Skip to content Skip to navigation

The iPhone Can it Walk the Talk?

August 5, 2008
by Joe Marion
| Reprints

I read with interest this week, Dr. John Halamka’s blog on the iPhone (http://geekdoctor.blogspot.com/, August 5 2008), wherein Dr. Halamka relays his experience in using the new G3 iPhone. As a consumer device, he concludes the iPhone is well suited due to its speed and graphics capability. As a business device, the iPhone falls down on several accounts, one being its lack of a tactile keyboard. I’ll let you visit his blog to get a few good laughs over the predictive word results of his typing. The part that caught my attention was the typing difficulty on the iPhone due to the lack of a real keyboard, and the need to navigate through three different screens to get to various symbols such as the backwards slash (\).

While the other features of the iPhone have merit for medical application, it’s the typing that seems to disappoint. I got to wondering – what might be a “killer application” for healthcare? What I came up with seems plausible, and well suited to medical applications. Why not couple the iPhone’s image management functionality and email capability with speech processing to create a more useable diagnostic and treatment assistance device?

I was impressed with McKesson Corporation’s RSNA 2007 exhibit last November where they mimicked the Apple and Microsoft hand gesture technology to view images on large flat panel displays. These technologies seem well suited to image reading in their ability to easily move and resize images, and change cases on the screen. McKesson also linked the display to the iphone as a convenient means for delivering diagnostic results.

The convenience of results delivery and viewing convenience seem to be complicated by the difficulty in a physician’s ability to respond! Imagine the convenience of viewing study results while commuting, but the frustration of the inability to respond. Now imagine if the iPhone’s capabilities could be married to speech processing technology, such that the accompanying email could be read to the physician, and the physician could easily use speech recognition to reply to the email? The application of speech recognition technology could offset the iPhone’s typing limitations and make it easier to use.

Clearly, one stumbling block is the processor horsepower and memory requirements of speech recognition applications. I can personally relate to this as an early IBM MedSpeak/Radiology reseller! The turning point for speech recognition was the advent of the Intel Pentium Pro 200 MHz. processor (anyone remember this!), which finally provided enough horsepower to make the application commercially viable. This meant lugging around a tower computer for demonstrations, which was a real challenge when traveling by air. IBM went so far as to ask Intel about packaging up the Pentium Pro in a laptop. They tell the story that when the Intel engineers stopped laughing, they explained that the Pentium Pro ran so hot that it would melt the plastic case of a laptop! So, resellers such as myself had to wait for the second version of software that optimized the application to run on a 166 MHz. Pentium MMX processor that would operate in a laptop.

So, the question is, does the iPhone have sufficient processing capability to support a speech recognition application? Probably not. But, there are now speech applications that will allow for voice files to be captured over a phone line and processed by a server. All that would be necessary would be an interactive email application that could use voice commands to play and record messages, and then process the recorded response into text that could be approved and sent as regular email. Now a user could rely on the iPhone’s gesture technology to access an email, review an image, and set up a response, and then use speech and voice recognition to respond.

For anyone who has used speech recognition applications, you will know that it is much faster to speak a response than it is to type one. I use the technology all the time when traveling, as I find it much easier to talk an email response than to type one on a small laptop keyboard.

Anyway, I wonder if someone will catch on to this, or if it would prove as awkward to use as the iPhone keyboard? Anyone care to comment?


Topics

Comments

Joe, I love this posting!! This is the stuff that will make Apple and independent programmers push their imagination to the limit!!

I am a big fan of the iPhone but I have to agree with you that the lack of a true QWERTY keyboard is a major drawback.

Thanks for the Blog by Dr. Halamka and again I agree with his assessments. While the iPhone has some stunning bandwidth and display capabilities it lacks the necessary data input features.

I really like your thoughts on speech processing and devices like the iPhone! The techo-geek manufacturers are definitely already thinking the same thing...Check out the Jott application for the iPhone which gets pretty close to the killer application you described but it requires too many steps for real-time Radiology? http://jott.com/default.aspx

I know a few imaging vendors that are working on some pretty cool applications for devices like the iPhone but as you indicated VR digital dictation is one that we will have to wait on. I'm sure we won't have to wait too long but I'm also sure it will cost major $$$$.

Joe, thanks for bringing this post together. I did go out to John Halamka's blog post, before commenting.

I'm not as critical of the iPhone's keyboard, but I seldom type even 40 characters at a time. In fact, I dont have an iPhone, I have an iPod Touch. But, it's an obviously amazing device. The combination of WiFi, Safari, Email, and now native applications with a zooming, multi-touch display is just unparalleled. I do use it for light email, around the house.

All that said, in my corporate life, I live in my Blackberry 8830 3G phone. The integration with the Enterprise Server is so seemless, I have no need to explore the iPhone for unified communications. I routinely type messages of arbitrary length on the tactile keyboard. The Bluetooth works well to a headset and as a way to convert the blackberry into an internet "modem" for my laptop. Curiously, there is no Bluetooth functionality in the iPod Touch.

Regarding speech recognition as an input modality, I've been extremely impressed by Dragon Naturally Speaking for the last 5 years. For a limited class of work, it's much faster and easier than typing. Generally though, I'm much happier at a normal-sized keyboard. My thinking-to-speech process is different from my thinking-to-keyboard process.

I can see your point the mobile device technologies are converging rapidly. Soon, we'll all have $300 MacBook Air(s) with multi-touch screens, broadband wireless, we'll phone via broadband with Skype Video conferencing, and we'll use voice, keyboard and screen gestures as defined by what's easiest, fluidly switching between them. I'll have mine in the black case, thank you.

this makes me think of the "information overload issue" that Joe touched on in another post - I think HCI may need to do a story on that sometime soon how CIOs and other strategic IT professionals can avoid succoming to the bombardment of emails and text messages which are now so much a part of business life

 


Joe Marion

Founder and Principal, Healthcare Integration Strategies

Joe Marion

www.hisconsultant.com

Joe Marion is founder and Principal of Healthcare Integration Strategies, specializing in the...