research process. In the arena of research, there are numerous players whose expectations need to be taken into consideration before a good research question is formulated. Foremost among them are our clients – the patients for whom we are doing research. We need to initially ascertain what the patient is expecting from us as a treatment for his ailment and what changes he is expecting in his Quality of Life. The second is our (investigator) views about the intended endpoints for a particular problem we are trying to solve. Third, unfortunately, we often neglect to ascertain the opinions of our co-workers regarding our proposed work. They might have a different perspective on our thought process. Finally, if we have agencies to fund our work, we need to ascertain that they are on board with us to support our research, agreeing with the problem for which we are trying to find a solution. If we analyze from the point of each stake holder mentioned above, each view is logical and it becomes a herculean task for the investigator to formulate and evolve the most appropriate research question. As a guide, researchers could follow the following criteria[1]: - The most appropriate question which is important to patient well-being - The most appropriate question relevant to our knowledge levels - The most appropriate question that can be addressed in a specific time frame - The most appropriate question that would interest you, your team and your patients the most - The most appropriate question that is likely to repeatedly present itself in your practice. As a piece of advice, I would state that, the scientific community always recognizes and appreciates researchers who address problems of diseases faced in their own local community and not some diseases which are very rare in a particular geographic setting. A researcher should always formulate an “Answerable” question. Here, a large broad topic needs to be split into smaller manageable units, which can then be addressed through a standardized protocol.[2] A young researcher by nature would be too ambitious to make a path breaking research to solve all problems. But, seldom does it happen. If your research question is too wide, you end up, lacking rigor in methodology. If your question lacks focus, it is almost close to impossible to replace it with another question once your work is commenced or completed. All questions you want to answer should follow the PICO format. This format is suggested because it helps you to specifically narrow down, refine and formulate your question to address one specific problem. Though PICO format is meant to address interventions, other research questions (diagnostic, prognostic, patient expectations) can be reframed to follow PICO format. The next key element of good scientific research is choosing the appropriate study design. We need to mandatorily sit with trained “BIOSTATISTICIANS”, explain our intended work and zero in on the most appropriate study design. Researchers need to do a thorough data search to find out how a similar question was designed and studied. As you navigate this process you will find your primary research question getting more and more focused which would help you to reframe your PICO components. The above exercise is definitely time consuming and test your patience, but remember – the extra time spent in this stage would save you many hours later if you proceed with an irrelevant research question, inappropriate study design or work on a topic which has already been exhaustively analyzed. Researchers feel quantitative research where you can categorize all parameters with numerical data is superior to qualitative research. But we oral physicians deal with a plethora of disorders especially oro-facial pain where the need to address emotions, feelings are more important than aiming at numerical value changes as a proof of your successful patient management. Qualitative research design are better suited address research questions dealing with feelings and emotions.[3] The NHMRC evidence hierarchy categorizes the most appropriate study design for specific type of research questions – interventional, diagnostic accuracy, prognosis, etiology, and screening.[4] On the contrary, in qualitative research, it is advisable to follow a typical practice-based approach to analyze the data through either a case study, grounded theory, phenomenology, ethnography,ethno methodology and narrative research. To conclude, researchers need to focus on the two essential pillars – the most appropriate answerable focused clinical question and choosing the most appropriate study design to specifically answer the formulated question. However well a question is framed and study is designed, the results will open the door for a next question to be investigated. That is how science grows and progresses!!