Top > Research > Human Modeling Team > Digital Patient

research

This research project was finished.

Digital Patient - Response Model & Pain Detection

When patients go under dental and/or otolaryngological treatments and surgical procedures, they tend to be nervous and uneasy. One of the reasons is that patients are aware of vibrations, noise, and pain that can not be suppressed by anesthesia. In this research, with the cooperation of patients who are undergoing actual surgery...

Cardiorespiratory Response Model for Pain and Stress Detection during Endoscopic Sinus Surgery under Local Anesthesia

[PDF: 492KB]

Surgical Procedure - Response Model for Surgical Training

At the National Institute of Advanced Industrial Science and Technology (AIST), the Institute for Human Science and Biomedical Engineering (HSBE) has been the center for developing a procedural training model for endoscopic surgeries. By using this model, the training for operational techniques of surgical tools, such as endoscopes and forceps, became possible.

Digital Human Research Center is studying the modeling of various interactions between surgeons and patients during surgery.

  1. The reproducible model of patients' reactions to surgical operations:
    reproduces vital patients' reactions, such as pain occurrence and fluctuation of blood pressure, to surgical operations, including removal of polyps and mucous membrane, destruction of bony walls, and aspiration.
  2. Action-Response pattern model of how experienced surgeons respond to patients' reactions:
    timing of when to take preventive measures (supplemental anesthesia) for pain, and the action-response pattern model for actions taken after pain occurs (length of rest, etc.).

By reproducing the possible patients' reactions that occur during surgery, we are developing a skill acquisition system that carries surgical procedures forward while keeping a the patient at ease.

Endoscopic Sinus Surgery

Figure 1: The scene at an actual surgical procedure (endoscope held in the left hand; a forceps in the right hand)
Figure 2: Example of heart rate change during an actual surgical procedure. Time under anesthesia/rest (gray) and operating time (white) is alternated. Significant changes can be seen in heart rate when pain occurs.

Findings of Study

The Measurement of Progress of Surgical Procedure and Patients' reactions:
An endoscopic video, an operating room video, and a bio monitor are used to measure the progress of a surgical procedure. As a record of progress in surgical procedures, types of operations, positions, tools and quantity of operations, conversation between surgeon, patient, and nurses (complaint of pains and instruction for administering supplemental anesthesia, etc.) are logged. As an index of a patient's mental stress, heart rate, blood pressure level, perspiration caused by mental anxiety, and breathing movements is recorded.

Results of Analysis:
At this time, we have conducted analysis based on the measurement data of 6 patients.

Figure 3 shows that when pain did not exist, patients' reactions varied depending on the type of operation. More specifically, when a forceps was inserted into the nasal cavities and the opening procedure started, patients were on guard. Consequently, their breathing was repressed, causing a decrease in heart rate. As the surgeon proceeded to aspiration and packing, breathing movements and heart rate were both found to increase.

On the other hand, when pain occurred (Figure 4), heart rate increased during the opening process. Also, as for frequency of pain occurrence based on regions, the sensation of pressure and pain occurred more both at the inside of middle nasal concha (MT-IS) and around the superior nasal concha (SRT) - perhaps since the surgical tools had difficulty reaching inside. Frequency of pain occurrence was highest at the maxillary sinus (MS) as it is located in proximity to the dental nerve. At ethmoidal sinuses (ES), the occurrence of pain was seen particularly while operating in the lower areas.

Figure 3: Patient response patterns (no pain)
Figure 4: Patient response patterns (feeling pain)
Figure 5: Frequency of pain occurrence based on regions

Modeling: Reproducible Model of Patients' Responses

Based on the results of the analysis, we constructed a reproducible model of patients' reactions toward surgical operations (Figure 6).

This model was developed using a probabilistic model, Bayesian network, and BayoNet, the Bayesian network construction system (developed by AIST; distributed by Mathematical System, Inc.).

We are also making a prototype of a patient simulator that works together with aforementioned model for surgical training (Figure 7).

Figure 6: Reproducible model of patients' reactions
Figure 7: A prototype of patient simulator

Top ∆