Q)=question A)=answer NQ)=new question for questionnaire NQA)=response to possible new question LL)=lesson learned BS)=brain storm thought C)=change made to simulator SC)=subject comment NC)=needed change!! OB)=observation by proctor during testing



4/10/97 Q:Chris Airola A:Dave Heskamp


Q) Need for manual trigger to toggle scoring timer to allow time for

subject to become familiar with scope and orientation.

-I think 'practice' in the trial directory will take care of this.

A) FYI: To stop the scoring timer press the freeze button. To restart it press run. And of

course you do have practice mode, as you suggest. Do we need more?



Q) Cannot feel (with haptics) distal part of middle turbinate while

trying to medialize it.


A) Haptics are only applied to the distal tip of the tool. There is a large indention along the

top of the turbinate. It looks as though the tool is going through the turbinate although the

distal tip of the tool is in the indention (no feedback). You must school the student to keep the

tip of the tool in view. I know this is NOT a good answer, but, with the technology we have it is

the best I can do.



Q) Alt-F3 error comes up and software locks up when attempting to play

back 2 consecutive recorded trials from Onyx. Must quit and restart.


A) Implemented. Looking at it. Fixed 05/09/97.


Q) Ability to separate Navigation, Injection and Dissection into three

separate tasks on novice level.

A) Not implemented. I'll think about it. Don't know quite what to do yet.


Q) When endoscope passes through the anatomy in Novice level, it should

be opaque to be more realistic. This would prevent subjects from

'cheating' during injection and dissection.


A) Not implemented. Maybe impossible with current limitations


Q) Audio cue for end of injection to 'begin dissection'.


A) Implemented. Done 05/09/97


Q) Blood spheres and bone chips toggle on/off on novice level.


A) I talked to Chuck and we decided to remove them from patientConfig. You'll get this change in Version 1.1 on Monday. (the CD is shipped!) On a subsequent release, I'll put in menu






4/29/97 Q:Chris Airola and Peter Oppenheimer


Q) Should Audio and Hoops1 on, be the default for NOV and INT?

Bloody scope off for NOV?


A) Audio, yes, Hoops1, no. In advanced level Hoops1 are default off.


Q) Should straight-biter be the first selected instrument in NOV dissection?

A) Setting order of difficulty. Can be changed in PatientConfig. Set to Chuck’s specifications on 05/09/97.


Q) The automatic selection of side biters needs to be mirrored in the left



A) Change in PatientConfig. Done 05/09/97.


Q) Should the notion of the "last target" be eliminated, have all targets look

and act alike and have injection completed when all targets have been injected?

Currently, in INT, injecting the green target generates a "Last target injected" audio prompt.

In NOV, injecting the green target switches to dissection task.


A) All targets are treated equally. Cue for "begin dissection" after last remaining target is injected. Done 05/09/97.


Q) Can we add a voice prompt cue for "dissection complete"?

Does timing automatically stop at the end of dissection?


A) "Congratulations you have completed the task" is cued after dissection is complete in Novice level. Nothing is cued for Intermediate or Advanced due to student’s discretion of a complete task. Yes, clock does stop at completion of dissection in Novice level. For Intermediate and Advanced, again, student’s discretion.


Q) Students had a bit of difficulty locating the next sphere during NOV



A) Location of spheres was taken care of by changing the color of a dissection sphere to a light shade of blue.


Q) Where is the active area of the instruments and the tissue. Students

find it difficult to know how to cut small 1 or 2 polygon bits of tissue.

Do you cut the vertex? the polygon center? and with which part of the

instrument? Does this need to be marked on the instrument or the tissue?


A) Not implemented. This was disregarded due to too much emphasis on the naive subjects tested. The resident surgeons should have no problems with this.


Q) When a sphere is partially dissected it is hard to distinguish the front

from the back since the lighting is straight on with little fall off.

This can be resolved by either coloring the inside of the sphere differently

from the outside or by applying a texture map that distinguishes front from



A) The invisible pieces were taken care of by adding a concentric sphere inside the existing light blue sphere. The inside sphere is colored yellow to give a depth cue and allow student to discriminate between the front and back of the sphere during dissection.


Q) Even for medical people, the dexterity of two handed operation can be

difficult at first. Hence we may want to freeze the endoscope and let

students inject and dissect one handed, before they try to operate the

instrument and camera simultaneously.


A) This should be implemented by proctor’s discretion. If the student is having troubles, freeze the scope, but note it in the Trial Matrix. Repeat trial until completion of task without freezing the endoscope is attained.


Q) Again we are finding a utility in deconstructing the procedure into smaller

subtasks and then integrating them in subsequent trials.


A) This should be implemented by proctor’s discretion. If the student is having troubles, break up the tasks using the ‘stop’ button. Describe the next task then continue the procedure by pressing the ‘run’ button. All these actions should be documented on the Trial Matrix sheet. The trial should be repeated without breaking up the tasks and a passing score should be attained by the student before proceeding.


Q) How do we delete incomplete trials from the student record, especially if

an incomplete trial is the best score?


A) Version 1.2, use the ‘delete record’ choice in the ‘file’ menu.


Q) Currently the simulator is compiling dissection percentages and other data to generate a student score,

storing that score and tossing the source data. That source data is not very large. Should we instead be storing the source data and generate a score based on a reprogrammable scoring algorithm as a

post process?


A) Version 1.2, The ‘studentViewer’ still outputs the same scores based on the algorithm, but now there is also a detailed record of all trials included in the viewer. This record includes times for each task, optimal times for each task, accuracy of navigation, accuracy of injection and accuracy of dissection.


Q) It is a bit unclear as to when the student record is written out.

Is it upon configure? reset? quit?


A) Version 1.2, on configure and quit it is written, on reset it is written, but overwrites. and on abort (or the rectangle in the upper left corner) nothing is written.


Q) One student suggested that a cross hair at the center of the scope image might

have facilitated easier navigation by providing an additional visual cue.


A) Not implemented. Trying to keep simulator as close to real surgery as possible.


Q) Is it possible to make the next hoop green rather than just the first hoop.


A) Not implemented. Decided that it would take more code and processing time and was not really needed.


Q) We have a possible solution to the slipping wire problem in the instrument

tracking assembly. By cutting a guide groove into the cylinder and quarter

cylinder blocks, the wire would be detained from longitudinal slipping.

Looks like a pulley.


A) Consideration by Immersion. Probably will make a cage to encase the encoder and cable to inhibit the endoscope from hitting them. This could feed the problem of the endoscope and tool colliding.


Q) The former tool still remains and is impossible to turn off if Null Tool

is not chosen before the end of the previous trial.

A) Version 1.2, Null Tool completely retracts and negates the tool. Fixed.


Q) When 'default record' is chosen in the configuration, replace the

previous name (if any) with 'default record' to clarify which is chosen.


A) Solved in Version 1.2. When any record is chosen in any of the menus, it is placed in the editable area.


Q) Zero degree scope's view should not be altered by axial rotation of the

endoscope. The lens of the scope is the only thing which is rotated,

not the camera, therefore the axial rotation of the endoscope should

only effect the 30 and 70 degree scopes.


A) Not implemented. Chuck did not want to remove the ability to rotate the camera. If the camera is rotated, the image is also rotated.


Q) Need to test the ideal relative position of the tip of the scope in the

ENDO.CFG on the PC. The tool is still in the way of the scope during

injection and dissection. Do we still have the option of cutting off

some of the end of the scope?


A) Cannot cut the scope. Both the tool and scope’s virtual eyepoint can be changed in the ENDO.CFG file on the PC. The positions must be calibrated for optimal performance.


Q) Possibly need to create a better counter balance for the scope.


A) Not implemented.


Q) Possibly need to shave hoops off at edge of anatomy in Intermediate

level, so a hoop cannot be scored if it is captured outside of the

anatomy. Possibly determine position of endoscope upon capture of the

hoop relative to anatomy and determine if the hoop should be scored.


A) Not implemented. The hoops are positioned better in Version 1.2.


NQ) For post test questionnaire, ‘Difficulty of task compared to each other task’ instead of just a numerical value.


NQA) Not implemented. It relied to much on subject’s unfamiliarity with the procedure. The ENT’s should have a reference to a real procedure.


Q) While moving through navigation, have the next hoop turn green.

A) Not implemented. Did not think it was necessary.


C) 05/08/97

* Virtual eyepoint of endoscope was pushed forward 0.33 inches, using ENDO.CFG on the PC.

* Increased the suction on the suction tool, giving it the ability to decrease the volume of the bleeders more quickly.

* Gave the ability to set haptics to either ‘full’ or ‘variant’ in forces. Full, is all tissue and bone have maximum force feedback. Variant, is tissue and bone have different forces.

* Polyps were placed in Advanced models.

* Added more grid points to Novice level to give a sense of depth. Changed top polygon to white to give a sense of orientation.

* Created two new middle turbinates, Concha and Little, to the anatomical models.

* Added 50% dissection of ethmoid cells requirement to the Intermediate and Advanced levels and added 300 seconds to optimal dissection time.

* Set up anatomical labels in intermediate level for middle turbinate, ethmoid bulla, uncinate and maxillary ostium.

* Fixed CD ROM on Onyx.

* Performed timing analysis.

* Fixed wires for encoders on tool.

* Fixed ‘Null tool’ on reset. No longer requires retraction to eliminate tool after reset or configure.

* Added "Trial Complete" message at Novice and Intermediate levels.


C) 05/12/97


* Loaded a new realTimeTask into ver 1.2 which increases the refresh rate of the graphics.


05/13/97 Subject #4&5 Comments


SC) Place more consideration in instrument used and instrument handling rather than reality of graphics.


SC) Use one handed and two handed navigation as a testing variable.


LL) Keep retract tool off in Novice level.


SC) Audio/Visual warnings for hazards (i.e. arteries, nerves, lamina...).


SC) Had problems with the scopes ability to ‘slide’ through the anatomy.


SC) Have the instrument disappear when it is ‘going through’ the anatomy. This will fore the student to keep all of the tool in view when dissecting.


SC) Allow suction tool to break open ethmoid air cells.


SC) Did not like the moment arm of the scope, the counterbalance tended to push/pull the scope.


SC) Liked the fact that the simulator was not ‘completely realistic’.


BS) Task for the sickle knife in novice level should not be cutting a sphere, it should be cutting out a surface.


Q) Ask Chuck if the three passes during navigation are done universally.


SC) Had problems determining where to dissect the anatomy; which parts were ‘active’ for dissection.


SC) Normally would grab bone chip at edge rather than at center. He spent a significant amount of time removing the bone chips.


BS) Allow suction tool to pull bone chips out at a faster rate.


BS) use Flash Frame or Slate at beginning of recording on Onyx to allow synchronization while editing recordings of Onyx and video camera. Will allow us to place the Onyx recording on the video recording.


SC) Angles of hoops towards the end of navigation passes seemed too excessive.


LL) Place the maxillary and sphenoidal sinus back into the intermediate models


05/21/97 Subject #6&7 Comments


SC) Want suction tool to pull bone fragments out and dissect.


BS) Change dissection cues to give relative scoring amounts. i.e. 50% uncinate = 100% uncinate dissected cue.


Q) Do bleeders increase in volume too fast?


NC) Adjusting/Changing the PC/Haptics interface. The vector of the instrument is drifting along the rotation about the z-axis. At tool change the vector seems to return to calibrated position. Over time with the same tool, the vector will drift either positive or negative depending on the amount of movement in either direction. The drift seems to be additive.


NC) Trim the hard pallet of the mannequin. While retracting tool, subjects tend to rub up against it/get caught on it. Causes them to pull harder on the tool for retraction.


BS) Get the cum_sim_time on the default student to calculate the total amount of time spent troubleshooting simulator.


Q) When does dissection time stop in Intermediate and Advanced levels?


C) 05/23/97


* Loaded a new realTimeTask into ver 1.2 which allows dissection using the suction tool.


C) 05/27/97


* Loaded new realTimeTask into ver 1.2.1 which allows removal of bone chips by the suction tool.

* Took out two inferior bleeders on uncinate in ver 1.2 and ver 1.2.1.

* This alteration actually moved the endoscope eyepoint, not the tool. ENDO.CFG on the PC only makes changes to the endoscope. Moved virtual eyepoint of tool back 0.005 inches which allows retraction of tool and removal of bone chips further inside nose and increases the amount of space between tip of endoscope and encoders on tool.


05/29/97 Subject #8 Comments


SC) Scope feels too heavy in lateral motions.


SC) Had problems with offset of virtual eyepoint of endoscope, he stayed outside the nose for the beginning of navigation.


BS) Have an adjustable mannequin (height, placed on an OR table) to accommodate people who sit down during the procedure. He is left handed and normally sits down.


SC) He saw larger movements than in real life with the tool (tool is too sensitive to movement) and also recognized a delay in the graphics update.


SC) The simulator forces you to move in very close with the endoscope (too close) to the anatomy you are working with.


OB) He had problems while handling the tool, keeping his thumb steady. This seems to be a recurring problem with almost all of the subjects.


OB) He did not rely on the image on the screen to initially navigate into the nose. He used the mannequin’s nose as a reference and would repeatedly get ‘lost’ inside the anatomy when looking up at the screen. This showed a major problem with the offsets of the scope.


NC) Need to pull the uncinate back or allow deeper dissection on its inferior/lateral wall to allow a more realistic view of the maxillary ostium.


SC) He did not like being forced to retract the scope when retracting the tool. He would normally keep the scope inside and change the tool. The encoders on the tool prevented this.


SC) Likes the haptics better on than off.


SC) The 30 degree scope is not right. It still resembles a zero degree scope. The view when rotating the 30 degree scope is unrealistic.


05/29/97 Chuck’s Comments on improvements to ver 1.2.1


* Increase the amount of dissection the suction tool is capable of.

* Graphics are still running too slow.

* The suction tool still needs a little bit more power when removing bone chips.


C) 05/29/97

* Took out two lower bleeders on the uncinate (X = 13.~) in ver 1.2, 1.2.1 and 1.2.2.


C) 05/30/97


* Loaded a new sinusim (sinusim1.exe) on the PC which decreases the haptics accuracy to one degree from original one tenth of a degree accuracy. This will allow Immersion Corp. to change the tracking of the tool, to allow for more freedom of operation.


C) 06/03/97


* Loaded a new realTimeTask and toolConfig into version 1.2.1 to improve the refresh rate and improve suction and dissection on the suction tool.

* Draw times: Version 1.2 = 55msec and Version 1.2.1 = 52msec with the ethmoid bulla, uncinate and posterior middle turbinate in view, microdebrieder at full oscillation (worst case).


06/05/97 Chuck’s Comments on improvements to ver 1.2.1


* Good refresh rate and good dissection for suction tool.


06/05/97 Subjects # 9 & 10 Comments


SC) Did not like the fact that he had to steady his thumb while using the tool.


SC) Wanted the haptics to ‘tear’ the tissue, wants the ability to feel the biting of the tools.


SC) Felt he had to unlearn real sinus surgery to steady his thumb.


SC) Wants an operating room bed to have the ability to raise and lower mannequin.


SC) #10, The simulator is harder than the real surgery.


SC) Both would prefer forces on the tool.


SC) Place and endoscope ‘sleeve’ on the scope to allow better grip.


SC) Want an elbow stand for stabilization.


SC) Give the mannequin head the ability to move laterally (tilt the head from side to side) to obtain better angles.


C) 06/11/97 Changes made by Jeff Miller and Chris Airola

LL) The tool was drifting across the heading and pitch during trials. It was believed that this drifting was caused by hardware that had become defective. Initially, rebooting the PC and recalibrating the tool was used to correct the problem. During testing of subjects 9 and 10, it appeared that the longer the system was powered on, regardless of the number of times the PC was rebooted, the drifting occurred exponentially faster. It was later found that the drifting was caused by interaction of the sectors (on which the encoder wires are attached) for heading and pitch, and the mannequin. This interaction moved the sectors, and also the encoder spindle, destroying the calibration of the tool.

This problem was corrected by replacing the set screws on both sectors and by placing Loctite™ on the outside of the sectors and rods.

C) Stabalized the sectors on the tool and tested a non-medical subject to test the stability.


C) Received new coated wires and set screws for the sectors and a new Allan wrench for the set screws.


LL) Put a new realTimeTask into versions 1.2a and 1.2.1 which moves the plane for retractions of tools and removal of bone chips more inside the nose. This change was made in response to the diagnosis of the tool drifting.


LL) The tool jaw 'jitters' when the haptics are turned on. This is possibly caused by the electromagnetic fields from the motors controlling the haptics.


* Version 1.2a was added as a possible replacement for version 1.2 for running subjects.

* Improved RealTimeTask.

* Planes for tool retraction and bone fragment removal were moved further inside the nasal passage to stop subject's tendencies to pull the tool too far outside the nose.

* Audio cue for tool retraction and bone fragment removal were placed before actual changing of the tool and removal of bone fragment to better synchronize the change and to stop subject's tendencies to pull the tool too far outside the nose.

* Version 1.2.2 was loaded onto the Onyx. Added features were:

* Improved movement of the middle turbinate during medialization.

* Heart monitor audio is no longer cut off by audio cues during the procedure.

* Announcement of tool after change (and retraction) is more synchronized with the system pause associated with the loading of the new tool. This change was made to lower the amount of contact between the sectors of the tool and the mannequin.

* Have the ability to press F1 over any item on the display to access its help file.

* Allows dissection of some of the nasal passage on the inferior uncinate, this allows a better view of the Maxillary Ostium after dissection.

*New patient models were created.

* OSC_7.0_edmond (Advanced Right in Version 1.2.2) was added with new polyps, ethmoid cells without ‘floaters’ during dissection and the ability to dissect more superior on the ethmoid bulla.

*OSC_7.0_Concha_bullosa = larger middle turbinate. Also has the features of OSC_7.0_edmond.

* OSC_7.0_Deviated_Septum = deviated septum. Also has the features of OSC_7.0_edmond.

* OSC_7.0_Small Turbinate = thinner turbinate. Also has the features of OSC_7.0_edmond.

* OSC_BP1.1 = New bullet path with entry wound on skin, hole through the middle turbinate and removable bullet fragment. Also has the features of OSC_7.0_edmond.


06/20/97 Change made by Dave Heskamp


C) Removed the option of turning on each individual hoop set in the menu bar. There is now one option for visibility of hoops, where each hoop set will become visible after the preceeding set is completed.


06/26/97 Subject #9 Comments


SC) The scope offset was noticeable and distracting.


06/26/97 Dr. Chuck Edmond's Comments on version 1.2.2


* Likes Middle Turbinate motion, constant heart monitor planes for tool retraction and bone fragment removal and Sinusim with one degree of accuracy.

* Need to allow dissection a little more superior on the Uncinate. This will allow a better view of the Maxillary Ostium and will be more realistic.

* Need a way to view scores other than the highest for subjects.


LL) Use 'archive' file to view previous scores, other than the highest score, for subjects.


07/07/97 Subject #14 Comments


SC) Use both the 'box' type simulator and the sinus simulator to determine whether the subjects are learning depth awareness or anatomical awareness and dexterity.


07/08/97 Dr. Chuck Edmond's Comments on new models and Version 1.2.2 (continued)


* Can we bypass the audio cues if the subject is 2 or 3 beyond what is cued up?

* Make sure to show non-Otolaryngologist and non-MD people see the intermediate instructional recording.

* Replace version 1.2 with 1.2a for the subject trials.

* realTimeTask for version 1.2a is located in version1.2/bin/ , and is named 'realTimeTask_1.2a'

* haptic_test, located in madigan/ , makes the tool the endoscope to test the haptics on a scope.

* Need more shading cues/visual cues/depth cues in the Ethmoid cells.

* Deviated septum model:

* Dont allow full medialization of the middle turbinate, only allow anterior tip to deform.

* Alter scoring algorithm for this model, take out pass 2a. This pass should not be allowed.

* Concha Bullosa is good.

* Small Turbinate is good.

* Bullet Model has a lot of illegal moves (i.e. the maneuverability through the hole), but is interesting.

* Likes the tool setting unfiltered. Sinu1pas.exe is the best, he likes the more real-time movements of the jaw. Sinu2pas.exe and Sinu3pas.exe are still too slow.

* Do the different filters on the tool affect the scope as well?

* Can we increase the frequency that the tool talks to the PC?


07/08/97 Subject #15 Comments


SC) Stabalizing the thumb is both unrealistic and adds too much difficulty to dissection.


SC) The depth cues in the novice model need to be improved.


07/09/97 Subject #16 Comments


SC) Track respiratory rate and ECG during trials to determine stress level and ability to perform.


SC) Does not like the virtual tool's ability to get in the way of the virtual scope.


SC) Would be useful for Artheroscopic and other more rigid environments rather than the abdominal cavity..


07/11/97 Subjects #18 & 19 Comments


SC) Wants adjustable table height. Ability to adjust the position of the mannequin.


SC) Does not like the ability to back through the hoops and capture them.


SC) Experienced tension in his hand, handling the tool.


SC) It is hard to determine depth in the Novice environment.


SC) Give a contrast area for dissection in Intermediate. Place a distinctive color over (highlight) the areas that are dissectable.


07/11/97 Subject #20 Comments


SC) Delay of tool jaw is bothersome.


07/16/97 New Features of Version 1.3


* Haptics

* Suction tool pulls you toward anatomy which is near its head.

* Microdebrider has a slight vibration in free space with an increase in vibration when interacting with anatomy.

* Jawed tools have a 'grab' and 'tear' feel.

* Needle has a 'pop' when entering the anatomy and a 'hold' when inside the anatomy, which prevents the needle from moving in the X,Y,Z directions, but allows heading, pitch and roll.

* All tools (except needles) have a 'jolt' of force when initially interacting with anatomy.

* Sickle Knife has feels like a straight razor cutting paper when interacting with anatomy.

* New GUI interface on SGI.

* 'Hoops' option in Training Aids pull-down menu has only one choice, either on or off. This simplifies the selection of hoops and complies with version 1.2's method of progressing through each set.

* Improved help within GUI interface. Place cursor over any item on GUI and press F1 to bring up help window.

* 2-Pass filtering on jaw of tool. (changed from 4-Pass)

* 3-Pass filtering on endoscope. (changed from 4-Pass)

* Ability to cut Uncinate more superior. Gives an ideal view, anterior to the Uncinate, of the Maxillary Ostium.

* When 'stop' is pressed on the GUI, or the PC stops receiving packets from the Onyx, the haptics on the tool are automatically deactivated.




C) Received an executable from Dave Heskamp which allows conversion of the Onyx's video of student's procedures to an ASCII file. This file will aid in statistical calculations specific to each student. Name of file: record2ASCII.


08/12/97 Comments from Subject #28


SC) Noticed the delay on the tool, tool jaw and endoscope during procedure. Also thought that her hands were too small to stabilize the tool during dissection.


08/12/97 Comments from Subject #31


SC) Does not like the tool's encoder box on its tip. Likes to get in close to areas of dissection and injection, which is inhibited by this box.


SC) Wants warnings, either audio or visual, when entering a dangerous area of the anatomy.


SC) Wants more visual cues in Ethmoid Cells. Need to be more anatomically correct both inside and around the cells.


SC) Create a procedure which guides the residents from point A to Z through the procedure. He has noticed that the residents tend to pause during the actual procedures because they have no 'plan' to follow.


SC) He usually has a table to rest his arm on for stability and to reduce fatigue.


SC) Need to word questions in Post Test Questionnaire with Model 1,2,3 only. When we use the words Novice, Intermediate and Advanced we are implying the difficulty of each model, instead of implying that they are purely learning tools.