font size
font size_s
font size_m
font size_l
Research Units

Bionic Medicine Research Unit

To Japanese version on this page

menu Research Outline Main Research Themes and Relevant Publications Principle Investigator of the group Staff and Research Focus

Main Research Themes and Relevant Publications

  1. Transneural treatment of cardiovascular disease
  2. Elaborate numerical models of biological systems: digital medicine
  3. Ultimate medicine: automated diagnosis and treatment

3. Ultimate Medicine: Automated Diagnosis and Treatment

Basic strategy of automated treatment

Acute left heart failure is a disease that can be effectively treated by bionic cardiology. Treatment for acute left heart failure requires specialized knowledge. When the pathological condition is identified, it is possible to automate the treatment. For a patient with hemodynamic instability, a medical specialist places an arterial catheter or Swan Ganz catheter to monitor arterial pressure, cardiac output, right atrial pressure and left ventricular pressure. From these data, the cause of low blood pressure or low cardiac output is diagnosed, and necessary treatment conducted. In other words, if the specialist interprets the measured data according to circulatory physiology, then the pathological condition can be accurately defined. Based on the diagnosis, treatment method (drug) is selected and treatment is initiated. Of course the treatment content has to be changed continuously with time, according to the response.

If the diagnosis and treatment process of the medical specialist can be automated, automated treatment may become possible. For this purpose, it is most important to establish a biological model representing the circulation. If the model is too complicated, it will not be possible to fulfill all the parameters in the model from the measured data. If the model is too simple, it will not be useful for diagnosing the pathology. We expanded the circulation model of Guyton, and developed an extended Guyton’s model that can be used for accurate evaluation also in left heart failure.

Block Diagram of Automated Treatment

This figure shows the outline of the system that automatically corrects the impaired circulatory state of acute left heart failure. The hemodynamic data are obtained from the patient and input into the extended Guyton’s model. Using the extended Guyton’s model, any problem with the pump function (SL), arterial resistance (R) and effective blood volume (V) is identified. Based on the evaluation result, treatment is given by: infusing dobutamine (DOB) if the pump function is insufficient, infusing nitroprusside (SNP) if vascular resistant is abnormal, or infusing dextran (DXT) solution or diuretic if effective blood volume is a problem. With continuous feedback of the hemodynamic parameters, treatment is conducted automatically.

Results of automated treatment

This figure shows the performance of an automated treatment system in a dog model of acute left heart failure. Left heart failure is induced by myocardial infarction. Following experimental production of myocardial infarction, blood pressure decreases by around 20 mmHg, cardiac output decreases by 40 ml/kg/min, and left atrial pressure increases by around 12 mmHg. In other words, the animal is in a state of hypotension, low cardiac output and pulmonary congestion. When automated treatment is started, the system simultaneously administers dobutamine, nitroprusside and dextran solution. Within approximately 10 minutes, the pump function, vascular resistance and effective blood volume are all normalized. As a result, hemodynamic parameters are also normalized. If this treatment is conducted by man, it will be difficult to start administering multiple drugs simultaneously, even though the person may be highly experienced. On the other hand, if an appropriate disease model is available, this kind of highly efficient treatment will become possible.

The fact that this system functions effectively in an animal model suggests that automated treatment for acute left heart failure is possible. However, many problems have to be overcome before this system can be put to clinical use. There is an urgent need to develop methods to acquire reliable hemodynamic data. Especially, since measurements of blood pressure and cardiac output are influenced by noises, stable recording of these basic data is extremely important. Successful solution of these issues will pave the way for automated treatment of diseases with established pathophysiology.

Relevant publications
  1. Uemura K, Kamiya A, Hidaka I, Kawada T, Shimizu S, Shishido T, Yoshizawa M, Sugimachi M, Sunagawa K. Automated drug delivery system to control systemic arterial pressure, cardiac output, and left heart filling pressure in acute decompensated heart failure. J Appl Physiol 100: 1278-1286, 2006.
  2. Uemura K, Kawada T, Kamiya A, Aiba T, Hidaka I, Sunagawa K, Sugimachi M. Prediction of circulatory equilibrium in response to changes in stressed blood volume. Am J Physiol Heart Circ Physiol 289: H301-H307, 2005.
  3. Kashihara K, Kawada T, Uemura K, Sugimachi M, Sunagawa K. Adaptive predictive control of arterial blood pressure based on a neural network during acute hypotension. Ann Biomed Eng 32: 1365-1383, 2004.
  4. Uemura K, Sugimachi M, Kawada T, Kamiya A, Jin Y, Kashihara K, Sunagawa K. A novel framework of circulatory equilibrium. Am J Physiol Heart Circ Physiol 286: H2376-2385, 2004.

  1. Transneural treatment of cardiovascular disease
  2. Elaborate numerical models of biological systems: digital medicine
  3. Ultimate medicine: automated diagnosis and treatment

▲Page Top