Aisthesis Medical Success Story: Test Before Invest – Fostering integration

This report presents a successful collaboration story between Aisthesis Medical and the European Digital Innovation Hub “smartHEALTH”.

Key Information

• Aisthesis Medical: Aisthesis Medical is a pioneering medtech startup transforming acute care with VIOSync, our AI-powered clinical decision-support system designed to anticipate sepsis and prevent multi-organ failure, enhancing patient outcomes and streamlining healthcare workflows. VIOSync proactively predicts sepsis, hypotension and multi-organ dysfunction syndrome, critical health risks affecting millions globally. Integrating seamlessly with wearable devices, bedside monitors and EHR systems, VIOSync creates an extensive digital twin for each patient, employing multimodal data for personalized, explainable predictive analytics.
• Technologies Used: Standard tools for data anonymisation and de-anonymisation, as well as known AI methodologies/tools (MONAI, Pytorch etc.)
• Collaboration Period: 01/12/2023 – 31/03/2024
• Services Provided: Consulting Services for scalability, portability and performance, services to respond to compatibility and interoperability demands.

The Challenge

Aisthesis Medical is at a fairly satisfactory level of digital maturity compared to other SMEs operating in Europe. However, in order to make further progress and maintain its competitiveness, it will need to address several needs that emerged during the assessment of its digital maturity, such as the following:

• Seeking funding
• Need to increase the level of technological readiness
• Harmonisation with current copyright regulations
• Lack of a universal data collection protocol
• Lack of continuous training of staff
• Reduction of the company’s environmental impact

The Analysis

Based on the responses to the questionnaire posed to the company, it would be to its advantage to focus on the following areas that will result in expanding its level of digital maturity:

• Copyright protection to safeguard the intellectual property of its software, as well as its financial and moral rights. In addition, the company will have to register, if it has not already done so, its trademark with the Trademark Directorate of the General Secretariat of Trade and Consumer Protection of the Ministry of Development & Investment.
• Information/Training of Staff: Aisthesis Medical staff should keep abreast of recent relevant technological developments through regular training programs, collaborations with academic and research institutions and participation in scientific conferences.
• Establish a Universal Data Collection Protocol to collect data from various institutions and open databases.
• Maturation of Services/Products through seamless monitoring of market trends in the sector in which the company operates and continuous adjustment of the company’s growth plan on an ongoing basis.

However, in addition to the general improvement in terms of increasing Aisthesis Medical‘s digital maturity indicators, there are also technical needs in terms of improving the technologies available and provided by the company.

The Solutions

Through the cooperation between Aisthesis Medical and the Technical University of Crete, the aim, in the first phase, was to provide Aisthesis Medical with a comprehensive picture of its level of digital maturity, compared to other SMEs operating in Europe, to be informed about the latest market trends and to receive suggestions for its further development. Moreover, in relation to its technical needs, specific guidelines were provided by the Technical University of Crete for the development of an optimal methodology to effectively achieve prediction in clinical cases such as sepsis based on modern algorithms and techniques. In this context, a specific action plan was proposed that includes the provision of consultancy services to enable the incorporation of algorithmic innovations to achieve effective and accurate diagnosis and prediction. These services are divided into two categories: a) Data Pre-processing, and b) Innovative Techniques in Machine/ Deep Learning.

The first service focused on managing the data in accordance with the current regulations, pre-processing them to reduce or even eliminate completely the noise introduced by the abnormal activity, while bringing out the desired information. It focuses on issues of data anonymization and de-noising, dealing with missing values and generating synthetic data, and merging data from multiple sources. The pre-processing process is a key step in ensuring the appropriateness of data before applying further analyses, ensuring the production of reliable results and interpretable in the field of medicine.

In the second stage/service, the plan focused on the detailed description of a set of methodologies leading to the desired prediction over inputs of a multimodal nature, as an extensive analysis was performed on current and innovative machine and deep learning algorithms, as well as automated machine learning, parameter and hyperparameter optimization, and specific interpretable artificial intelligence approaches and self-reference data techniques were considered.

The Implementation

Based on the needs of Aisthesis Medical, the Technical University of Crete delivered organised material containing a detailed action plan per consultancy service, based on scientific publications and the experience of the experts, while appropriate computational tools were also provided, where available. The total delivery time lasted four months, covering the period from 01/12/2023 to 31/03/2024. This period was divided into two sub-periods of two months each, for the completion of the two reports/deliverables, relating to “Data Pre-processing”, and “Innovative Techniques in Machine/Blind Learning”, with some meetings with Aisthesis Medical in between.

Data Pre-processing:

• Data anonymisation to remove the personal details of the patient concerned.
• Data de-noising using digital filtering, statistical analysis and decomposition of digital signals, deep learning techniques, etc.
• Dealing with Missing Values using statistical, engineering and deep learning methods to address the issue of incomplete information.
• Synthetic Data Generation to create new data to extend or improve the existing data set using time and/or frequency domain focused approaches, etc.
• Merging Data from multiple sources to produce more consistent and useful information than that provided by any single data source.

Innovative Engineering and Deep Learning Techniques:

• Algorithms/Techniques in Machine Learning and Deep Learning, where the most up-to-date algorithms and techniques in the field of machine learning and deep learning, related to related research in the field of interest, were analysed.
• Automated Machine Learning with the aim of automating the process of applying machine learning models, and can be exploited in the field of parameter and hyperparameter optimization.
• Interpretable Artificial Intelligence, where it is about artificial neural networks, machine learning models and algorithms designed so that their decisions are easily interpretable.
• Self-insight techniques, where data is characterized in an automated way without the supervision of an expert.

The Benefits

The benefits gained by the company through the cooperation with the Technical University of Crete, first of all, relate to the company’s information regarding its level of digital maturity compared to other SMEs in Europe, as well as the latest market trends. Additional benefits, which are mainly expected in the near future, relate to the following areas that directly concern the extension of its level of digital maturity:

• Copyright protection.
• Information/Training of Personnel: Through appropriate training programs, and in cooperation with academic and research institutions, as well as through participation in scientific conferences, AIMENT’s personnel will be kept up to date on technological developments in relation to the services provided and the products produced by the company.
• Establishment of a Universal Data Collection Protocol: the establishment of a universal data collection protocol will contribute to the secure and rapid collection of data from various institutions and open databases to achieve the collection of a wealth of data for analysis in a shorter period of time, thus pushing the company’s momentum.
• Maturation of services/products: Aisthesis Medical will achieve market consolidation through the creation of competitive products and services, by maintaining constant dedication to this goal and keeping abreast of current trends.

In addition, in relation to meeting the technical needs of Aisthesis Medical, specific guidelines have been created by the Technical University of Crete for the realization of the ultimate goal of effective and accurate prediction of sepsis. In this context, the company received consulting services in the form of two deliverables with detailed descriptions of modern data preprocessing techniques and methodologies and innovative machine and deep learning techniques.

Lessons Learned

There were no communication problems with the current recipient in relation to the description of the problem and the provision of services. Initially there was a lot of discussion as to a better understanding of the company’s needs and the specific problems that needed to be addressed by the TUC.

In terms of the main mutual benefit, through the services offered, common research interests emerged which resulted in joint research proposals and a broadening of the circle of partners.

Follow “smartHEALTH” on social media for the latest news and updates: Facebook, LinkedIn, Twitter.