Recent Posts

Pages: [1] 2 3 ... 10
General Chat / Re: Why doesn't nature's cells consume all of Earth?
« Last post by korrelan on Today at 12:13:46 pm »
The Martians had a similar problem back in the early 19th century with red weed; they just gave up trying to control it and invaded another planet lol.

I’m surprised that some entrepreneur hasn’t launched Kudzu burgers or some similar method of profiting from the stuff.

Robotics News / Faster analysis of medical images
« Last post by Tyler on Today at 12:00:52 pm »
Faster analysis of medical images
18 June 2018, 5:00 am

Medical image registration is a common technique that involves overlaying two images, such as magnetic resonance imaging (MRI) scans, to compare and analyze anatomical differences in great detail. If a patient has a brain tumor, for instance, doctors can overlap a brain scan from several months ago onto a more recent scan to analyze small changes in the tumor’s progress.

This process, however, can often take two hours or more, as traditional systems meticulously align each of potentially a million pixels in the combined scans. In a pair of upcoming conference papers, MIT researchers describe a machine-learning algorithm that can register brain scans and other 3-D images more than 1,000 times more quickly using novel learning techniques.

The algorithm works by “learning” while registering thousands of pairs of images. In doing so, it acquires information about how to align images and estimates some optimal alignment parameters. After training, it uses those parameters to map all pixels of one image to another, all at once. This reduces registration time to a minute or two using a normal computer, or less than a second using a GPU with comparable accuracy to state-of-the-art systems.

“The tasks of aligning a brain MRI shouldn’t be that different when you’re aligning one pair of brain MRIs or another,” says co-author on both papers Guha Balakrishnan, a graduate student in MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and Department of Engineering and Computer Science (EECS). “There is information you should be able to carry over in how you do the alignment. If you’re able to learn something from previous image registration, you can do a new task much faster and with the same accuracy.”

The papers are being presented at the Conference on Computer Vision and Pattern Recognition (CVPR), held this week, and at the Medical Image Computing and Computer Assisted Interventions Conference (MICCAI), held in September. Co-authors are: Adrian Dalca, a postdoc at Massachusetts General Hospital and CSAIL; Amy Zhao, a graduate student in CSAIL; Mert R. Sabuncu, a former CSAIL postdoc and now a professor at Cornell University; and John Guttag, the Dugald C. Jackson Professor in Electrical Engineering at MIT.

Retaining information

MRI scans are basically hundreds of stacked 2-D images that form massive 3-D images, called “volumes,” containing a million or more 3-D pixels, called “voxels.” Therefore, it’s very time-consuming to align all voxels in the first volume with those in the second. Moreover, scans can come from different machines and have different spatial orientations, meaning matching voxels is even more computationally complex.

“You have two different images of two different brains, put them on top of each other, and you start wiggling one until one fits the other. Mathematically, this optimization procedure takes a long time,” says Dalca, senior author on the CVPR paper and lead author on the MICCAI paper.

This process becomes particularly slow when analyzing scans from large populations. Neuroscientists analyzing variations in brain structures across hundreds of patients with a particular disease or condition, for instance, could potentially take hundreds of hours.

That’s because those algorithms have one major flaw: They never learn. After each registration, they dismiss all data pertaining to voxel location. “Essentially, they start from scratch given a new pair of images,” Balakrishnan says. “After 100 registrations, you should have learned something from the alignment. That’s what we leverage.”

The researchers’ algorithm, called “VoxelMorph,” is powered by a convolutional neural network (CNN), a machine-learning approach commonly used for image processing. These networks consist of many nodes that process image and other information across several layers of computation.

In the CVPR paper, the researchers trained their algorithm on 7,000 publicly available MRI brain scans and then tested it on 250 additional scans.

During training, brain scans were fed into the algorithm in pairs. Using a CNN and modified computation layer called a spatial transformer, the method captures similarities of voxels in one MRI scan with voxels in the other scan. In doing so, the algorithm learns information about groups of voxels — such as anatomical shapes common to both scans — which it uses to calculate optimized parameters that can be applied to any scan pair.

When fed two new scans, a simple mathematical “function” uses those optimized parameters to rapidly calculate the exact alignment of every voxel in both scans. In short, the algorithm’s CNN component gains all necessary information during training so that, during each new registration, the entire registration can be executed using one, easily computable function evaluation.

The researchers found their algorithm could accurately register all of their 250 test brain scans — those registered after the training set — within two minutes using a traditional central processing unit, and in under one second using a graphics processing unit.

Importantly, the algorithm is “unsupervised,” meaning it doesn’t require additional information beyond image data. Some registration algorithms incorporate CNN models but require a “ground truth,” meaning another traditional algorithm is first run to compute accurate registrations. The researchers’ algorithm maintains its accuracy without that data.

The MICCAI paper develops a refined VoxelMorph algorithm that “says how sure we are about each registration,” Balakrishnan says. It also guarantees the registration “smoothness,” meaning it doesn’t produce folds, holes, or general distortions in the composite image. The paper presents a mathematical model that validates the algorithm’s accuracy using something called a Dice score, a standard metric to evaluate the accuracy of overlapped images. Across 17 brain regions, the refined VoxelMorph algorithm scored the same accuracy as a commonly used state-of-the-art registration algorithm, while providing runtime and methodological improvements.

Beyond brain scans

The speedy algorithm has a wide range of potential applications in addition to analyzing brain scans, the researchers say. MIT colleagues, for instance, are currently running the algorithm on lung images.

The algorithm could also pave the way for image registration during operations. Various scans of different qualities and speeds are currently used before or during some surgeries. But those images are not registered until after the operation. When resecting a brain tumor, for instance, surgeons sometimes scan a patient’s brain before and after surgery to see if they’ve removed all the tumor. If any bit remains, they’re back in the operating room.

With the new algorithm, Dalca says, surgeons could potentially register scans in near real-time, getting a much clearer picture on their progress. “Today, they can’t really overlap the images during surgery, because it will take two hours, and the surgery is ongoing” he says. “However, if it only takes a second, you can imagine that it could be feasible.”

Source: MIT News - CSAIL - Robotics - Computer Science and Artificial Intelligence Laboratory (CSAIL) - Robots - Artificial intelligence

Reprinted with permission of MIT News : MIT News homepage

Use the link at the top of the story to get to the original article.
XKCD Comic / XKCD Comic : Irony Definition
« Last post by Tyler on Today at 12:00:51 pm »
Irony Definition
18 June 2018, 5:00 am

Can you stop glaring at me like that? It makes me feel really ironic.


General Chat / Re: I have a question
« Last post by Art on Today at 11:44:50 am »
Yesterday I was sitting in the office of a notary (I have lost my dear dad some days ago  :( and needed a notary). And you know what?

On his desk there was a monitor (Lenovo). On the wall, behind his chair, a giant television (Sony) displaying the same image (the desktop).

On the desktop: "Windows XP Professonial" was written in a corner, with a lot of "normal" icons on the screen. A shock in my sadness, this guy had a true computer!

Despite the fact this man is young (I guess he's also not poor), and of course French. But he can survive without the terrible Windows 8.

Since yesterday I'm wondering if, if I change my computer (of course I don't need a new computer anymore), I must buy a Macintosh...  ::)

Welcome here.

I too am very sorry to hear of the loss of your father. I lost mine last year and it can be devastating so keep the faith and time will help lessen your burden.,

As for your computer, I like Windows XP but is no longer supported. Windows 10 is a very good OS and it's pretty forgiving as well. Just about everything can run on in in the way of programs and apps.

Realize that while it might sound chic to own an Apple product, the Worldwide Market Share is held by Windows OS and divided among several manufacturers like Lenovo, Dell, HP, Asus, and a few others. Apple holds only 7.4 % of the Market Share!! Very costly software and especially Hardware make it a determining factor for me and others.
One of Apple's main sliver of the pie is their use of "iStuff" in various schools. My grandkids HAD to have iPods, iPhones, iPads, and laptops, so yes the educational market is more geared toward Apple as they got their foot in the door early on in the game.
General Chat / Re: Why doesn't nature's cells consume all of Earth?
« Last post by Art on Today at 11:29:28 am »
Yeah, as if the Kudzu hasn't already done enough damage...

Keep it up and we'll be an all Green country before anyone else!! (but in the wrong way of course).
General Project Discussion / Re: The last invention.
« Last post by korrelan on Today at 11:16:21 am »
Will your AGI discover the solution for cancer by mental kung-foo?

Mental Kung-foo… lol… good name for a chatbot.

We humans like to think of ourselves as intelligent, though how intelligent and indeed the limits of intelligence possible remains to be seen.  At our level of intellect the best method we have devised so far is the ‘scientific method’… and with good reason.

New knowledge has to be based on testable/ reproducible facts; there is no substitution… no assumptions or guesses… cold hard testable facts.  No matter how intelligent a system becomes these methods must be adhered too if advancements are to be realised.  Obviously if the system is intelligent it will realise/ adopt these same methods… they are a logical conclusion/ outcome… a law of nature/ information/ reality.

How intelligent my AGI could become, I have no idea.  There are several factors that limit human intelligence, axon propagation speeds, size of our craniums, etc. Although my AGI won’t be limited by these finite properties there are properties to information/ knowledge its self that could be limiting factors. 

Everything we know and understand about reality has been derived from a human point of view, the limitations imposed on our senses and mental faculties have shaped our understanding of reality and the tools we design to enhance our abilities… we only understand what is observable/ detectable by us.  When we ‘think’ we compare/ calculate at a resolution/ level of abstraction relative to the problem we are trying to solve, again constrained by our experience of reality.  For an AGI to be useful it must experience/ experiment/ design for our reality.

You have probably heard the old saying ‘A jack of all trades is a master of none, but often-times better than a master of one’. 

We have many savants amongst us, people who have extraordinary mental abilities… but it’s always at the deficit of other mental abilities.  I pride myself in being a ‘jack of all trades’, if I was a mathematical savant… there is no way I could have got my research this far. So at our level of intellect too much intelligence in one sphere can become a limiting factor… that will probably hold true for higher levels of intelligence… again its the ‘general’ bit of AGI that counts.

My main goals include self-awareness/ consciousness and human level intellect; the true power won’t come from one single extremely intelligent machine but from a collection of human level machines working in perfect harmony, experimenting and exchanging information seamlessly... tirelessly.

We humans could achieve a similar scenario… but there are too many exposed backs and knives available.

General AI Discussion / Re: I want to ask a question
« Last post by LOCKSUIT on Today at 12:53:20 am »
Wow Art, holdin out that all this time I see! Definitely check that out korrelan!
General AI Discussion / Re: I want to ask a question
« Last post by Art on Today at 12:03:58 am »
It's like a giant Tree of Knowledge (biblical references aside) with the ROOT of the tree containing the main Brain and as each branch needs attention or has to deal with a problem or situation, the ROOT, this CORE of knowledge sents this branch detailed information it needs to be able to confront and deal with the problem.

Our brain is a lot like that...a Command Center dealing with those little things that come up in our lives and assigning a weight of importance to that little issue and deals with it before it grows into a bigger issue. And so it goes, experiencing, learning, digesting, categorizing, determining, judging, ruling, assigning, executing and relaxing and so many more adverbs and verbs. Gaining knowledge based on learned things and experience will help serve as its own reward.

So, One main algorithm overseeing the sum of its parts! (the way I see it).
General AI Discussion / Re: Can artificial intelligence have a soul and religion?
« Last post by Art on June 18, 2018, 11:54:42 pm »
I absolutely agree with your assessment and also regard politics the same way.

Emotional simulation however misplaced, might help in making an AI more believable or connected with its human counterparts.

I don't require religion nor politics in my AI. Common Sense, good judgment, and fair practices are about all that's necessary.

As far as the would probably be the essence of the AI at its very core, what it believes, knows and how it acts based on those traits and abilities.

In as much as it is thought that humans possess some sort of soul, that, to my knowledge, has never been scientifically proven either.
General AI Discussion / Can artificial intelligence have a soul and religion?
« Last post by infurl on June 18, 2018, 11:34:55 pm »

Will we be able to manipulate AI via its programming or its various environments to incline it toward certain religious propensities? Are there limits to “evangelising” AI? Is reprogramming a Christian Siri to be a Muslim Siri off-limits? What happens to AI that refuses to cooperate with the religious fundamentalism of its designer? Is it simply turned off?

A rather long-winded article that it is funny because it takes itself so seriously. Religion is ridiculous, why would you even bother?
Pages: [1] 2 3 ... 10

Users Online

38 Guests, 1 User
Users active in past 15 minutes:
[Trusty Member]

Most Online Today: 45. Most Online Ever: 208 (August 27, 2008, 09:36:30 am)