MICSI - Higher resolution MRI with faster scan times

Unlocking the Future of MRI: MICSI's Revolution in Medical Imaging

In the world of modern medicine, diagnostic accuracy is often the key to effective treatment and improved patient outcomes. Among the many tools at a physician's disposal, Magnetic Resonance Imaging (MRI) has long been a cornerstone in the realm of non-invasive diagnostics. However, the limitations of MRI technology, including long scan times and lower resolutions, have posed challenges for both patients and healthcare providers. MICSI, a groundbreaking startup founded in 2018, has set out to transform the landscape of MRI imaging. With their innovative AI software, MICSI is poised to double resolution, halve scan times, and revolutionize the field of medical imaging. But how exactly does MICSI plan to accomplish this? Let's delve into their vision, technology, and the potential impact on the healthcare industry.

Who Are the Visionaries Behind MICSI?

To truly understand the innovation driving MICSI, we must first get to know the individuals behind this groundbreaking startup. The co-founders, Gregory Lemberskiy and Benjamin Ades-Aron, bring a wealth of knowledge and expertise to the table, making them the driving force behind MICSI's vision.

Gregory Lemberskiy: Pioneering MRI Advancements

As the CEO of MICSI, Gregory Lemberskiy's journey began with a passion for bringing cutting-edge image processing tools from the research setting into clinical practice. Gregory's background includes a PhD in experimental physics, which he earned in 2019. During his doctoral work, he focused on developing algorithms for image enhancement and biophysical modeling of the MRI signal. These innovative methods have already found applications in improving image quality on low-field MRI systems and characterizing the physical properties of the prostate glandular lumen.

Benjamin Ades-Aron: The Mastermind of Computer Vision

Benjamin Ades-Aron, another integral part of the MICSI team, holds a PhD in electrical engineering, which he obtained in 2022 from NYU. His specialization in developing machine learning-based computer vision algorithms for MRI data visualization has positioned him as a leader in the field. Benjamin excels in developing software to process and route diffusion and functional MRI data, with a primary goal of enhancing image signal, reducing noise, and creating clinically viable diagnostic imaging biomarkers.

Together, Gregory and Benjamin have spent over a decade developing machine learning software aimed at improving the quality and diagnostic utility of MRI. Their expertise, combined with a shared vision for the future of MRI, is the driving force behind MICSI's mission to make MRI quantitative and reproducible, ultimately enhancing patient satisfaction and outcomes.

The MICSI Launch: A Quantum Leap in MRI Technology

MICSI, short for Microstructure Imaging, has embarked on an ambitious journey to bring higher resolution MRI scans with significantly faster scan times to the medical world. Their initial offering serves as a stepping stone towards a larger vision: transforming MRI into a truly quantitative instrument capable of providing highly reproducible data for more accurate diagnoses and patient management. But what problems does MICSI aim to solve, and how do they plan to do it?

Problem: MRI Centers are Overbooked

For anyone who has experienced the process of getting an MRI, the common grievances include waiting weeks for an available slot and enduring lengthy scan times. These challenges not only inconvenience patients but also contribute to hospital backlogs and appointment delays. Such delays can have critical implications for conditions that require prompt diagnosis and intervention, leading to increased healthcare costs, heightened patient anxiety, and potentially compromised treatment strategies.

Compromise: Why Are MRI Scans So Time-Consuming?

The seemingly excessive duration of MRI scans primarily stems from the need to preserve signal-to-noise ratio (SNR) to produce high-quality images. Conventional SNR-preserving methods, such as line-by-line acquisitions and image averaging, significantly extend scan times. For instance, a typical MRI scan of the whole brain could be completed in seconds if not for the focus on SNR preservation.

Additionally, there's a compromise in spatial resolution, as MRI captures data in 3D using "voxels." Doubling resolution in each dimension requires an eight-fold increase in SNR, a challenge that has proven difficult to overcome even with decades of innovation.

Solution: Boosting SNR with MICSI-RMT

MICSI's revolutionary solution to these challenges lies in their patented approach, MICSI-RMT, which is currently pending FDA510k clearance (anticipated in Q4-2023). MICSI-RMT employs a self-supervised AI method to learn the noise properties of MRI exams and uniformly remove noise across all images in the dataset. This method can be likened to a "smart averaging" approach, where multiple MRI images are combined while preserving their unique image properties and eliminating noise. The result is a significant boost in SNR, approximately proportional to the square root of the number of images included in the dataset and inversely proportional to their dissimilarity in physical properties and artifacts.

MICSI-RMT is the commercial embodiment of the MP-PCA algorithm, developed by MICSI's research group led by Drs. Dmitry S. Novikov and Els Fieremans at NYU Radiology's Center for Biomedical Imaging. Notably, MP-PCA is the most popular denoising approach in the MRI community, with the original paper accumulating over 1300 citations since its publication in November 2016.

The Impact of MICSI-RMT: A Paradigm Shift in MRI

With MICSI-RMT, MRI centers are poised to provide not only higher quality imaging but also drastically reduced scan times, up to 50% faster. This improved efficiency will enable centers to scan more patients per day, significantly enhancing their operational capacity and throughput. From a financial perspective, this boost in capacity translates into substantial revenue increases for every MRI center, estimated at an additional $2 million in annual revenue per MRI scanner. Moreover, the enhanced image quality and resolution promise more accurate diagnoses, early disease detection, and the potential to reduce overall healthcare costs over time.

Conclusion: The MICSI Revolution

As MICSI continues to push the boundaries of MRI technology with their innovative AI-driven solutions, the future of medical imaging looks brighter than ever. The startup's mission to enhance the quality and efficiency of MRI scans has the potential to revolutionize healthcare, making diagnostic imaging more accessible, accurate, and patient-friendly. Gregory Lemberskiy and Benjamin Ades-Aron's dedication to bridging the gap between cutting-edge research and clinical practice serves as an inspiring example of how innovation can transform an entire industry.

In the coming years, we can expect MICSI to play a pivotal role in reshaping the landscape of medical imaging, offering hope to patients, healthcare providers, and the broader healthcare industry as a whole. As they work toward FDA clearance and the widespread adoption of their technology, we anticipate a future where MRI scans are not just faster and more precise but also more affordable and readily available to those in need. MICSI's journey is a testament to the power of innovation to drive progress and improve the quality of healthcare for everyone.