Our Projects

CCF PI: Jun Qin
IBM PI: Gavin Jones

CCF PI: Daniel Blankenberg
IBM PI: Omar Shehab

This is a federal award from the United States Defense Advanced Research Projects Agency (DARPA) to develop a quantum computing benchmarking framework for quantum computing architectures that can overcome any failures in hardware or software. It will lead to the development and analysis of various computational analysis problems to test and evaluate scenarios relevant to a variety of fields — including genomics, computational biology, protein conformational analysis, protein-drug interaction, radiation therapy planning, healthcare process optimization, computational biology and computational chemistry.

CCF PI: Mina Chung
IBM PI: Omar Shehab

This project will explore the benefits of quantum computing to study the risk factors for cardiovascular complications following non-cardiac surgery. Major adverse cardiac events are a leading cause of non-cardiac morbidity and mortality around the time of surgery. By applying quantum machine learning techniques, we can enhance our understanding of the risk factors, improve patient outcomes and potentially reduce healthcare costs and utilization by facilitating clinical decision-making. The results of this project can also help determine if quantum computing-based machine learning algorithms can improve patient outcomes in other clinical settings.

CCF PI: Timothy Chan
IBM PI: Sara Capponi & Laxmi Parida

Cleveland Clinic and IBM are combining artificial intelligence and quantum computing to advance a key biomedical task in cancer immunotherapy design. The project’s goal is to engineer optimal immune T-cells and immune T-cell receptors that are highly efficient at killing cancer cells. The team will achieve this by combining knowledge of classical and quantum computing algorithms with Cleveland Clinic’s novel experimental data sets.

CCF PI: Kenneth Merz & Susanta Das
IBM PI: Joseph Morrone

CCF PI: Xiaojuan Li & Mihriban Karaayvaz
IBM PI: Omar Shehab

This project aims to leverage the power of quantum computing algorithms to optimize feature identification and feature selection in healthcare and life sciences (HCLS) data sets. Cleveland Clinic and IBM will identify optimal ways to represent HCLS datasets using fewer features. This will help to represent noisy datasets that are common bottlenecks for performance. By creating a better understanding of this problem in the context of quantum computing, we will help guide future research efforts, as well as hardware and software development. A deeper understanding of this field will also help with the improvement of classical AI algorithms.

CCF PI: Peter Mazzone
IBM PI: Laxmi Parida & Hokan Doga

Cleveland Clinic and IBM are evaluating the potential for quantum computing to develop early lung cancer detection with blood tests and other easy-to-use biomarkers.

Early detection of lung cancer improves patient outcomes. Despite this, the success of lung cancer screening is limited by strict eligibility criteria, slow uptake, poor adherence to annual screening and the impact of false positives (e.g. lung nodules) on patient outcomes. This project will explore an easy-to-use, accurate and inexpensive molecular biomarker (e.g. blood test) to improve lung cancer screening outcomes. Cleveland Clinic and IBM are evaluating the potential for quantum computing to improve the accuracy of early lung cancer detection biomarkers.

CCF PI: Sandip Vasavada & Glenn Werneburg
IBM PI: Laxmi Parida & Kahn Rhrissorrakrai

Cleveland Clinic and IBM are combining artificial intelligence with quantum computing to advance a key biomedical task for prediction of antibiotic resistance. Antibiotic resistance is rampant and increasing worldwide, and a main driver of resistance is the inappropriate use and overuse of antibiotics. This project aims to create a model using utility-scale quantum machine learning to predict antibiotic resistance for urinary tract infections. This model can help decrease the amount of antibiotics unnecessarily prescribed. It will also be one of the first utility-scale quantum experiments for healthcare and life science applications using a targeted 50-100 qubits.

CCF PI: Kenneth Merz & Timothy Chan
IBM PI: Gavin Jones

This project aims to perform quantum simulations to study chemical reactions in important biological processes. Through simulations, we may develop a deeper understanding of biomedical reactions that can lead to human disease. Currently, performing these types of simulations is quite limited by existing computing technology. Cleveland Clinic and IBM plan to address this by developing new quantum simulation algorithms as the first of their kind to be demonstrated on a quantum computer. One of the most significant benefits of this work is that it will allow researchers to stop relying on low-accuracy methods and instead make use of more precise quantum simulations that can potentially lead to new therapies.

CCF PI: Nancy Foldvary
IBM PI: Erhan Bilal & Jeff Rogers

This project aims to improve the apnea-hypopnea index (AHI), which is a diagnostic tool for determining the presence and severity of obstructive sleep apnea. Currently, the AHI depends on manual scoring that can result in different findings based on the observer and is subject to night-to-night variability. The study's goal is to create an alternative approach to obstructive sleep apnea diagnosis using modern self-supervised deep learning models. These models will be more efficient, reliable and informative. They may also be extended to gather data from non-medical grade sleep sensors like smartwatches, smart rings and others.

CCF PI: Thaddeus Stappenbeck
IBM PI: Bo Wen & Jeff Rogers

A longitudinal study of disease phenotypes and signatures of diverse Crohn’s -IBM and Cleveland Clinic are conducting a digital health study to understand how social determinants of health, such as education and neighborhood, impact Crohn’s disease (CD) progression. The study will focus on diverse populations that are experiencing an increase in CD diagnosis. To better understand patients’ symptoms, the IBM Accelerated Discovery Platform will monitor patients’ symptoms daily and combine the results with their clinical assessments. Additionally, an AI-powered data collection technology will be created to monitor other patient and environmental factors. This will provide a better understanding of how social determinants of health affect disease.

CCF PI: M. Cecilia Lansang
IBM PI: Vince Siu & Jeff Rogers

Cleveland Clinic and IBM are using data from electronic health records (EHRs), clinical notes and continuous glucose monitors (CGM) to develop experimental prototype artificial intelligence (AI) models that predict insulin dose adjustment in kidney transplant patients on a prednisone taper.

About 20,000 kidney transplants occur annually in the United States, with diabetes listed as the most common clinical condition listed for kidney transplant (31%). The Cleveland Clinic Kidney Transplant Program reduced the length of stay (LOS) for kidney transplant patients to three days. However, post-transplant readmissions or emergency room visits for hyper- and hypoglycemia still occur. To prevent this, Cleveland Clinic and IBM are using data from electronic health records (EHRs), clinical notes and continuous glucose monitors (CGM) to develop experimental prototype artificial intelligence/machine learning (AI/ML) models designed to predict insulin dose adjustment in kidney transplant patients on prednisone taper.

CCF PI: Willy M. Valencia Rodrigo
IBM PI: Vince Siu

Cleveland Clinc and IBM are leveraging IBM's Digital Health platform and analytics capabilities to develop a digital tool that can help validate an efficient and reliable screening procedure for geriatric evaluations. This tool will help to properly determine glycaemic targets, prevent diabetes complications and tailor treatment intensity and associated risks. The goal is to provide results that physicians can use to adjust their treatment plan.

CCF PI: Animesh Tandon
IBM PI: Kely Norel

Severe congenital heart disease (CHD) and prematurity can cause physiological instability in infants. In an outpatient setting, these high-risk infants can suffer from clinical deterioration leading to hospital readmission, morbidity and mortality. Wearable biosensors (wearables) have been developed to monitor infants, creating an opportunity to collect continuous physiological data (CPD) from home. Cleveland Clinic and IBM will explore the use of AI to predict low cardiac output syndrome and other relevant outcomes through wearable-derived analytics for high-risk infants.

CCF PI: James Liao
IBM PI: Erhan Bilal

Cleveland Clinic and IBM are using modern computer vision techniques with deep learning to analyze videos of patients’ gaits to detect normal pressure hydrocephalus (NPH) and predict patients’ responses to treatment.

Normal pressure hydrocephalus (NPH) is one of the few causes of dementia that can be controlled or reversed with surgical treatment. Gait impairment is an early NPH symptom, but it is often undiagnosed or misdiagnosed with other neurodegenerative diseases due to the lack of quantitative gait analysis systems. Cleveland Clinic and IBM are working on collecting gait parameters from videos by using modern computer vision techniques with deep learning to identify gait markers for detecting NPH patients and predict their treatment responses. The goal of this work is to explore the predictive capability of IBM’s video-based gait analysis for long-term symptom improvement after a standard-of-care shunt surgery.

CCF PI: Andreas Alexopoulos
IBM PI: Bing Dang

Cleveland Clinic and IBM are studying the detection of certain types of seizures for adult patients based on Cleveland Clinic data and smartwatches. The unpredictability of epileptic seizures exposes individuals with epilepsy to potential physical harm, restricts day-to-day activities and impacts mental well-being. In routine clinical practice, seizure tracking that guides epilepsy management relies on subjective patient and caregiver recall and self-reporting, which are known to be extremely unreliable. Wearable electronic devices may be used to collect and monitor health-related data (e.g., monitor and detect seizures), which can help with managing the condition and improving the patient's quality of life.

CCF PI: Giuseppe Sautto
IBM PI: Wendy Cornell

Cleveland Clinic and IBM are using a biomedical foundation model (BMFM) framework to develop next-generation vaccines and monoclonal antibodies (mAbs) that can prevent infectious disease.

Cleveland Clinic and IBM are using a biomedical foundation model (BMFM) framework to develop next-generation vaccines and monoclonal antibodies (mAbs) to prevent infectious diseases. The team will use both public and Cleveland Clinic data sets of influenza vaccine strains and mAbs to predict the range of immune responses elicited by vaccination. The overall goal of this project is to develop cutting-edge next-generation vaccines and to broaden the antibody response against diverse influenza strains. Importantly, the methodologies that will be developed in this project could also be applied to other pathogens and infectious diseases, especially in the context of the pandemic preparedness.

CCF PI: Timothy Chan
IBM PI: Jeff Weber & Wendy Cornell

This project builds on a previous study that detailed how the immune system recognizes immunotherapy targets through artificial intelligence. The team will apply biomedical foundation models (BMFM) to expand on this work, identifying potential vaccine targets and generating therapeutic T-cell receptors. The goal of the project is to develop the next generation of immunotherapies through in silico evolution.

CCF PI: Shaun Stauffer
IBM PI: Wendy Cornell

This project will be a unique opportunity for Cleveland Clinic to work with IBM’s computational experts, advanced computing platforms and molecular modeling tools for small molecule discovery/optimization in the areas of COVID antiviral drug development and inhibitors of a protein implicated in cancer development. The project will have chemists, structural biologists and computational scientists working together as a closely aligned team. The effort, if successful, will result in discoveries in chemistry that can ultimately lead towards novel drug development candidates for clinical testing.

CCF PI: Vara Prasad Josyula
IBM PI: Wendy Cornell

Cleveland Clinic and IBM using biomedical foundation models (BMFM) and medical chemistry techniques to identify non-peptide inhibitors suitable for the treatment of osteoarthritis.

The goal of this project is to use the generative and predictive capabilities of IBM’s biomedical foundation model (BMFM), in tandem with Cleveland Clinic’s medicinal chemistry techniques and experiments, to identify and develop effective and selective non-peptidic inhibitors of HtrA1 suitable for treatment of osteoarthritis (OA) by intra-articular administration to affected joints. The team will use their skills to expedite the discovery of small molecules that bind to the target protease with high binding affinity. If successful, the team will identify potent compounds with drug-like properties for clinical studies that could treat OA in a novel way, i.e., disease modification.

CCF PI: Feixiong Cheng
IBM PI: Yishai Shimoni

IBM and Cleveland Clinic aim to identify existing drugs that have the potential of improving management, slowing deterioration and/or delaying the onset for neurodegenerative diseases including Alzheimer’s disease. This study will leverage network medicine analysis technologies of Cleveland Clinic and IBM’s prototype toolset called the drug repurposing engine (IBM DRE) which will be applied to a large data set. The team will focus on Alzheimer’s disease as the first instance to show acceleration of discovery with the combined tools.

CCF PI: Thaddeus Stappenbeck & Julia Liu
IBM PI: Joao Bettencourt-Silva & Vibha Anand

Physicians and researchers are struggling to stay up to date with the vast amount of scientific information being published each year. There is a need to identify and summarize what really matters from the data so that new information can be incorporated into clinical practice and drive new research avenues. Cleveland Clinic and IBM aim to use artificial intelligence to enhance the way in which researchers and physicians search and review large amounts of documents from multiple sources, such as scientific literature or clinical trial protocols. The team will build tools to help identify knowledge gaps, generate new hypotheses and extract hidden factors (e.g., clinical, social, biological) that may be important for the way in which physicians study and treat IBD.

CCF PI: Larisa Tereshchenko
IBM PI: Ching-Huei Tsou

Cleveland Clinic and IBM are developing a deep-learning electrocardiogram (ECG) model capable of diagnosing cardiac arrhythmias beat-by-beat (on each cardiac beat).

This project is developing a deep-learning electrocardiogram (ECG) model capable of diagnosing cardiac arrhythmias beat-by-beat (on each cardiac beat). This AI-ECG model will permit accurate diagnosis of cardiac arrhythmias in medically underserved populations and improve the efficiency of ECG monitoring in hospitals/intensive care units and clinical and research laboratories.

CCF PI: Thaddeus Stappenbeck
IBM PI: Vibha Anand & Joao Bettencourt-Silva

This project will study the cellular malfunctions that contribute to Inflammatory Bowel Disease (IBD), Ulcerative Colitis (UC) and Crohn’s Disease (CD). Using advanced computing methods, including Biomedical Foundational Models (BMFM), the team will identify targets that are potential drivers of these diseases. This model will be trained to identify individual cells as either healthy, disease or pre-diseased. The findings from this will help provide new insights into target discovery and cell-to-cell interactions.

CCF PI: Feixiong Cheng
IBM PI: Joseph Morrone & Parth Suryanarayanan

IBM and Cleveland Clinic are building and validating biomedical foundation models (BMFM) for small molecule drug discovery. This project aims to build a BMFM that is pretrained from multiple views/modalities of small molecule data. This data can be used for a wide variety of downstream predictive tasks in computer-aided drug discovery. The BMFM will then be fine-tuned and validated on relevant tasks, including those related to the discovery of Alzheimer’s disease therapeutics.

CCF PI: Christopher Weight
IBM PI: Michal Ozery-Flato

Cleveland Clinic and IBM are using a multiomics approach and machine learning to predict outcomes for patients with renal tumors.

Cleveland Clinic and IBM are using a multiomics approach and machine learning to predict outcomes for patients with renal tumors. To achieve this goal, the team will incorporate pathology, radiology and clinical data.

CCF PI: Animesh Tandon
IBM PI: BC Kwon

Patients with congenital heart disease receive a variety of diagnostic tests to help determine the next steps for their treatment. Tests include imaging studies, electrocardiograms(ECGs) and exercise testing. Analyzing the results is a challenging task generally done in the clinician’s head. This project will combine various test results including cardiovascular MRI and ECGs with novel machine learning techniques to improve predictions of outcomes for patients with congenital heart disease, initially focusing on a repaired tetralogy of Fallot (rTOF).

CCF PI: Thaddeus Stappenbeck
IBM PI: Vibha Anand & Joao Bettencourt-Silva

Two major forms of inflammatory bowel disease (IBD) are Crohn’s Disease and ulcerative colitis. While they are similar, their symptoms differ among patients based on race, ethnicity, disease location at onset and diagnosis, as well as varying treatment responses. There is an urgent need for more accurate identification of the IBD sub-types to effectively provide therapeutic interventions and guide the development of new therapies. To achieve this, the team aimed to: (1) Generate hypotheses about novel disease progression sub-groups and (2) Define characteristics of those who respond to existing drug therapeutics while focusing on specific clinical outcomes. The data-driven approach provided novel insights and generated new hypotheses for future IBD drug discovery and clinical trial enhancement, potentially using data from EMR.

CCF PI: Lara Jehi
IBM PI: Yishai Shimoni

This project will explore a very large national data set that contains information from insurance claims to see if certain medications that reduce inflammation have also reduced the occurrence of seizures in epilepsy patients after brain surgery. The team will then confirm any information discovered by studying a smaller patient population from Cleveland Clinic where detailed clinical information is available.

CCF PI: Xiaojuan Li
IBM PI: Simona Rabinovici-Cohen

This project will develop pre-trained models as the basis of a foundation model for medical image analysis. The foundation model will be able to identify various features and patterns to make diagnosis assessments. The team will focus on knee MRIs in the first stage, before expanding to multiple body parts (neural, cardiac, breast, musculoskeletal, etc.) and multiple modalities (X-ray, CT etc.). The resulting models will significantly impact clinical analysis and enable downstream applications for automated medical image analysis.

CCF PI: Tara Karamlou
IBM PI: Peter Staar

IBM will provide Cleveland Clinic with IBM Deep Search technology with the aim of enabling researchers to search for, and within, medical studies and publications. Additionally, IBM aims to further enhance and develop Deep Search to improve its functionalities to allow Cleveland Clinic to more efficiently utilize published literature and datasets to design clinical trials, develop clinical prediction tools, and identify patterns/associations.

CCF PI: Bruce Trapp
IBM PI: Ching-Huei Tsou

Cleveland Clinic and IBM are using pre-trained biomedical foundation models (BMFM) to identify novel therapeutic targets for multiple sclerosis (MS) in the drug discovery process.

Cleveland Clinic and IBM are studying target identification for multiple sclerosis (MS). The team will use IBM’s computational methods and pre-trained biomedical foundation models (BMFM) to identify novel therapeutic targets in the drug discovery process. IBM will fine-tune the pre-trained foundation model using public data and data from Cleveland Clinic for target discovery in MS. The project will focus on targets related to a novel Cleveland Clinic compound designed to enhance remyelination in individuals with multiple sclerosis.

CCF PI: Timothy Chan
IBM PI: Jeff Weber

This project will find new insights and develop new technologies that accelerate next-generation cancer immunotherapy discovery. Cleveland Clinic and IBM will use artificial intelligence and molecular simulations to develop better ways to identify targets for cancer vaccines. By creating and leveraging a combination of mechanistic simulations and AI models, we will expand our knowledge related to immunotherapy, vaccine design and clinical response to cancer immunotherapies.

CCF PI: Lara Jehi
IBM PI: Ehud Karavani

Cleveland Clinic and IBM are working to improve predictions of which patients are more prone to seizure recurrence after cranial surgery in epilepsy by using RNA sequencing expression data that can drive or predict patients’ response.

Resection surgery is a last resort treatment for stopping seizures in patients with uncontrolled and debilitating epilepsy, in which the focal of epileptic seizures is being removed. While this treatment usually provides short-term cessation of seizures, a significant proportion of patients suffer a recurrence of seizures after a few years. Cleveland Clinic and IBM are working to improve predictions of seizure recurrence after cranial surgery in epilepsy by using RNA sequencing expression data to identify driving biomarkers. The goal of this study is to predict which patients are more prone to recurrence by leveraging RNA expression markers that drive or predict patients’ response.

CCF PI: Feng Lin
IBM PI: Leili Zhang & Pablo Meyer Rojas

Cleveland Clinic and IBM are working to fine-tune AI foundation model technology to design nanobodies that selectively inhibit the catalytic activities of complement convertases based on their 3D structures.

CCF PI: Feng Lin
IBM PI: Ching-Huei Tsou

Cleveland Clinic and IBM are studying drug target identification using advanced computational models and pre-trained biomedical foundation models developed by IBM for age-related macular degeneration (AMD). The goal of this project is to advance the field of identifying novel therapeutic and diagnostic targets in the drug discovery process. IBM will fine-tune the pre-trained foundation model using public data for target discovery in specific diseases of interest with a focus on the complement proteins as potential targets. Cleveland Clinic will evaluate the model-predicted candidate targets.

CCF PI: Jianjun Zhao
IBM PI: Simona Rabinovici-Coher

CCF PI: Christine Moravec
IBM PI: Joel Chudow

IBM and Cleveland Clinic will launch and evaluate a self-paced online learning journey focused on AI and Data Science for a group of medical students, PhD candidates, nursing students and postdoctoral researchers. This program will utilize IBM-developed courses and educational and training assets. The goal is to evaluate the relevance of IBM’s content on AI and data science for student and researcher skills training and to benefit their research and future goals.

CCF PI: Daniel Blankenberg
IBM PI: Borja Peropadre

IBM and Cleveland Clinic will develop educational resources to instruct Cleveland Clinic’s key researchers and technical personnel about the fundamental aspects of quantum computing and enable them to utilize IBM’s quantum computers via the Quantum Information Science Kit framework (Qiskit), a programming interface to utilize quantum computers. The goal of this initiative is to prepare researchers to do state-of-the-art research on quantum computing applications in healthcare.