National Cancer Center Research Institute of Japan Tele-medicine Asks for Global Addresses

Hiroshi Mizushima
Director, Center for Medical Genomics


Hiroshi Mizushima at National Cancer Center Research Institute is one of the pioneers of the Internet in Japan. He experimented extensively with Gopher, long before Web became a common information publication tool, for example. He is the Director at Center for Medical Genomics now, making extensive researches on the relationship between genes and diseases, while actively involved in tele-medicine projects. We asked him about the present and future of medicine, the Net, and IPv6.


IP-connected Devices Change Medicine

IPv6style:You are a medical expert. What do you think of the necessity of IPv6 in networked medicine?

Mizushima:We used IPv6 for tele-conferencing with an organization in Hawaii. We currently use IPv6 to satisfy such ad-hoc needs. These things can be done with IPv4. As for the things that can only be done with IPv6, they are still only in our minds.

But current IPv4 has significant limitation, such as the inability to assign global addresses to connected devices, requiring NAT. Need for NAT makes devices and mechanisms for net connection. It would be very handy if all devices were to have IPv6 addresses.

I'm involved in a project on tele-medicine for intractable disease patients. We would like to place artificial respirators and other devices at patients' homes. Hospital stay is stressful for patients. We would like to let them stay at home as much as possible. Patients requiring constant attention need to be close to doctors, but home is the best location for those at chronic stage. Patients at home, however, fear that they cannot see or talk with doctors when they need them.

That's why it gives them a piece of mind to connect medical devices to the Internet. For example, artificial respirators are very difficult to configure, and any trouble with it can immediately lead to death. So people providing care for those patients feel uneasy. It helps ease such anxiety if we could offer assistance through video conferencing or if we could remote control these devices. Hospitalization also causes social stress and cost.

For on-demand control or monitoring of medical devices, it is important that every one of them gets a global address through IPv6.

IPv6style: Would you say that tele-medicine with video conferencing is shifting from experimental stage to actual deployment stage?

Mizushima: I think so. I'm involved in tele-medicine for 10 years, and I attend tele-medicine conference every year. Conventional tele-medicine projects were typically large in scale, such as those involving High Definition live video transfer among cancer centers in Japan. They were only possible with national budget and with participation by large hospitals.

I see tele-medicine being applied to patient care at home. It is triggered by the fact that mobile phones and PCs are getting more popular and easier to use, while broadband Internet connection is getting widely available at home. With mobile phones being connected with more bandwidth, we'll have even better environment for tele-medicine for everyone.

IPv6style:Some IPv6 trials attempt remote health management through devices directly to the Internet. What do you think of the possibility of health management on network?

Mizushima:I have been riding bicycle to work for 20km (32 miles), for several years now. That means 1,500km (2,400 miles) a month, and it's a great way to stay fit. I would like to connect my bicycle to the Internet. My bike has got a GPS receiver, and I download the information to my PC everyday. I hope this could be automatically done through networking. It would be very convenient if pedometers and thermometers were all directly connected to the Internet.
These devices should be directly connected for resulting data to be managed properly. When I see accumulated data, I feel I want to do more. Some devices on the market come with infrared connection, but I want them to be directly connected to the Internet for direct communication. Each device should have global IP address to communicate on its own.

Daily-recorded body temperature data helps a lot to detect change in my body. Normal baseline data becomes suddenly meaningful when the person gets out of shape.

Diabetes control is a great example. Current diet therapy depends largely on self-reporting by patients. Patients report doctors what they ate, for calculation of calories. But doctors cannot tell if patients are cheating, as no objective data is available. But if blood sugar can be monitored real time, we can calculate how much he/she ate. They can't eat unnoticed. This could work for diet in general, as well as diabetes control.

Challenge in this mechanism is enabling noncontact blood sugar measurement, while it is done by blood examination at present. I would like to see it happen, because the system also helps detect diabetes shocks that occur when blood sugar gets abnormal in real time.


New Way of IPv6 Address Allocation

IPv6style:What do you think IPv6 lacks or is not convenient enough at present?

Mizushima:Using IPv6 requires special service contract or configuration at present. A lot of devices are IPv6-enabled, but ISPs as well as users are not making good use of IPv6. That is probably the biggest reason that IPv6 is not popular yet.

I think that killer applications are indeed necessary for IPv6. I'm not saying that we need applications that consume a lot of IP addresses. But IPv6 cannot be popular unless there are some applications that can only be realized with IPv6. Otherwise, IPv6 will not be used as the default protocol.

I hope that 10,000-yen (85-dollar) home routers support IPv6 by default, as many home electronics products get connected, with a home server monitoring the state of each device.

For medicine, these can be used to help promote medical care at home. Helping these home patients is a serious issue, and popularization of IPv6 may provide a great opportunity to improve the situation.

IPv6style:You talked about killer applications. What sort of applications do you think would effectively use IPv6?

Mizushima:Filling the lack of address space may be one advantage, but I am hoping that such huge address space could be used for authentication. A certain bits in IPv6 address may be assigned for specific industry to be used for medical purposes.

Such mechanism should help control communications, effectively creating VLANs, assigning a certain network path or destinations, which are not provided in current IP address. I don't want just large address space, but some meaning in it. One PC could be used for various purposes by re-assigning IP addresses.

This may be a laymanÅfs idea. IPv6 experts may have thought about it. But I hope this is realized if possible at all.

IPv6style:Address allocation is an important and sensitive topic. IPv6 basic header is basically designed for topology aggregation.

Mizushima:That's right. IPv6 is based on hierarchical address allocation to solve the mess in current IPv4 routing. What I'm saying might be the opposite of what IPv6 designers have thought.

I used to do network administration. Network administrators want to control routing using IP addresses. But for users, it is more convenient if they can use different services just by switching IP addresses. I hope it would be enabled with IPv6.

IPv6style:Any other comments on IPv6?

Mizushima:IPv6 address is being consumed at a rapid pace. I thought transition to IPv6 would have been earlier, but people generally manage to do with IPv4 at present. Here is the danger. People are getting so used to using private IP address, that they begin to believe that anything can be done with it. But global IP address is the basis of communication, so I hope we move to IPv6 early.

When I was a network administrator, we didn't accept the use of private IP address. Something weird happens once we allow the use of the same addresses at two or more locations. We hated DHCP, because we couldn't specify terminals that way, without administrators directly allocating address to each one. I hope that we could manage communication based on each terminal having an IP address.