an ecosystem with integrated clinical decision support tools, artificial intelligence applications and novel medical devices, for use in austere environments with limited resources
Learn moreCritical care and trauma management requires significant expertise and special diagnostic and therapeutic devices. Critical care is a data intense speciality where multi-attribute decision making is the key for therapeutic decisions, which are taken under stress and time-pressure. This makes management of critically ill patient prone to error, even for an experienced intensivist. In extraordinary situations and austere environments this problem is pronounced.
In all situations the major problem is the lack of experienced medical staff. We aim to solve this problem by developing an ecosystem composed of clinical decision support tools, artificial intelligence applications and novel medical devices.
01
In disasters and pandemics number of critically injured/ill patients increases suddenly and logarithmically. This increase is out of proportion compared to available resources. Mobilization of additional resources is often delayed and insufficient.
02
Military conflicts and combat situations require special expertise and equipment. Nature of combat may delay evacuation and advanced care of the critically injured. Within the context of modern combat, the troops need to be self sufficient in advanced care of the injured soldier. Harptek Ecosystem aims to ensure efficient advanced care of the wounded at as early as tactical field care (TFC), and tactical evacuation care (TEC) stages of TCCC.
03
International military conflicts and potential food and water shortages make refugee crises a real threat for all countries. Large scale refugee crises is a major health care crises where patients are treated with limited resources. Humanitarian missions also will benefit from our solutions.
04
Space missions and polar expeditions bring additional problems like care of patient in zero gravity and within a significantly limited area, in an environment exposed to solar radiation. This requires design and development of special equipment that can operate reliably in zero gravity and robust against the EMF imposed by solar radiation. The major problem is limitations in number of experienced medical staff recruited in these missions.
05
Increasing costs of treatment and limitations in staffing imposes a new health care crises for many countries. Even the highly industrialized nations are prone to suffer from the effects of this crises. Many of our solutions are effectively applicable to solve these problems.
We clearly know the problem and its nature.
Prof. Kılıç, founder and main researcher of Harptek, is an experienced surgeon and critical care specialist, who has significant experience in disasters. His almost 30 years long journey is the basis of our vision. During this journey he had gained significant knowledge and experience on many disciplines of engineering. This enables us to view the problems from both medical and engineering perspectives at the same time. Being able to recognize the extent of the problem and limitations of the potential solutions is one of the major aspects that makes our team unique.
We know and respect the complexity and adaptive nature of the problems we face with. Despite the common tendency to use the most advanced algorithms for decision support and artificial intelligence solutions, most of them are not reliably applicable to problem solving in medicine. We select the most appropriate AI algorithm for the specific problem, because there is no one size fits all solution applicable to all problems. These include neural networks, fuzzy logic, expert systems, deep learning and combinations of these as appropriate.
A talk with Prof. Kılıç on life, technology and entrepreneurship...Our ultimate goal is to develop completely autonomous medical devices for patient care, which are robust and reliable. This goal requires a clear understanding of clinical problems, treatment algorithms and response to treatment. Believing to be able to solve clinical problems by simply feeding a deep learning algorithm with data is an illusion. Disease states and treatment processes are complex adaptive systems and medicine is an art.
Our short term goal is to develop a reliable and robust ecosystem for care of critically injured/ill patients in austere environments with limited resources, both in respect to medical devices and experienced personnel.
The products and prototypes are as follows. Some of the softwares are listed are developed for internal use and are only available to selected partners we collaborate in R&D projects.
01
A paradigm shift is coming in the care of injured/critically ill patients in austere environments with limited resources. A major adaptable medical counter-measure, applicable to TCCC, TECC, US&R, and critical care in the field for disaster response, pandemics, and combat settings is on the horizon. Stay tuned for updates on this groundbreaking development.
02
Novel critical care monitor with real time entropy monitoring capability as part of multiparameter vital sign variability analysis. Uses advanced algorithms to calculate entropy in short term vital sign time series and frequency-domain variables, time-domanin variables, nonlinear analysis, geometric analysis in long term data series in real time. Also monitors entropy of entropy. R&D phases partially funded by KOSGEB.
03
Pocket size transport monitor (ECG and pulse oximeter only) with decision support capabilities. Fully embedded, fast bootup. Wireless connectivity.
04
Monitor (ECG and pulse oximeter only) and decision support system specially designed for TCCC. Fully embedded, fast bootup. Wireless and wired connectivity. Long battery life. Guides treatment of wounded at all levels of TCCC, including evacuation.
05
Tactical personal status monitor prototype. Monitors ECG, oxygen saturation and body temperature, and communicates over BLE. Can be integrated with central monitors, transport monitors, and other wearable systems. Can monitor health and physical performance of combatant. Can be used in team activities.
06
The most comprehensive critical care information system in the world. Integrated critical care database having capability of disease severity/mortality prediction and multiple organ failure scoring. Provides context sensitive decision support and tools for quality improvement and benchmarking. Regular performance measurements and quality improvement reports can be provided if needed. A.I support and medical device integration is optional.
Access Online07
coLearn is an online platform for collaborative learning on unlimited diversity of subjects. Provides many tools that enhance distant learning and is highly extansible. Access through mobile apps and web browser. Collaborate in education and discussions anytime, anywhere, as an educator or a learner.
Access Online08
Most comprehensive diagnostic and decision support application available for evaluation of arterial blood gas and electrolyte disturbances. Provides a detailed diagnosis using both algorithmic approach and strong ion difference. Makes detailed differential diagnosis and provides user with practical knowledge related to pathophysiology, clinical findings and treatment of the problem. Generates a detailed plan for further diagnosis and treatment of the specific patient. Can also be used as an educational tool. Mobile, desktop and web based applications available.
09
Mainly developed for field and TCCC settings. Uses expert systems and fuzzy logic interpreters for generating clinical decision support for management of severely injured trauma patient.
10
Mechanical ventilator simulator that can simulate basic and advanced modes. Can be used on desktop, mobile and online platforms. Can be integrated as GUI for new mechanical ventilators.
11
Advanced node editor for designing and testing mechanical ventilation algorithms. Components are combined in a Lego pattern. Respiratory dynamics can be simulated as needed. Inputs and outputs of created ventilator algorithm can be further connected to additional python libraries of artificial intelligence and machine learning for algorithm development and testing. [Internal use]
12
Performs detailed calculations related to respiratory dynamics an mechanical ventilation. Can both use one compartment and two compartment models. Calculated values are required for advanced mechanical ventilation modes.
13
Generates REST API code for php and mysql in Model-View-Controller pattern. Additionally generates code for related mobile applications. [Internal use]
14
Python code generator for AI applications with neural networks, fuzzy logic and neuro-fuzzy applications. Can generate code for multilayer fuzzy systems. Has a visual rule editor for fuzzy systems. Development is ongoing for C++ code generation. [Internal use]
15
Software for generating bare metal code for commonly used MCUs and boards. One can generate embedded code at the level of registers visually. [Internal use]
16
Novel mobile app with multilingual intuitive interface and novel controls that calculates and evaluates FOUR score, IMPACT model, CRASH model and GCS score. You can try a limited version online.
Try Online17
Multiplatform mobile app for calculation of body mass index and evaluation of diagnosis. Intuitive multilingual interface has novel controls that allow access to all functions with one hand ergonomically. Presented here as an example of our capabilities in mobile development. Available in PlayStore.
Watch Video18
Educational games on complex areas of critical care and trauma management. One example is the ABG Game where user predicts primary and secondary diagnosis of a randomly generated arterial blood gase data and try to win the app. Slide the red circle to randomize, press the central button to predict diagnosis. See whether you can predict all the details Harptek ABG Algorithm diagnoses. A detailed summary of how the diagnosis is reached is also presented for educational purposes.
19
We work on artificial intelligence tools related to clinical decision support and autonomous medical devices. Mostly based on neural networks and fuzzy logic, these tools include mechanical ventilator algorithms, weaning algorithms, shock management, trauma management, electrolyte and acid-base diagnosis and treatment. You can access an early version of our weaning algortihm.
20
Geolocation project aiming coordination of disaster response and TCCC. Disaster response includes coordination of SAR (search and rescue) and medical reponse. Disaster response and TCCC also can coordinate evacuation and transfer of wounded. Planned to integrate RFID tagging for appropriate follow-up of victims rescued from buildings and transferred for further treatment. Will be available as mobile app, desktop and web based.
The primary problems that we are working on are