HOME ABOUT GAME CHANGERS GET THINGS DONE TRACES CONTACT

ABOUT GAME CHANGERS GET THINGS DONE TRACES CONTACT SEARCH

Aicebear
HEALTHCARE SYSTEMS IMPLEMENTATIONS

HEALTHCARE SYSTEMS IMPLEMENTATIONS:

WE GET THINGS DONE

Who we are

We are an independent provider of research and advisory and implementation services for the life sciences industry, health insurences, caregivers and public administrations.

Our foundation

Aicebear was founded in response to a growing demand for better analysis of environmental impact issues, meta-design of large infrastructure projects and efficient management of public administrations. Healthcare systems has become our most important issue over that time.

Why you should choose us

We are strong in interlinking different topics and perspectives at different scales. We know the healthcare system in depth, from the legal, operative, economic and political side. You keep your independency - as an indepent consultant we do not take profit from your decisions.

EMERGING CHALLENGES

GAME CHANGERS IN HEALTHCARE

Precision medicine

The adaptation of a therapy to the individual patient has always been a task of clinical medicine. But now new diagnostic and therapeutic possibilities are constantly being added. They are changing the entire interplay between research and clinical treatment and reimboursement, and the availability of data is becoming a strategic cornerstone.

Data as a strategic asset

Longitudinal data is needed for personalized medicine. Medical interoperability must be ensured across medical specialties and medical treatment sites. This are core competences and not pieces of software. Healthcare companies that succeed in this tasks gain a strategic advantage.

AI has come to stay

AI methods are older than you think. But it was the availability of data and computing power that gave them the boost. " No data, no AI - bad data, bad AI" points to the necessary boundary conditions. Implementation in a clinical setting requires careful consideration of the training database, the explanation of results, the handling of uncertainty, the variability of results due to ongoing upgrades, and the price/benefit ratio.

Changing awareness of personal identity and integrity

Real world data (RWD) are potentially the far larger data pool than clinical trial data. Resistance to the long-standing practice of unasked-for data use is forming on the basis of health data. Fair, efficient and transparent consenting processes provide a sustainable comparative competitive advantage.

Paradigm shift in reimbursement

The current reimbursement system is currently only to a limited extent able to keep pace with medical progress. In order to make modern therapies available to patients within a reasonable period of time, the evidence base must also increase in strength, speed and robust methods. The opportunities and limitations are well known from the AI world. In medical applications, however, much higher requirements apply.

Market entry of digital tech giants

The most important changes taking place in medicine are data driven. The digital tech giants have the capability for both, data integration over long pathways and the abandonment of market activities that are not attractive. In the scalable market segments, a positioning competition is taking place with the major companies in the healthcare market, the outcome of which is still open.

Key driver of change in healthcare

Aicebear is an independent provider of research and advisory and implementation services for the life sciences industry, health insurences, caregivers and public administrations.

MEET YOUR NEEDS

Healthcare Systems Implementations - we get the things done

We follow our customers on different journeys to meet their need: from road map to implementation, from scientific overview to patient reported outcome, from clinical data capture to real world evidence

In the end, we want to see ongoing clinical and clinical trial processes that are efficient and generate resilient results.

IMPLEMENTATIONS OF HEALTHCARE STRATEGIES

We implement strategies focused on personalized medicine, AI learning in clinic and clinical trials, data-driven business and reimbursement models

Formulation of strategic principles, indicators for controlling

political environment analysis, alignment with other objectives

Implementation of guidelines, controlling information cockpits

Strategic data partnerships

IMPLEMENTATIONS OF CLINICAL AI LEARNING

We implement processes in clinical and clinical trial environments that are ready to generate and use AI learning.

Clarification of the learning perspective

Strategic data partnerships, strategic platform partnerships

Formation of appropriate AI teams

Alignment of AI models, predictive analytics, data flow

AI model training and testing

Implementing the AI model in clinical practice, adapting clinical processes, dealing with uncertainty.

IMPLEMENTATIONS OF HEALTH DATA GROWTH

We implement clinical environments and clinical trial environments for sustainable growth of real-world data.

Ensure medical interoperability from the first data entry. Simplify and eliminate multiple data entry by medical staff.

Quality-driven ontology management.

Establishment of sustainable data repositories, establishment of FAIR principles.

Efficient data provisioning and ELT processes in compliance with regulatory requirements for clinical and CT AI applications

IMPLEMENTATIONS OF EVIDENCE GENERATING PROCESSES

We implement seamless process chains that generate evidence.

Documentations for and negotiations with authorities.

Numerical analytics, computer-aided modeling.

Quality-assured processes, in-house and cross-institutional

PITCH PATCHES OF OUR WORK

Click on the images to get more informations

Oncology is a highly complex field in which a wide variety of subfields are intertwined. Each subfield in turn is characterized by a high degree of technology. Differentiated scoping is essentiel for every clinical AI project

Clinical AI learning is focused on the patient's condition. Patient condition data is the most important data. However, this data category also has the greatest heterogeneity in coverage and quality. Processes to offload medical and nursing staff are an important success factor for clinical AI learning.

AI for Quality Assurance intends to show We implement processes in clinical and clinical trial environments that are ready to generate and use AI learning. the probabilities of therapeutic success and undesirable side effects for given analyses and possible forms of therapy. These will be taken into account when deciding on the therapy.

AI for Research intends to propose new forms of therapy through intelligent optimization. This approach must always be integrated into the legal processes on ethics and data privacy.

Interpretive approaches are changing. Long-term data strategies are required. Nevertheless, thoughtful scoping is required because of the wide variety.

Real world data (RWD) and real world evidence (RWE) means that in clinical practice, the cycle of observation, data collection, and scientific evaluation and discussion is closed so well that it becomes evidence-ready. This can only be achieved if efficiency in clinical operations is taken into account too.

In personalized medicine, the study population is getting smaller. It is imperative that this be compensated for by longer observation chains. For applied real world projects, the clarification of the collaboration map is an important entry point.

Data structures and work processes must be aligned with personalized medicine. The Medical Data Set is the methodological backbone for ensuring medical interoperability at all times.

Artificial intelligence can be used in a variety of situations, but not in all of them at the same time. <br> AI - and especially deep learning - owes its success to the ability to classify situations with an acceptable error rate. A temporally fast sequence of classifications leads to process control. The acceptable error rate depends on the subject.<br> A wide field is also opened by AI to be able to map knowledge into a complex, mathematical computer model that forms the quantified prior to a question. The comparison with situation-related data enables an evidence evaluation as posterior.

Markov Chain Monte Carlo model with explicit drift field

Image segmentation: AI algorithm aproaching iteratively the target zones

The applied use of complex functions has been largely displaced by numerical methods.

The construction of AI models is still distinctly heuristic. Model updates also require clear release management in the clinic since updated models may exhibit altered behavior.

Each category of data requires its own processing.

Efficient clinical AI learning needs efficient pipelines.

Each artificial neural network is a computational prediction/correction scheme: (1) guessed model parameters are used to 'predict' the output of the labeled data, (2) the error between predicted and labeled output is evaluated, (3) the model parameters are corrected based on heuristic error back-propagation principles. <br><br> For some sophisticated model types, the structure of the model is also adjusted and optimized. Explainability of AI models is one of the main tasks in AI development.

CONNECTING TRACES

AICEBEAR IMPLEMENTATION SERVICES

Aicebear is an independent provider of research and advisory and implementation services for the life sciences industry, health insurences, caregivers and public administrations.

AICEBEAR LEAD

Peter Schuhmacher, PhD
Head of Aicebear Healthcare Systems Implementations

Peter Schuhmacher looks back on a long experience in healthcare from different perspectives and has led far-reaching change projects at hospitals, public health administrations, pharma and medtech companies and insurers. For about 10 years he was an elected member of a parliament and as a former healthcare politician he knows the political processes in healthcare in depth.

Peter Schuhmacher holds a degree from ETH Zurich in Management, Technology and Economics (MTEC ETH) and a PhD degree in Computational Fluid Dynamics and Atmospheric Sciences (Dr. sc. nat. ETH). He applies his knowledge of handling data and building and validating complex computer models for machine learning and artificial intelligence in medical applications.
Peter Schuhmacher is a lecturer at two universities.