During the last meeting, I was asked if I had any ideas on how to approach the problem of automatically segmenting the different features of the heart. While I currently lack the background to speak definitively about the "low level" technical details of creating such an algorithm, I have several usability ideas that pertain to how the program will eventually be used by the clinician. I believe that, while the accuracy of the algorithm is certainly an extremely aspect of the program's design, the interface itself must be intuitive enough to be used easily while conforming to the user's expectations in operation and implementation. In many ways, the interface itself is just as important as the algorithm behind it.
1. Users like control
Currently, the algorithm developed by Noel Codella is highly automated, but in doing so, takes much of the control away from the user. To operate the program, the user must simply identify the left ventricle in one slice of the cine MRI with a single click. The algorithm handles the rest. But to an experienced clinician who is used to performing all of the tasks manually, this degree of automation is a little disconcerting. I believe adding an option for the user to "tweak" the algorithm's segmentation decisions would lessen this sense of detatchment. While it is debatable whether a computer or a human can provide the best segmentation decisions, giving the option to perform manual corrections would make the user far more comfortable in using the program.
2. Users like feedback
As much as everyone loves to hate the little status bars the creep across the screen while we perform tasks on the computer, they are essential. I never really realized this until I saw the automatic segmenter in action. The program, while relatively fast, requires up to 30 to 45 seconds to fully complete segmentation after the initial user input. Unfortunately, there is no visual indication concerning the program's progress during this time. In every trial thus far, the program has finished successfully every single time it has been run, so it would seem that a status bar would serve little purpose other than to provide a little visual entertainment for the user. In reality, I've found that this has been one of the major complaints about the program's current implementation. Users like feedback about what the computer is doing!
These are all pretty minor suggestions that have little to nothing to do with the algorithm itself, but I feel changes such as these would make the program far more user friendly and will make it more likely for the program to be used in an actual clinical setting in the future.
Friday, July 20, 2007
Thursday, July 19, 2007
Hearts and Lungs
Adventures in the OR
Since my last post, I have seen quite a few operations. The one I found particularly interesting was an arterial switch (known as a Jatene procedure), which is performed within the first two weeks of the patient's life. The purpose is to repair dextro-transposition of the great arteries: a congenital defect in which the pulmonary artery and the aorta are transposed. In other words, the aorta is where the pulmonary artery should connect to the heart, and vice versa. What effect does this have on the body? I'm glad you asked. Essentially, it results in two circulatory loops: one that pumps the blood continuously through the lungs, and the other that bypasses the lungs altogether. As a result, oxygenated blood will not be distributed to the body.
I watched Dr. Jonathan Chen from Cardiothoracic Surgery perform this operation on one of the patients from the neonatal ICU. I was very fortunate to be able to witness such a long, arduous surgery, and I was very impressed at the level of orchestration compared to some of the previous surgeries I have seen. Dr. Chen required almost complete silence during most of the operation, which is something I found out the hard way...
Change of Direction
Regarding my work with Dr. Frayer, I have begun looking into computational fluid dynamics as a new direction for the lung model. The original model was based on a lumped parameter circuit analog, but we decided to abandon it for a number of reasons. For example, although such models can be fitted to clinical data, they lack the predictive power that can be achieved with a 3D CFD model, particularly for studying the upper airways. Also, it assumes homogeneous ventillation to the acinar units of the lungs, which is a poor assumption especially when we are interested in modeling how the acinar units are ventillated.
I have begun working with Fluent, which is a popular CFD software package. I spent the last couple days working through tutorials for both Fluent and Gambit (Fluent's 3D modeling program). So far I haven't made much progress. The picture shown below is supposed to be a trachea with the two primary bronchi.
Wednesday, July 18, 2007
Cranioplasty with Insertion of Medpor
As I discussed last week during our weekly rendezvous, by far the coolest operation I have seen has been a craniofacial plastics-neuro tag team effort to reconstruct a man’s frontal skull bone after an unfortunate encounter with a baseball bat to the face. Poor guy. Anyway, to reiterate from last week, essentially this young man had very little frontal bone left and a shattered nasal bone due to said trauma.
So, now for some details…first, an incision was made from ear to ear across the crown of his head. The face was eventually peeled forward to reveal the defect and then the neurosurgeons, wearing super-geeky, thick double-lensed, black-rimmed glasses, started picking his brain…the not-so-fun, invasive way. I believe all they were doing was taking biopsies as very little tissue was removed, but I could very well be mistaken. It was so cool. There I was gawking at this guy’s brain pulsing to the rhythm of his blood flow. Awesome. It was actually quite curious, the location of the defect and extent of the face peel really looked like his eyeballs should have been starring back at me…yet, they weren’t…weird. I guess they were hidden under his face flapped over his nose. That sounds icky.
Anyway, tag team high five and plastic surgery’s on the scene. The procedure began by a leisurely drilling of his skull to isolate a portion of cranium destined for an autograft. Bone filings went EVERYWHERE…the surgeons’ faces, the floor, half of the observers - everywhere. Following this mess was a rather barbaric technology (if you ask me)…the hammer and chisel. Yes, we can put a man on the moon, genetically engineer all sorts of things, and yet, in order to take a simple graft we still use the age old, chisel to beat on someone’s skull until chips of bone fly into a nice warm, blue bowl of saline…there’s gotta be a better way to harvest bone grafts. This dude took a beating…again. So, just for a visual (since the pictures I took were on the fellow’s camera, and thus, gone forever), this guy’s face is peeled off, his parietal skull bone is drilled and chiseled away to remove a bone graft and blood is everywhere...and he's still missing most of the frontal bone.
Now for the cool innovations. First, a material developed by the medical device company, Stryker, provided us with a quick-drying (~30-40 seconds) cement of hydroxyapatite…a precursor to bone. This paste was spread over the donor site created on the parietal bone…and eventually along the interface of the implant and the skull. The next cool biomaterial used was Medpor (the implant), which is a custom made, perfectly contoured, porous sheet of high-density polyethylene microspheres sintered together...kinda like the small piece shown below.
Medpor is used instead of say, PMMA (polymethylmethacrylate), due to the lower risk of complications. Potential side effects of PMMA may include local tissue damage caused by its exothermic reaction, release of toxins, and the possibility for a high rate of infection. Medpor doesn’t have these. However, Medpor has several very enticing benefits, including its flexibility and ease of coverage, its biocompatibility, and its porous nature, allowing for the rapid ingrowth of vasculature and soft tissue. The latter is very important because it promotes wound healing and stabilizes the implant.
As an aside, one study I read about had zero post operative infections, zero wound breakdowns, and zero follow up surgeries after a period of 4 years in 611 procedures (Liu et al., Neurosurg. Focus 16, March 2004). Pretty cool stuff. I hope I invent something totally sweet like this.
Anyway, this sheet of Medpor flawlessly fit into the man’s cranial defect and was screwed into the skull for support with titanium (inert and biocompatible) screws. It looked something like this, but this is not from the actual case.
To finish the case, the man’s face was stapled and sewn back on…and despite the head bandage, he looked pretty normal. Amazing.
tying a knot laparoscopicly
I have heard of laparoscopy for several times. It’s minimally invasive and also known as a keyhole surgery. Patients don’t need to be hospitalized for a long time after operation. Improved cosmesis and reduced wound complications associated with large scars are also other major advantages. But not until this summer immersion did I have a chance to see a laparoscopy. Using one word to describe it from my own feeling, AWESOME.
The patient had a retroflex sigmoid colon. Because of gravity, the redundant colon constantly drags other part of colon and pushes rectum, which also brings pain to the patient. What the doctor will do is that he will use threads to knit a net to pull up the redundant colon. It sounds like a good idea.
At the begi
nning, the doctor made several small incisions on the abdomen and put come metal pipes through the abdominal wall as channels, so that a laparoscope and some other surgical devices could easily go in. Then CO2 was inflated and abdomen was blown up like a balloon to provide working and viewing place. The laparoscope captured real time movies and transmitted it to a high definition LCD. This is doctors’ “eyes” in the operation. Those long thin surgical devices are doctors’ “hands”.
Then we began a wild exploration. After identified several organs, the doctor thought that the position of uterus was hindering the operation so he delicately tacked the uterus onto the abdominal wall. He afterwards reassured me that considering the patient’s age, the chance for her to use her uterus was extremely low. That’s the reason he did so and it wasn’t a common procedure.
Finally, the retroflex colon was found. What happened next was I saw needles and threads going into the metal pipe. The long thin picker was so agile and efficient in such a small space. Within less than half an hour, some knots magically shown up on the monitor, and gradually assembled into a net. The redundant colon was fixed.
When the doctor was tying a knot, he tried to convince me that making a knot laparoscopicly is simple. I don’t believe him this time. Why? Just check out the underneath animation.
The patient had a retroflex sigmoid colon. Because of gravity, the redundant colon constantly drags other part of colon and pushes rectum, which also brings pain to the patient. What the doctor will do is that he will use threads to knit a net to pull up the redundant colon. It sounds like a good idea.
At the begi
nning, the doctor made several small incisions on the abdomen and put come metal pipes through the abdominal wall as channels, so that a laparoscope and some other surgical devices could easily go in. Then CO2 was inflated and abdomen was blown up like a balloon to provide working and viewing place. The laparoscope captured real time movies and transmitted it to a high definition LCD. This is doctors’ “eyes” in the operation. Those long thin surgical devices are doctors’ “hands”.Then we began a wild exploration. After identified several organs, the doctor thought that the position of uterus was hindering the operation so he delicately tacked the uterus onto the abdominal wall. He afterwards reassured me that considering the patient’s age, the chance for her to use her uterus was extremely low. That’s the reason he did so and it wasn’t a common procedure.
Finally, the retroflex colon was found. What happened next was I saw needles and threads going into the metal pipe. The long thin picker was so agile and efficient in such a small space. Within less than half an hour, some knots magically shown up on the monitor, and gradually assembled into a net. The redundant colon was fixed.
When the doctor was tying a knot, he tried to convince me that making a knot laparoscopicly is simple. I don’t believe him this time. Why? Just check out the underneath animation.
Neurological Surgery
This past week, I had the chance to observe several more procedures in the OR, including observing a kidney transplant surgery with Dickson. I watched the removal of the kidney from one patient, handling of the kidney while the recipient was opened up, and attachment of the kidney into the recipient. The connections that must be attached when put in the recipient are shown in the figure. I found the handling of the kidney external to the human body to be especially interesting. Since there were delays in opening up the recipient, the transplant doctor kept the kidney in slush. Then, when the recipient was ready to receive the kidney, the doctor wrapped the kidney in some wet cloths packed with slush in what he described as a ‘sushi roll’. He used this as an ice pack to keep the kidney cold while attaching the kidney to the patient and mentioned the need for a better kidney preservation machine that could mimic the human body and would not require cooling the kidney down.
Neurological Surgery Aneurysm Treatment
As I’ve mentioned previously, there are two major techniques to treat brain aneurysms – clipping the aneurysm or using filling the aneurysm with coils through endovascular means, also known as coil embolization. Coil embolization is much more attractive because it is less invasive, however there are limitations on when coil embolization can be used, depending on aneurysm shape, as well as problems with aneurysm recurrence (defined as aneurysm neck enlargement or recanalization) when the aneurysm is not adequately occluded.
Computer illustration of process of clipping an aneurysm. www.thejns-net.org
Figure illustrates filling of aneurysm with coil. www.brainaneurysm.com
Traditionally, coils have been made out of platinum. One popular coil is the gugliemi detachable coil (GDC) which has a shape memory effect. It is routed into the aneurysm and detached from a pusher wire using a low current. Due to the problems demonstrated with significant recurrence rates, researchers have been trying to come up with techniques to increase the percent occlusion (defined as the volume of the coil over the volume of the aneurysm). A modification of the GDC coil was introduced, called the Matrix coil, which has an absorbable polyglycolic-polylactic acid (PLGA) copolymer coating on platinum coils. The PLGA is absorbed by the body within 90 days. The premise for why this technology is an improvement over GDC is that it provides a much higher initial occlusion. I am working on analyzing clinical data to investigate and compare the outcomes of the different treatments.
Figure illustrates mechanism for inserting matrix coils as well as different configurations of coils. www.bostonscientific.co.uk
Portable X-Rays Acquisition
This past week I followed a portable X-ray technician as he went around to image patients. Portable chest x-rays are done generally because the person can't really leave where there at. But the x-ray is important, as it needs to be done anytime any tube or line is inserted or withdrawn from the body. So if they can't go down to radiology, bring radiology to them.
www.floridaxraysales.com
Above are some portable scanners to give you an idea of what they are like. For reference the ones I saw were specifically GE Medical Solutions. And Kodak makes the film (Surprise! or not). From a technical point of view, portable X-ray machines are actually rather intriguing. They don't expose film that a person can read right away. Instead you shoot at a cartridge (a $400 re-usable cartridge) that you place behind a person while the are laying or sitting down. The cartridge is basically comprised of a box with a phosphor plate inside. Once the scan is acquired, you pass into a machine and it gets scanned into the hospitals PACS (picture archive effectively). Then it can be passed to radiology for reading digitally, no film running here! Also, a portable X-ray machine has a far more concentrated beam and much less scattering, so no lead coats are required, which is useful.
So I traveled the 4th floor with the tech (nice guy), and got to see how he acquired the scans. Most people are very understanding when he comes in (I think its because of all medical procedures they understand x-ray the most). The tech usually leans the patient forward, slides the plate under their chest, steps 70" back, yells "X-ray", hits a button, yells "Clear", and we are done. Its quick. The general result is below
www.med-ed.virginia.edu
Sometimes there are problems. Like if a doctor doesn't put in orders, but need it done 'stat'. Or if there is double-exposure. Or the patients obese. Or the patient is rotated. For that last one sometimes there nothing that can be done.
Above are some portable scanners to give you an idea of what they are like. For reference the ones I saw were specifically GE Medical Solutions. And Kodak makes the film (Surprise! or not). From a technical point of view, portable X-ray machines are actually rather intriguing. They don't expose film that a person can read right away. Instead you shoot at a cartridge (a $400 re-usable cartridge) that you place behind a person while the are laying or sitting down. The cartridge is basically comprised of a box with a phosphor plate inside. Once the scan is acquired, you pass into a machine and it gets scanned into the hospitals PACS (picture archive effectively). Then it can be passed to radiology for reading digitally, no film running here! Also, a portable X-ray machine has a far more concentrated beam and much less scattering, so no lead coats are required, which is useful.
So I traveled the 4th floor with the tech (nice guy), and got to see how he acquired the scans. Most people are very understanding when he comes in (I think its because of all medical procedures they understand x-ray the most). The tech usually leans the patient forward, slides the plate under their chest, steps 70" back, yells "X-ray", hits a button, yells "Clear", and we are done. Its quick. The general result is below
Sometimes there are problems. Like if a doctor doesn't put in orders, but need it done 'stat'. Or if there is double-exposure. Or the patients obese. Or the patient is rotated. For that last one sometimes there nothing that can be done.
da Vinci...man was a genius
This week I observed the removal of a prostate that had been affected by cancer. The prostate is a gland that provides nutrients for the seminal fluid that leaves the seminal vesicles. It is located below the bladder and is also connected to the seminal vesicles. The urethra starts at the base of the bladder and runs through the prostate to the end of the penis. In the past, at NYPH, the patient's abdomen was cut open to remove the prostate, which is a relatively invasive procedure. Currently, at NYPH, a machine termed "da Vinci" is used to perform the operation. There are several benefits to this technique: 1) it is minimally invasive (as I will soon describe) 2) the patient can leave the next day and 3) there is a lot more accuracy where cuts are made. The main drawback of this technique is that it requires a large learning curve.
The da Vinci surgical system consists of two major components. The first component is comprised of three robotic arms (top image). The center arm is used to control the endoscope and the other two arms control the two instruments. The other major component is the control center (bottom image). For this procedure, 5 incision points are made in the abdomen of the patient. One incision point in made near the medial line of the body, which is where the endoscopic arm will be inserted. Two more incisions, one on each side of the middle incision, will be made for the two robotic arms that will control the surgical instruments. And two final incision points will be made on the abdominal sides, for two additional instruments used by the nurses to assist in the surgery. Below is an image of what whole procedure may look like.
Throughout the surgert, the doctor will sit at the control station and control the arms. It was truly an awesome sight to see the large robotic arms swaying around and hitting the assisting nurses as the instruments cut through the tissue. Below is a picture of what one of the tools may look like. The precision that the system allows is incredible. One of the main issues of the surgery is to not cut the nerve that may be responsible for sensation. This system allows the surgeon to clearly see what he/she is cutting.
In the end, the prostate is disconnected from the bladder, the seminal vesicles and vas deferens are clipped, and the urethra is reattached to the bladder. The man will obvioulsy not be able to have children, but if the surgery went well, he would still be able to have sensation in his penis and will still be able to have an orgasm, although there would be no ejaculation. Leonardo didn't design this system, but if he were still alive today, I know he would have.
The da Vinci surgical system consists of two major components. The first component is comprised of three robotic arms (top image). The center arm is used to control the endoscope and the other two arms control the two instruments. The other major component is the control center (bottom image). For this procedure, 5 incision points are made in the abdomen of the patient. One incision point in made near the medial line of the body, which is where the endoscopic arm will be inserted. Two more incisions, one on each side of the middle incision, will be made for the two robotic arms that will control the surgical instruments. And two final incision points will be made on the abdominal sides, for two additional instruments used by the nurses to assist in the surgery. Below is an image of what whole procedure may look like.
Throughout the surgert, the doctor will sit at the control station and control the arms. It was truly an awesome sight to see the large robotic arms swaying around and hitting the assisting nurses as the instruments cut through the tissue. Below is a picture of what one of the tools may look like. The precision that the system allows is incredible. One of the main issues of the surgery is to not cut the nerve that may be responsible for sensation. This system allows the surgeon to clearly see what he/she is cutting.
In the end, the prostate is disconnected from the bladder, the seminal vesicles and vas deferens are clipped, and the urethra is reattached to the bladder. The man will obvioulsy not be able to have children, but if the surgery went well, he would still be able to have sensation in his penis and will still be able to have an orgasm, although there would be no ejaculation. Leonardo didn't design this system, but if he were still alive today, I know he would have.
Subscribe to:
Posts (Atom)