Introduction
Cars can be plugged in at the mechanics for electronic diagnosis, customer issues logged in enterprise support systems receive immediate potential solutions to their issue prior to a customer service representative looking at it, and computers send error reports when an application crashes. In industries across the world automated diagnostics becomes more and more prevalent leveraging continually advancing algorithms that become increasingly intelligent in identifying solutions to known problems. Yet in the health care industry Doctors have out dated and limited access to potential solutions and details from a patient’s case are seldom fully available to be investigated holistically.
There has been significant research on automated diagnosis, but limited practical application and integration of systems. The idea, well represented in Magnus Stensmo’s Ph.D thesis, Adaptive Automated Diagnosis, clearly paints the picture of how powerful and needed such an approach is.
Bias in today’s computer assisted diagnosis
“Enter symptom, disease type, test name or code” requests one physician diagnosis database. As with any human search that begins with keywords chosen by the user, bias inherently influences the results. If a Doctor has an assumed diagnosis, they will immediately begin searching for further evidence that their assumption can be validated. And if it isn’t, then they will have missed other potential diagnoses. Additionally, if the Doctor begins searching by symptoms, while these may be accurate, the order or weight given to any one symptom will give a bias toward related diagnosis when in fact, there may be a symptom not given any credit and thus not included in the search. Regardless of whether you consider today’s databases or the older process of researching in books, the results are always influenced by the bias of the researchers’ initial assumptions.
An Automated Medical Diagnosis System
What is needed instead is an approach that minimizes human bias and considers all relevant and irrelevant data in determining a diagnosis. Computer software does this well. With an automated medical diagnosis system, Doctors could be presented with multiple potential diagnoses based on all of the patient’s current and past details. Such a system could be designed for automated medical diagnosis that is based on probability, utility and decision theory (Read Adaptive Automated Diagnosis). Essentially, the computer software could be fed human observations of symptoms, test results, and any machine data collected such as blood pressure, heart rate, oxygen levels, etc. The software would then compare these observations with a database of potential diseases and external agents (e.g. , viruses, bacteria) to determine the most probable diagnosis.
These results would then be presented back to the doctor along with a probability rating indicating which ones are likely most relevant or accurate. Each diagnosis could also then be presented with additional direction to the doctor to further explore for additional symptoms and/or order an additional test. These additional observations and/or test results would again then be fed into the system where it could reevaluate the probable diagnoses canceling out some while raising the probability of others.
In addition to immediate interactions with the software, Intensive Care Unit’s machine observation data (e.g., oxygen levels, heart monitors) could be constantly fed into the system to allow the software to be looking for patterns that match other known diagnosis that would never be able to be caught by a human as it would take too much time to evaluate the data. Nurse’s notes could also be used in a similar manner.
The most challenging data for the automated medical diagnosis software to interrupt would be results of imaging systems. However, current advances in face recognition technology can and have been applied to reviewing images such as x-rays. This is important as “image interpretation is an error prone task. The number of lawsuits filed against medical imaging professionals that are related to the miss of a diagnosis is close to 70% (Berlin, 1995). The most common errors are perceptual errors that lead to diagnoses misses, representing about 60% of the cases (Renfrew et al., 1992).” – Application on Reinforcement Learning for Diagnosis Based on Medical Image . With a software solution, images could be reviewed against known patterns and then presented to a Doctor for a final review. Similarly, following a codification process, a standard could be defined to notate what one sees in an image that would be relevant to a medical diagnosis, which could then be understood by the software.
The database of potential diagnoses should be an online service that all medical institutions interact with in a real-time basis. This would allow for two key additional benefits. The first would be to connect Doctors in real time with their peers at other facilities with patients experiencing a similar condition. This would allow for immediate collaboration that could lead to a faster treatment. And second, by having the system online, you ensure that every doctor has access to the latest scientific diagnoses.
The system should then be designed to become increasingly more intelligent with time from increased information about each disease and ailment as well as feedback as to the accuracy of its results. Specifically, after each diagnosis, the Doctor would be responsible for submitting feedback as to whether or not the chosen diagnosis and subsequent treatment successfully resolved the patient’s problem. These responses could then go through an automated peer review with the results updating the probability factors to each disease and associated symptoms, test results, etc. Additionally, the software would record the other data learned from each patient in consideration of relevance to future cases. Results could be rated in accuracy based on how they are determined. For example, the results obtained from a biopsy or autopsy may be weighted as being more accurate than simply an observation from the Doctor that the patient recovers.
Available today
Today there are a number of point-of-care (POC) testing solutions, (also called near patient or bedside testing), but no large scale application of an automated medical diagnosis system. Today’s applications are for very focused blood tests and are far from providing the capabilities described above. Some examples:
- http://www.icutracker.net/
- http://www.rals.com/
The machine data required for the software has been developed such that it could be fed into such a system. There are many systems available today that would likely only require some standards to be developed for the resulting data. Some examples:
- Life support systems
- Positive patient ID
- Meds tracking
- Lab work tracking
- (see http://www.handheld.com/lp/healthcare.htm)
Conclusion
Software engineers everywhere can read this idea and will instantly recognize that today’s technology supports this solution. However, the greatest obstacles would be to gain public support for the requirements of fully electronic medical records and Doctors learning to work together. Progress is being made on changing the culture on both of these topics; unfortunately we have a long ways to go.
This article is dedicated to Grace Allen, a beloved mother, grandmother, great-grandmother, friend, cousin, partner and overall amazing woman. Had this system been in existence today, we would have had some more wonderful years with her.
References
- Stensmo, M & Sejnowski, T. J (1996) Automated Medical Diagnosis based on Decision Theory and Learning from Cases, World Congress on Neural Networks pp. 1227-1231
- Stensmo, Magnus (1995). Adaptive Automated Diagnosis. PhD thesis, KungligaTekniska Hogskolan (Royal Institute of Technology), Stockholm, Sweden.
- Stelmo Magalhaes Barros Netto, Vanessa Rodrigues Coelho Leite, Aristofanes Correa Silva, Anselmo Cardoso de Paiva and Areolino de Almeida Neto (2008). Reinforcement Learning ISBN 978-3-902613-14-1. Application on Reinforcement Learning for Diagnosis Based on Medical Image.
- Point of Care Testing
- http://www.pointofcare.net/
- http://www.meditech.com
About half past midnight I left my house to head to the Apple store on Stockton Street in San Francisco. I had on multiple layers of clothing and my motorcycle leather jacket, stocking cap, gloves and a back-pack filled with more layers and even a towel. I was as prepared for sleeping on a sidewalk as I figured I would need to be. I’ve been known to be able to sleep anywhere, so I figured I didn’t need any fancy tarps, tents, blankets or other creature comforts I was sure would be on display in the line.
When I got to the Apple Store there were about thirty people already in line. I had heard that the first person started out Wednesday evening. It was a fairly quiet group of mostly geeks and artists (inclusive of Apple developers). The ratio of boys to girls was unfortunately very high (come on girls get into tech!). I walked to the last place and just stood there. I had no idea what to do now that I was actually in line. Now, many people are thinking, “this guy is nuts waiting in line outside in a city to buy a phone.” I even wondered to myself if I was maybe crossing some imaginary line between the already obsessed Apple enthusiasts and well, part of the cult of Apple. I figured it was going to be a great adventure to say the least, a first person experience of where society comes together for a single moment to celebrate innovation. To be honest, the first hour was pretty boring.
A guy came over with his laptop asking for a MG. He pinged a few of us in line, but none of us was who he was looking for. By the way I saw him glancing back at his computer, I assumed he had made a connection with someone in line, online. He had. A couple people up from me was MG. The two connected live and a whole group began to mix into conversation. The line was starting to interact. I’m sure the folks at the front had their moments already, but the tail was starting to wag. We were now a group. Occasionally, someone would need to run to the bathroom or grab coffee from the 24-hour Starbucks. When they went to go, they’d just turn to their neighbor and ask, “will you watch my stuff?” It was an instant nod and the bond of trust amongst the line increased. There were laptops, cell phones, music players, back-packs, blankets, tents, and even a dog. Everything was safe. Everyone was together.
I happened to be one person away from the guy “holding a place” for an editor from 
cards. The news crews starting showing up in the early morning, still before sunrise. They’d walk up and down the line interviewing the “die hards”. Application developers were also there pitching their new applications on handbills or demoing them in some instances (Pandora cofounder was walking the line I believe – Pandora is awesome). The excitement kept growing. By 5am no one was even trying to sleep. We were all anxiously awaiting 8am. We were now a part of something. As the sun began to rise, more and more people showed up. We broke a hundred before dawn, but by the time it was day light the line stretched around the block, thicker, stronger and buzzing.
Security was at the doors, even the police had arrived. Everyone was up on their feet, stuff picked up, and all moved forward about twenty feet. We bunched up. We had been waiting for this moment. We all watched the clock intensely for about twenty five minutes prior. There were film crews from the news stations and photographers everywhere. The doors were opened and thirty people were invited in. First it was about a half of a dozen quickly allowed in, then as we all blinked and wondered if we had really seen anyone enter, the rest of the first group were allowed through the giant glass doors. It had begun. As I was 34th in line I was now at the front of the door lined up to be in the second group (we had all heard they were letting in 30 at a time as folks were still online with their laptops and iPhones reading about what the other experiences were from the East Coast onward. Nothing happened for a while. We could see the first group in line up the stairs. No one came out. We all calmed ourselves by reminding everyone that it was going to take about 15 minutes to get one person through the process. It was 8:06. It felt like it was 9. 
Finally, the guy that was first in line, came busting out the door all fired up. He was upset because nothing was working and he still didn’t have an iPhone. He had been apparently told that he had to move his tent or the police were going to confiscate it, so he was allowed to come back outside. As he did he was mobbed by the media and went off on a tyrant. He was really peeved. The media ate it up. I was surrounded by cameras, journalists and #1. We just wanted our iPhone 3G. The guy next to me had an original iPhone with a bit of juice left. He said the stock was starting to really slip. I knew it was bad. We heard the servers had crashed and that everything had come to a grinding halt. I knew what happened. Too many transactions all happening at once on systems that weren’t ready for this volume. I’d been in that situation before. I was very happy to not be in the call center fighting this time, but was hopeful that the IT folks would figure out a solution quick. After a while, the first iPhone activated walked out the door. He wasn’t first in line, but his worked first. Slowly a few would come out at a time and eventually, the second group was allowed in the door.
When it was my turn I debated quickly, “white, black, white, black, white, black, … which one do I want”. I choose black. As the employee started the transaction, it quickly failed. She couldn’t get me setup. Nothing worked. She wasn’t sure why either. I had heard from some others before me that there were problems with people that had corporate discounts, not corporate accounts mind you, but just discounted person accounts. As it turns out, Apple and AT&T didn’t setup Apples point of sale system to handle customers that had a corporate discount. What?! That has to be most customers out there! I was distressed to say the least. I was tired, dirty and felt like a two year old having all their toys suddenly destroyed. I was really upset. I slept over night for this. I was a part of this. I was a part of the group. I couldn’t walk out without an iPhone 3G. Multiple people tried to help. Everyone said it was impossible and that only AT&T could do my transaction. I refused to believe it. I was so upset. But, I knowing how these systems worked, I knew that if they didn’t build it to support that transaction, there wasn’t anything I could do. The only option they offered was to get a new phone number (I was an existing AT&T customer). That wasn’t an option.