Accelerating healthcare diagnostics with Intel oneAPI and AI

Accelerating healthcare diagnostics with Intel oneAPI and AI

The use of Artificial Intelligence in healthcare is booming. According to an Intel study, 84% of US healthcare leaders have begun or are planning to use AI since the COVID-19 pandemic.

AI offers promising possibilities for improving healthcare, preventing disease and saving lives. It can also relieve healthcare professionals of routine tasks, enabling them to focus more on patients and research. In fields such as medicine and pharmaceutical research, Deep Learning and Machine Learning are greatly enhancing personalized care and the patient experience.

The use of AI to accelerate medical diagnostics relies on complex computational models. These require high-performance systems, particularly in terms of HPC calculations, combining CPUs, GPUs and other specialized processors. Explore below four examples of applications that have benefited from the efficiency of Intel oneAPI tools and AI.

GE Healthcare solutions accelerated by Intel oneAPI toolkits

GE Healthcare has worked with Intel to improve the performance of its medical imaging equipment (X-rays, scanners, etc.) using Intel oneAPI toolkits and Intel Xeon processors.

The company estimates that this will save millions of dollars in configuration costs and engineering effort. In the future, it plans to extend the use of oneAPI and DPC++ in its R&D activities.

Mental health diagnosis: Hipposcreen

With electroencephalogram (EGG) signal processing and AI technology at its core, HippoScreen develops diagnostic tools for psychiatric disorders and diseases.

Thanks to Intel oneAPI Base and AI Analytics Toolkit, this Taiwanese start-up has succeeded in accelerating by 2.4 times the construction times of the Deep Learning models used in its Brain Waves AI system, thus considerably improving their efficiency.

Cervical cancer detection: KFBIO

KFBIO, a provider of complete pathology solutions, uses Deep Learning algorithms to improve cervical cancer screening.

Thanks to the use of the Intel Distribution of OpenVINO toolkit and Intel Xeon processors, inference performance for cancer screening has increased by a factor of 8.4.

Lung disease detection: Accrad

South African company Accrad, has successfully developed a medical AI solution, CheXRad, capable of labeling certain pathologies in chest X-rays up to 160 times faster than radiologists, with comparable levels of accuracy, sensitivity and specificity.

Using Intel oneAPI AI Analytics, Intel Distribution of OpenVINO and Intel Developer Cloud for oneAPI, Accrad was able to train, optimize and deploy its Machine Learning model faster and more cost- effectively.

To see the full article, click here.