Capella has been designated a National Center of Academic Excellence in Information Assurance Education (CAEIAE) by the National Security Agency and the U.S. Department of Homeland Security. Capella’s information security curriculum meets the NSA Committee on National Security Systems standards 4011, 4013, 4014.
The NSA’s Information Assurance Courseware Evaluation (IACE) Review Committee has validated that Capella’s information security curriculum meets the Committee on National Security Systems (CNSS) National Standards 4011, 4013. Additional courses are needed for 4014. Learn what it means to a Capella learner.
While we normally worry about computer systems being hacked, it now turns out that medical devices can be hacked as well. This article from Information Week discusses one example — but thinking about all the new technologies combining medical and technology, did they ever think of security?
Here are some other articles and discussions from Blackhat, EmaxHealth and Health IT Exchange.
Share your thoughts.
4 Comments
The referenced link returns an error page but I assume you referred to the Radcliffe wireless hack presented at Defcon.
I might reason the problem this way. Consider for a moment that an intelligent device undoubtedly uses some form of control program or turnkey operating system, using privileged instructions to function and make the device operate and do whatever it is designed.
To ease the task of building this control programming, code is probably modular. I would reason that there are storage locations where breakpoints could be enabled for troubleshooting. Subroutines and program modules may follow some form of standardized call to handshake data. Also, calls to physical interfaces would probably be standardized. Usually banners are used to label the routines, and these may be found in memory dumps.
So any programming misstep in these areas might expose the device operation to a non-privileged operation – an unauthorized program so to speak.
Those ideas, plus a knowledge of electronics, and the determination of the individual, I would reason, may expose such a device to the possibility of hacking.
I’d speculate that perhaps one way to combat this type of product tampering is to have the device compute a hash value of its programming on start-up and compare it to another hash value stored elsewhere in its memory, possibly in the control program loader. A mismatch between values would require the device to not complete initialization and indicate an error. A vendor’s maintenance upgrade would have to take into consideration the comparison of these values.
I’d like to contribute an additional bit of speculation.
Interfaces to legitimate users perhaps could be protected by using a system similar to Kerberos, where multi-step authentication verifies a user or another system’s identity and assigns privilege levels appropriate to the communication. Unauthorized users could be immediately disconnected.
Emergency access to an implanted device perhaps could be augmented with some form of non-invasive physical detection built into the logic and hardware, i.e. ultrasound, magnetic, infrared. These are sensor types that are both common and inexpensive.
Yes, they do consider security when developing these devices. But remember, many medical devices may be a generation back technologically as only well proven tech is typically used in their development coupled with the amount of time it takes to get FDA approval. That does not mean that review of the devices should stop, or that improvements could be made.
Many medical devices are linked to control servers with dedicated SSIDs on the access points and controls on the devices themselves to deter tampering. Just like in a light wireless access point coupled with a wireless controller, many medical devices are configured by the corresponding control server, and if something doesn’t match, the device is either reconfigured or an alert is triggered.
Additional security options under IPv6, such as cryptographically generated local addresses, are evolving to help protect communications to keep hackers from reaching the devices.
For more information on cross discipline research on increasing the security of medical devices:
http://www.secure-medicine.org/
Sorry about the link- things move fast on the Net..it should be correct now.
I wish hashing would help – it might, I’ve just seen a number of products fail, not the encryption part -just the implementation. Even with encryption and hashing, if a person was brought to an emergency room in an unconscious state, would those protective technologies help?
I think one of the most distributing aspects of the attack is that it was done with over the counter everyday electronic products and with the……proprietary ways of communicating….we saw this fail with the encryption debate in the 90s and the clipper chip. Hopefully with the news out, new products will be developed with more secure open protocols.
-