July 18, 2022
What are latency-sensitive applications and why are they needed for the future?
A latency-sensitive application is a application which is supposed to respond fast on specific events. Latency is defined as the time between the occurrence of an event and its handling. There was a time when it took minutes or even hours for us to send a simple email. But that time is long gone now. Technology has evolved in tune with the requirements of the times. Right from things such as phone calls to the most critical medical and military technology, the ‘time factor’ assumes a pivotal role. Medical imaging is a time-critical application it require real-time low-latency image processing with high throughput is very important. This is where latency sensitivity comes into play. Medical imaging continues to evolve rapidly the latency targets for real-time services remains a constant challenge. Latency is the total delay between the sensor and the receiver. The various levels of latency decides the quality and timely delivery of the images. The medical imaging field sustained considerable growth in terms of technological development, innovation, and market expansion, as a result production of a large-scale of data that probably fully poses diagnostic imaging in the context of big data. The most intensive concern in the medical imaging scenario is moving large gigabytes of studies over the internet is considerable slower. In such application, there is a high demand for real-time, low-latency image processing with high throughput. Decreasing latency is highly desirable, mainly where real-time analysis is essential.
Latency sensitive applications
A number of modern technological applications are termed as latency-sensitive applications. But what does that mean? ‘Latency’ refers to the interval or time gap between actions. Latency sensitive applications mean that these applications simply cannot perform their intended tasks if there is a significant time gap in the processing of signals or information. Most critical technologies are latency-sensitive. A delay in processing signals can end up in deadly consequences with regard to these applications. The compute-critical process of training convolutional neural networks (CNNs) can be enormously upgraded in the cloud, cloud communication introduces the problem of latency which may lead to lagging inference performance in edge devices and mission-critical applications. With the development of real-time AI-based services, latency becomes more and more important facet of inference performance. Not only is high throughput critical, but so is delivering high throughput within a specified latency budget to optimize end-user experience. In medical imaging field, communication latency is a critical issue that still not properly resolved by many big organizations in the business.
Medical Imaging and Latency Sensitivity
Medical imaging is a critical technology which ought to be latency-sensitive. Medical image processing has been established as a important field of innovation in modern health care combining medical informatics, neuro-informatics and bioinformatics. Modern imaging entirely establish intricacy of the attainable dataset, encircling different imaging techniques, each with several advantages and drawbacks. The speed of data generation as well as data collection interprets a typical workflow that is related to the diagnostic imaging procedure. Millions of diagnostic images are scanned on a daily basis for the patients who undergo diagnostic imaging procedures, that consists of a clear workflow, the scanned image is acquired through the imaging device, stored in Picture Archiving and Communication System (PACS), these images are visually inspected on a DICOM viewer by a radiologist. PACS infrastructure works out high cost as it needs dedicated server room, temperature controlled, high cost of electricity, 24X7 Server maintenance staff, upgradation cost, civil work including wiring across network in Imaging centers and hospitals. Over the years RIS PACS becomes easily outdated because of the limited infrastructure scalability. Beyond this conventional obsolete workflow, new technology must be considered like teleradiology, cloud computing of complex data with AI capabilities. Sharing medical imaging data is in itself a great task owing to the complexity of the images and the diagnostic standards which have to be adhered to. In addition to this comes in the unavoidable aspect of latency sensitivity. Hence the whole process becomes way more complex. Any efficient medical imaging technology or appliance should be able to address all these considerations. The failure to do so can lead to catastrophic consequences.
LifeVoxel – setting the standard for latency-sensitive applications
LifeVoxel is a major emerging player in the field of AI-based medical imaging techniques. The company uses GPU-powered supercomputing cloud technology in this domain. Its efforts have already been recognized with several patents and partial funding for Research and Development by the National Science Foundation for Intelligent Visualization. LifeVoxel’s segmented viewing of high quality imaging enables the detection of abnormalities occurring within one’s body at initial stage, proactive detection of diseases can ensure better cure in early stages. Interactive visualization is the key, LifeVoxel is a visualization and workflow solution that delivers and stores diagnostic quality images. LifeVoxel has set the benchmark regarding the aspect of latency-sensitive applications. The degree of success can be gauged by the fact that while top-rated content delivery networks have a latency of 0.5 to 3 seconds, LifeVoxel is able to consistently provide a latency of a mere 0.01 seconds. This is done with the help of predictive buffering. Graphics Processing Units (GUPs) are made use of by the firm in order to offer seamless interactive viewing over the cloud. Dynamic transcoding using evolutionary adaptations ensure image quality while increasing the speed of sharing and viewing. LifeVoxel is thus able to provide near-zero latency in such a critical technology. Unlike conventional PACS which 20 year old technology, LifeVoxel platform has interactive streaming, which will be the future on 3D Big data visualization.
Latency sensitive applications are going to play a prominent role in the coming years. From self-driven cars to remote location guided surgeries, latency-sensitive technologies of the top quality are going to be in great demand. A lot can be learnt from the innovative techniques utilized by firms such as LifeVoxel to provide top-notch latency. GPU cloud predictive intelligent streaming overcomes large data access speed and latency over the internet. LifeVoxel uses GPU to manipulate GB of patient data remotely without transmitting data to the end user. It allows to access visualizations on any device on-demand and in real time. Streaming of visualization has been done by predicting next frames. Fast FPS from GPU has been enabled for discarding incorrectly predicted frames and generating the new ones. Predicted frames are buffered to clients for overcoming latency. Artificial Neural Network is used for pattern recognition. Metaheuristic or Genetic algorithms are used for implementing Heuristic search. The fact that it does this with really complex data such as medical images is evidence of the powerful computational methods and techniques used.
Most of the cloud based imaging services don’t provide diagnostic quality images, and the ones who claim to provide do with a heavy lag time, and very slow process. LifeVoxel has the ability to quickly process and transmit high quality diagnostic images without any delay. LifeVoxel meets today’s medical environment demands, by providing efficient, cost-effective workflow tools that can empower faster and more accurate diagnosis within an extremely affordable fee structure. More investment should be made in the Research and Develop for latency-sensitive applications recognizing the role that this technology would have to play in the near future. Successful developments should be further popularized and further built upon to take latency sensitivity to new and promising fields.
 Stream processing and AI share the same qualities that make them unique but also share the same challenges and bottlenecks in their adoption. How can enterprises overcome these challenges and make the right investments and actions related to their data architecture?
 The RIS PACS is constantly storing and retrieving images to and from the storage system. This includes tasks like pre-fetch, short-term transfer, long-term transfer, and more. Low latency is required for any action between the RIS PACS and the storage system, because higher latency will cause the entire system to become sluggish. It’s recommended to monitor this latency and set thresholds for when any of the RIS PACS tasks take too long.
 Decreasing Latency for live streams and increasing reliance on Artificial intelligence (AI) will be the most significant new tech trends of 2020, according to a survey of 540+ industry experts from 100+ countries. The survey – the Bitmovin Video Developer Report – provides insights into the evolving technology trends of the digital video industry.
It’s the third year for the report, and, consistent with last year’s results, latency is the most significant issue faced by more than half (54%) of video developers, closely followed by getting playback on all devices (41%). More than half (53%) of respondents said they expect to achieve live streaming latency of less than five seconds, while 30% believe hitting under one second latency could be possible. Meanwhile, over half (56%) of those taking part in the survey expect to start using AI/ML (Machine Learning) by the end of 2020.
Where to now?
For more information on this article or to view our software in action, please don't hesitate to schedule a virtual tour.
 https://blog.paessler.com/healthcare-it-4-ways-to-monitor-a-pacs By Shaun Behrens Jul 19, 2019