Remote Robotic Ultrasound Platform

An echocardiogram is a noninvasive and clinically necessary healthcare test that allows healthcare providers to assess for both urgent, life-threatening conditions and monitor chronic cardiac disease progression. Patient demand for this cardiac ultrasound is rising, driven by an aging population and increasing cardiac ailments. However, the accessibility to this diagnostic tool is hampered by several factors:

1. Expertise Scarcity: Cardiac sonographers, those trained to conduct echocardiograms, are in short supply. This has led to long waiting times for patients, particularly in areas underserved by healthcare professionals.
2. Consistency in Imaging: Echocardiograms are operator-dependent. Variability in operator skill can lead to inconsistencies in image quality, affecting diagnosis accuracy.
3. Geographical Disparities: Rural and remote regions globally often lack the facilities and professionals required for echocardiogram procedures. This creates a healthcare disparity where certain populations are disadvantaged due to geographical location.

Our team is working with US communities, urban and rural, to build a solution that caters to both customer groups. Realizing that there is great potential to improve this problem beyond US boarders is what has inspired us to apply with MIT Solve.

Scale and Statistics:

• As of 2023, with the increasing patient population over the age of 50, the global diagnostic ultrasound market has been valued at around $9.3 billion.
• In the U.S. alone, approximately 7,000,000 echocardiograms are performed annually, and this is just a fraction of the global need.
• The World Health Organization reports that up to 40% of global populations lack access to essential diagnostic services, including echocardiography.

Factors Relating to Our Solution:

Our solution specifically targets the shortage of specialized personnel and the inconsistencies in echocardiogram imaging:

1. Automated Assistance: our device’s robotic arm is guided by state-of-the-art AI algorithms, offers consistent, high-quality imaging, and reduces the variability seen with human operators.
2. Telehealth Integration: its teleoperation capabilities allows experts from urban centers can remotely guide or review the echocardiogram procedure, bridging the accessibility gap for those in remote or underserved locations.
3. Training and Scaling: our device has the potential to function as a training tool, aiding the next generation of cardiac sonographers in mastering the intricacies of echocardiogram imaging.

Our mission is not just to bring state-of-the-art technology to echocardiography but to ensure that every individual, regardless of their location, has access to this critical diagnostic tool.

Our solution a tele-robotic tool for remote heart imaging. Imagine a device that can grip and ultrasound transducer and is equipped with two way communication, where the cardiac sonographer can see and speak to the patient from their location while completing their ultrasound. Our technology can automatically move and adjust itself to capture clear images of the heart. Here's a breakdown:

1. Robotic Assistant: we utilize a robot arm that can gently and accurately place an ultrasound device on a patient to see the heart's images.
2. Intelligent Guidance: This robot is guided by artificial intelligence, which means it's trained to know where to move and how to adjust to get the best images. Trained from the top experts in echocardiography, it's like having a virtual expert guiding the robot's every move!
3. Remote Expertise: specialists from anywhere can remotely watch and guide the process. So, even if the patient is in a small town in a different country, they can benefit from an experienced cardiac sonographer.
4. Consistent and Clear: the images of the heart are consistently clear and precise, helping doctors diagnose heart conditions more accurately. This also allows for standardization which, in the imaging world, makes monitoring chronic conditions easier.

In a nutshell, our device combines the best of robotics and artificial intelligence with experienced, healthcare expertise to make heart imaging easier, clearer, and accessible to all.

Our target population primarily comprises individuals residing in underserved and remote communities where specialized medical expertise, particularly in cardiology, is limited or sporadically available. Here's a closer look:

1. Rural and Remote Populations: Many individuals in rural regions lack easy access to major medical facilities, which means they often miss out on or don't have the means to acquire specialized, but critically important, diagnostic procedures like echocardiograms. This delay can result in late or missed diagnoses of heart conditions.
2. Economically Marginalized Communities: Even within urban settings, economically disadvantaged populations often struggle to afford specialized medical services. They might live in "medical deserts" where certain medical specialties are scarce or inaccessible due to cost barriers.
3. Aging Populations: As people age, the risk of heart diseases increases, necessitating regular cardiac check-ups. Elderly individuals, especially those with mobility issues, find it challenging to travel for specialized medical tests.

Prioritizing Patient Needs:

• Accessibility: By deploying our device in local clinics or community health centers, we aim to bring expert echocardiographic services closer to these populations. They won't need to travel long distances or wait for an expert's availability.
• Affordability: With the automation of our device, the cost per procedure can be significantly reduced. This affordability can be passed on to patients, ensuring that more individuals can access and benefit from this essential diagnostic service.
• Quality and Safety: the coupling of high human expertise with high technical capabilities ensures high-quality imaging. The results are consistent with what a patient might receive in a top-tier hospital. Additionally, the robot is designed with patient safety in mind, ensuring gentle interactions.
• Empowerment through Education: Apricity Robotics isn’t just thinking about diagnostics. We envision a suite of applications for our technology that amplifies the human impact in healthcare.

In essence, our solution strives to bridge the healthcare disparity gap, ensuring that every individual, regardless of their location or economic status, has the right to timely, affordable, and quality cardiac care.

Our team’s composition and collaboration are rooted in a profound understanding of the community we serve. Our strength lies not just in technical expertise but also in our direct experience and relationships with these communities. Here's a detailed look:

1. RN on the Founding Team: One of our team members, having previously served as a Registered Nurse (RN), brings a wealth of practical insights. Their firsthand experiences in treating patients in Columbus, OH, especially from underserved communities, have provided our team with a deep understanding of their unique challenges and needs.
2. Advisory Board with Direct Community Experience: Our advisory board boasts a cardiologist from the Cleveland Clinic, who serves as the Director of Population Health for Heart Failure. In this role, he is focused on scaling efforts to improve heart failure care across a large integrated system. He is a member of the Heart Failure Collaboratory (HFC) and serves on their digital health committee. He is a staff physician in the Cleveland Clinic Amyloidosis Center. This association provides us with a direct line to understand the intricacies of cardiac health and the specific challenges faced by the larger community. His role aligns perfectly with our mission to serve broader communities, ensuring that our solution isn’t just technically proficient but also culturally sensitive and relevant.
3. Engagement with Community-focused Hospitals: Our lead investor is Orlando Health—a hospital that has a strong focus on community medicine. By partnering for impact, we ensure our solutions are tailored for real-world challenges. Their hospital sites—in the US and Puerto Rico— give us a platform to test, iterate, and refine our robot based on actual patient feedback and medical staff insights.

We believe in a bottom-up approach, where community input is not just welcomed but actively sought. Our engagement strategy includes:

• Feedback Loops: Periodic consultations with community health workers, patients, and local clinics give us feedback on the effectiveness and user-friendliness of our device. This ensures that our solution remains grounded and responsive to the community's evolving needs.
• Educational Workshops: We conduct workshops in collaboration with our community-focused hospitals to educate residents about cardiac health and the importance of timely diagnostics. These workshops double up as feedback sessions where we learn directly from potential end-users.
• Community Representation: Our collaboration with hospitals and the presence of an RN on the team ensures that the voice of the community is always present in our design and decision-making processes. Their experiences and perspectives guide our design choices, ensuring that our device is not just a product but a solution shaped by the community.

In conclusion, our team, guided by the community's insights, is in a unique position to deliver a solution that's not only innovative but also deeply resonant with the real needs and challenges of the communities we serve. We are not just creating a solution for the community; we are crafting it with the community.

Apricity Robotics has completed a Phase I National Science project where the objective of creating a minimum viable robot capable of performing transthoracic echocardiograms was achieved. Our device demonstrated accurate placement of the transducer on a phantom and the ability to move the transducer between keypoints as well as angle manipulation at keypoints without ever losing contact nor exceeding safe pressure thresholds. This testing showed  that the robot was both safe and accurate in positioning the transducer, and Apricity commenced human testing.

A single window was also evaluated on humans. The robot was able to place the transducer at the correct location on the human body and acquire images; however, at first, the images were not sufficient for clinical use. Sonographers are skilled at looking at images and optimizing their quality by manipulating the transducer to vary angle and rotation. Taking inspiration from our target customers, Apricity developed a set of algorithms to autonomously segment TTE images, assess their quality, and determine motor outputs to move the transducer in a way that optimizes image resolution.

Our team is dedicated to creating a device that is equal and equitably trained to support patients of all backgrounds. Unfortunately, it is easier to create artificial intelligence that works best on white, healthy males because those are the most affordable volunteers to access during the technology development process while under the constraints of a startup budget. In a proactive effort to diversify our data and in an intentional act of inclusivity to groups that require resources and capital to reach, we need a partner in Solve. Solve’s resources, human experts, and connections to international interests will ensure that there are voices guiding not only what American investors want built but what the universal citizen needs. 

The remote robotic ultrasound platform is innovative in the way it applies technology to healthcare. While robotics have been used for decades in operating rooms, the technology has been unrealized in other applications. Our product will augment the skill of cardiac sonographers, reduce injury in their work, create efficiency and access in patient care, and allow for the education of others to perform echocardiograms. Ultimately, the R-REA will significantly improve how specialized sonographers provide ultrasounds, thus catalyzing broader impacts in the healthcare space.

Our device enables specialized healthcare professionals to remotely provide cardiac ultrasound services–transthoracic echocardiograms, to be exact. This approach dramatically expands access to specialized care in areas where attracting or maintaining skilled technicians is a challenge. This fundamental shift in service delivery transcends geographical constraints and ensures that patients in remote or underserved areas receive the same quality of care as those in urban centers.

Our solution enhances healthcare system efficiency by allowing staff talent to be optimally utilized and reducing the need of transporting a patient to a more expensive, populated health center. Specialized professionals can focus on their core areas of expertise, while routine ultrasounds can be conducted by remote experts. This new approach streamlines resource allocation, reducing patient waiting times and improving the overall quality of healthcare services.

We employ technology to up-skill general ultrasound technicians in cardiac imaging. This approach not only addresses the issue of a limited number of graduates with echocardiography certification but also ensures that healthcare institutions have a continuous supply of skilled professionals. This innovation enhances the adaptability and versatility of the healthcare workforce.

The knowledge and expertise gained through this solution can be applied to various ultrasound specialties, including venous, obstetrics, and pediatrics.

Our solution extends beyond national boundaries, enabling its application in countries struggling to provide effective ultrasound services. This has the potential to revolutionize healthcare access on a global scale, reducing disparities and improving global health outcomes.

By reducing the time required to acquire ultrasound images, our solution promotes a shift from reactive to proactive care. Early access to diagnostic information enables more effective disease management, particularly in the context of cardiovascular conditions. This transition has the potential to improve patient outcomes and reduce healthcare costs.

Our solution aligns with the concurrent development and implementation of AI programs in healthcare. By developing our technology to acquire more ultrasounds, we complement these AI advancements to create innovations which speed up the time between ordering an echocardiogram and having it completed and interpreted.

In conclusion, our innovative solution not only addresses the critical healthcare challenge of providing democratized access to cardiac imaging but also has the potential to catalyze positive impacts throughout the general ultrasound industry, foster global health equity, and drive market changes that prioritize accessible, efficient, and high-quality care.

Our device will provide hospitals with the opportunity to perform transthoracic echocardiograms (TTEs) in the same or less time and with better image quality consistently without the necessity of having a cardiac sonographer at the patient site.

We expect this solution to have a positive impact on immediate outputs like the introduction of the new device into clinical settings and training sonographers to use the device effectively. We also expect the implementation of our device to fit into hospital workflows and ordering processes.

Immediate outcomes include reduced time required to travel to perform TTEs, subjective scoring by interpreting Cardiologists as to increased quality and consistency of TTE image quality and thoroughness, reduction in sign-on bonuses hospitals need to offer to be “fully staffed,” and increased access for patients in rural areas to receiving their cardiac ultrasounds.

Our longer term outcomes include improved patient outcomes for emergent and chronic conditions that TTEs test for, reducing waiting times for patients both in-patient and out-patient settings. Reduced spending by hospitals and lessened burnout of cardiac sonographers.

We will utilize third-party research to validate safety, quality, and workflow of our device to back up the key claims we make. Additionally, we will utilize formal process evaluations to gauge implementation into clinical care. Finally, data from interviews with users, patients, and customers will be collected at every stage of the product lifecycle.

By addressing the need for more efficient and accessible TTE procedures in hospitals, our solution aims to directly impact patient care and hospital operations positively.

Impact goal: Reduce musculoskeletal injury sustained by cardiac sonographers who perform TTEs remotely using our device compared to in-person. 

We will measure our progress by collecting pain scores from sonographers before and after using our device to collect echocardiogram ultrasounds and compare it to control group of sonographers performing scans at the bedside. 

Impact goal: Decrease hospitalized patient wait times for echocardiograms among customers who employ our technology. 

We will measure patient wait times among customers and compare to their wait times before implementing our device. We can also collect data relative to hospital acquired infections, which is heavily influenced by prolonged hospitalization time. 

Impact goal: Perform TTEs with less applied pressure with our device than by a bedside sonographer. 

Our team is capturing pressure applied from transducer onto patient chest wall during remote TTEs performed via our device and will compare to the pressure delivered during a bedside TTE from a sonographer. 

The core technologies of our remote robotic ultrasound platform are robotics and artificial intelligence. Their collaborative effect is a cutting-edge system that employs a sophisticated Convolutional Neural Network (CNN) trained to accurately identify key anatomical landmarks on the patient's torso—enabling precise initial placement of the ultrasound transducer. Our team utilized the application of ultraviolet (UV) markers and illumination to enriching the training dataset and significantly improving model precision.

To achieve precise transducer positioning and maintain optimal contact pressure and orientation for superior quality imaging, our system leverages depth sensing and real-time keypoint detection algorithms. These algorithms control a robotic arm equipped with a Proportional-Integral-Derivative (PID) controller feeding into an admittance control system.

Upon establishing contact with the patient's skin, the system streams the captured ultrasound image through an auto-encoder, which, in conjunction with a scoring algorithm, inputs into a pre-trained neural network. This network, in turn, calculates and outputs the necessary linear and angular velocities for the transducer to capture the optimal view for each specific TTE window.

The scoring algorithm utilizes a U-net convolutional neural network trained the model on a diverse dataset of 350 images. These images, derived from a volunteer patient's heart, represented a comprehensive spectrum of PLAX views, including optimal, suboptimal, and views lacking essential anatomical features for TTE.

The scoring algorithm developed in tandem quantitatively evaluates the echocardiographic images to determine their diagnostic utility. This evaluation considers both the visibility and size of key anatomical structures, assigning a base score for each identified region. Notably, this scoring methodology was rigorously validated against established imaging standards and further endorsed by a board-certified cardiologist, affirming its clinical accuracy and relevance.

An auto-encoder is used to feed information into the optimization algorithm. An auto-encoder is a deep learning model that compresses high-dimensional ultrasound images into a condensed, latent space representation. This compression not only preserves the critical visual features necessary for diagnostic evaluation but also significantly reduces the computational burden.

The imitation learning framework adopted for this project is based on the principles of behavioral cloning, supplemented by Dataset Aggregation (DAgger), a method designed to iteratively refine the model's performance through expert feedback. The neural network, configured with three fully connected layers employing Rectified Linear Unit (ReLU) activations, is tasked with predicting the optimal control actions necessary for precise transducer positioning.

The deployment of the trained model in a live clinical setting revealed the challenges inherent in behavioral cloning, particularly the issue of compounding errors when the robot encounters novel scenarios not covered in the training dataset. To address this, the DAgger technique was employed, allowing experts to provide corrective feedback in real time.

We have two full-time employees, a part-time employees, and one contractor. We are currently hiring for a full-time and a part-time role. 

Our company was founded in 2022 with the awarding of a Phase I grant from the National Science Foundation. Our solution was prototyped in 2023, and we have filed a provisional patent this year. 

We attribute the headway our team has made to the diversity of perspective that our small team boasts. Our CFO and robotic engineer have academic backgrounds; our Product Manager is a RN with a clinical background; our CEO is from a global conglomerate. 

Further more, our team includes a female and an international graduate student. Among our applicants for interns are women and minorities. 

Diversity, equity, and inclusivity are integral to our mission and vision. We believe that to truly revolutionize cardiac care, we need a team and a solution that understands, values, and caters to the myriad nuances of the global community.

1. Diverse Representation: Our team reflects a diverse array of backgrounds, both professionally and personally. By having leaders from different ethnicities, genders, and life experiences, we ensure a variety of perspectives that drive inclusive decision-making.
2. Community Engagement: Recognizing that diversity goes beyond our organization, we've forged partnerships with hospitals focusing on community medicine. These hospitals often serve diverse populations, ensuring that our solutions are tailored to a wide audience. Our collaboration with these institutions ensures that R-REA is not just technologically advanced, but also culturally competent.
3. Feedback Mechanisms: We have established channels where team members can provide feedback on diversity and inclusion matters. This not only helps us identify areas of improvement but also empowers our team members to actively participate in shaping a more inclusive environment.
4. Advisory Board Composition: Our advisory board, including the Director of Population Health for Heart Failure from the Cleveland Clinic, ensures that we stay informed about the diverse needs of patients. Their expertise and insights from various backgrounds enable us to design solutions that cater to a broad demographic.
5. Designing for Inclusivity: In the development of R-REA, we ensure that the data sets used for AI training are diverse. This ensures that our AI models are not biased and can cater to the varied needs of different populations, addressing healthcare disparities in the process.
6. Team's Proximity to the Cause: With a team member who has previously served as an RN, we have firsthand insights into the needs and challenges of healthcare providers. This proximity to the ground realities ensures that our solutions are not just technologically driven but are also deeply empathetic and patient-centered.

Based on feedback from customer discovery, Apricity has developed a go-to-market (G2M) strategy that prioritizes customer segments based on market opportunity, level of need, product-market fit, and willingness to pay. We have also developed a business model and sales strategy for our customers’ organizations that will demonstrate Apricity’s value to all stakeholders and minimize sales cycle time - a common hurdle for new market entrants. Hospitals are moving from sonogram device ownership to leasing models, which reduces their operational costs when a device needs to be serviced or replaced with newer technology. Based on value level delivered, alternative strategies to solve pain points, competitor pricing, and revenue potential for our customers calculated from reimbursement rates we will offer two options for our customers to acquire our devices: an upfront purchase at a cost of $200,00 per unit with options for additional annual support and maintenance (est. $7,000/year/robot), or lease our system to our customers on an annual basis for $60,000 per robot. Other customer-derived revenue categories include setup fees ($5k per device) and additional controllers or customization services. There are also licensing opportunities for other robotics control applications that utilize Apricity’s IP, though this will not be a primary area of pursuit during the G2M execution but more for revenue growth at mature stages. Based on customer discovery feedback, we believe it is in our best interest to include user training as a part of the annual lease to ensure we continue to have highly engaged and highly trained users behind our product.We will initially target large hospital systems which include an urban campus, such as Orlando Health (refer to the Letter of Support). These systems, which often include multiple satellite locations including in rural locations, face challenges in evenly distributing critical and specialized services like echocardiography. This approach aligns with our pricing strategy and meets a critical need for high-quality, consistent care across diverse settings. Large independent hospitals, although standalone, face similar staffing and operational efficiency challenges, making them ripe for the adoption of our technology.

They are defined as 500+ bed hospital in metropolitan designated area and not associated with other hospitals. They are competing with the larger hospital system for talent and thrive based on their reputation for quality service. While the number of units per sale cycle is less than hospital networks, they are still appreciable through the bureaucratic process which is much more simplified. Rural hospitals are the next customer target segment. They have much less access to capital (leasing option preferred), represent small volume of sales but a much shorter sales cycle. IT will take a larger sales force to make appreciable penetration in this segment and this is the primary reason for later-stage expansion. Our lowest priority customers are independent cardiology clinics, critical access hospitals, teaching hospitals, and urgent care. Their pain points and willingness to pay simply do not make sense as a good fit for Apricity until the platform becomes more of a de facto standard.

Apricity proposes a leasing model at $60,000 per unit annually, with a $5,000 initiation fee for new leases to cover setup costs. Lease prices will be adjusted annually by approximately 3% to reflect inflation, ensuring the offering remains competitive and sustainable. Leased units will be replaced every seven years at no additional initiation cost. An outright purchase option is also available for $200,000, with an annual $7,000 premium support contract for maintenance and technical assistance. This option caters to institutions preferring in-house maintenance or requiring vendor support, depending on their operational setup.

Projected first revenues are conservative, with four pre-FDA-certified units expected in 2026 sold at cost-of-goods, scaling up significantly post-FDA certification based on clinical evidence and word-of-mouth validation within the healthcare community. Break-even is forecasted for late 2028, by which time Apricity anticipates a substantial increase in leased or sold units, driven by demonstrated efficacy and operational efficiency.

In order to secure the resources needed to implement the above vision, Apricity has developed a 3 phase private investment financing strategy in addition to the NSF Phase II SBIR funding. Each phase will have a specific milestone that represents a value inflection point that Apricity will leverage to raise the next round of financing at a favorable valuation. 

Seed Round: Apricity is raising a $1M Seed Round with our lead investor,Orlando Health. These funds will support the initial development, efficacy, safety, and usability testing of the prototype robot in clinical settings. It will also kickstart the data collection for FDA approval under an IRB, following a pre submission meeting to guide definitive studies. 

Series A: In the Phase II project's final six months, leveraging development outcomes and FDA pre-sub meeting insights, Apricity will seek $7-9M in Series A funding. This capital injection will fund the completion of the FDA approval process and preparations for a limited market release, encompassing initial manufacturing, marketing, and sales channel development. This round will attract venture capital firms with healthcare technology and life sciences expertise, particularly those experienced in guiding companies through regulatory hurdles to market entry. Strategic investments from healthcare and medical device sector companies will also be pursued for their added value beyond capital. 

Series B / Venture Growth Round: Post-FDA clearance, approximately 18 months following Series A, Apricity plans a significant fundraising round to scale manufacturing, expand the market, and bolster marketing and sales initiatives, propelling the company towards profitability and converting early pilot customers into full-scale deployments. This round, anticipated to be in the realm of $15-20M, will draw interest from later-stage VCs, growth equity firms, and possibly debt financing sources, aligning with Apricity’s expanded market presence and product development based on initial feedback.