1964-1965
July 1, 2022
Canada
July 1, 2022Biofourmis
Overview
The Biofourmis team is a multidisciplinary one with engineers, clinicians, and researchers from around the globe operating at the intersection of data science, biotechnology, and software technology. They aim to free patients from the stressful and expensive cycle of clinical exacerbations and hospital readmissions and facilitate more harmonious communication with the medical team.
Biofourmis A.I Aspect
Their field-tested solution enables very advanced pattern recognition that can facilitate visualizing a patient’s future to help change it for the better.
The Biofourmis Biovitals® system allows teams at different locations to monitor physiological signs and other essential indicators in patients with acute and chronic conditions. It then compares that information with a centralized database. This analysis provides clinical teams with a real-time study of a patient’s disease trajectory that may enable them to detect, predict, and prevent adverse medical outcomes before they occur.
Their field-tested solution enables very advanced pattern recognition that can facilitate visualizing a patient’s future to help change it for the better.
The Biofourmis Biovitals® system allows teams at different locations to monitor physiological signs and other essential indicators in patients with acute and chronic conditions. It then compares that information with a centralized database. This analysis provides clinical teams with a real-time study of a patient’s disease trajectory that may enable them to detect, predict, and prevent adverse medical outcomes before they occur.
Biovitals® is a modular solution easily customized to unique patient populations or areas of research. Biovitals® brings disease, data, and machine learning together to revolutionize chronic care management.
AI-Based Ecosystem
Biofourmis uses software-based therapeutic intervention achieved by using treatment algorithms relevant to the patient’s condition. This therapeutic approach allows physicians to optimize treatment, allow for the proper drug dosing, and modify treatment plans appropriately.
These algorithms help predict and detect serious medical events and decompensation via advanced analytics that process active and passive data, allowing for quick, efficient interventions.
Personalized feedback is also available for patients via a patient-directed application to coach, engage and promote treatment program adherence. Individualized content is proactively served to the patient’s app supporting specific diseases and comorbid conditions.
Biofourmis uses software-based therapeutic intervention achieved by using treatment algorithms relevant to the patient’s condition. This therapeutic approach allows physicians to optimize treatment, allow for the proper drug dosing, and modify treatment plans appropriately.
These algorithms help predict and detect serious medical events and decompensation via advanced analytics that process active and passive data, allowing for quick, efficient interventions.
Personalized feedback is also available for patients via a patient-directed application to coach, engage and promote treatment program adherence. Individualized content is proactively served to the patient’s app supporting specific diseases and comorbid conditions.
Biofourmis’ IT Solution
BiovitalsHF: This platform is a prescription software to enable early detection of heart failure exacerbation and guideline-directed improvement of heart failure medication use. The purpose is to reduce the length of stay, 30-day rehospitalization rate, and overall healthcare spending while improving the quality of life and function in patients with heart failure.
RhythmAnalytics: is a cloud-based system that automatically detects more than 15 different cardiac arrhythmias. A deep neural network architecture was trained on 4 million EKGs to achieve the highest degree of detection performance. RhythmAnalytics is a service-offering software with scalable APIs to enable easy integration.
Biovitals Sentinel: Biovitals Sentinel by Biofourmis allows for the continuous remote monitoring of critical vitals and patient-reported symptoms using Everion, an innovative new multi-sensor armband wearable device. It uses advanced AI algorithms to process the collected data and detect physiological changes. This algorithm allows it to predict events in various disease conditions, facilitating early intervention and dramatically improving outcomes of many gold-standard therapeutics.
Painfocus: This platform augments pain management by objectively assessing pain to guide personalized therapeutic decision-making. There are currently two ongoing multi-center clinical trials in postoperative pain and cancer pain that use Painfocus.
Gaido: Prescription software to continuously monitor and detect early signs of deterioration in oncology patients, whether pre-or post-treatment. Gaido prevents avoidable readmissions and helps spare patients from developing complications. The purpose is to reduce rehospitalization and overall healthcare spending while improving the quality of life of cancer patients.
Biovitals Research: This is a digital clinical trial platform with software that allows the collection of real-time study data anytime and anywhere, from multiple sources, including phone apps and other connected devices. The platform provides a fully integrated suite of tools for digital clinical trials and remote passive monitoring. Biovitals Analytics also identifies novel patient-centric biomarkers as surrogate endpoints to achieve the desired endpoints and speed up clinical trials.
Biofourmis' Financial Aspects
In 2019, Biofourmis secured a $35 million series B round. In 2020, the company announced an investment of a $100 million series C financing round led by SoftBank, Vision Fund 2. The participating investors included Openspace Ventures, MassMutual Ventures, Sequoia Capital, and EDBI. Since then, the company has increased its revenue significantly through new partnerships and growth with seven pharmaceutical companies and ten health systems globally.
Also, Biofourmis will use the funding for developing, validating, and commercializing several released and unreleased digital therapeutic solutions. These solutions will be applied across cardiology, respiratory, oncology, and pain, focusing on the United States and key Asian markets, including Pacific Asia, China, and Japan
In 2019, Biofourmis secured a $35 million series B round. In 2020, the company announced an investment of a $100 million series C financing round led by SoftBank, Vision Fund 2. The participating investors included Openspace Ventures, MassMutual Ventures, Sequoia Capital, and EDBI. Since then, the company has increased its revenue significantly through new partnerships and growth with seven pharmaceutical companies and ten health systems globally.
Also, Biofourmis will use the funding for developing, validating, and commercializing several released and unreleased digital therapeutic solutions. These solutions will be applied across cardiology, respiratory, oncology, and pain, focusing on the United States and key Asian markets, including Pacific Asia, China, and Japan
Feasibility and Usability Aspects of Continuous Remote Monitoring of Health Status in Palliative Cancer Patients Using Wearables
Patients in palliative care are a population where deterioration of health is common and expected. Around 50% of emergency visits and readmissions in this patient group are deemed avoidable. The use of health monitoring and sensor-equipped arm bracelets could help in the prediction of a decline in health status. The study aimed to evaluate the feasibility as well as the patient’s acceptance of remote monitoring using wearables in palliative cancer patients.
The researchers selected 30 cancer patients treated with palliative care in an inpatient setting with an estimated life expectancy of more than eight weeks and less than twelve months. They were provided with a smartphone with an “Activity Monitoring” app and a sensor-equipped bracelet and monitored over twelve weeks.
Patients in palliative care are a population where deterioration of health is common and expected. Around 50% of emergency visits and readmissions in this patient group are deemed avoidable. The use of health monitoring and sensor-equipped arm bracelets could help in the prediction of a decline in health status. The study aimed to evaluate the feasibility as well as the patient’s acceptance of remote monitoring using wearables in palliative cancer patients.
The researchers selected 30 cancer patients treated with palliative care in an inpatient setting with an estimated life expectancy of more than eight weeks and less than twelve months. They were provided with a smartphone with an “Activity Monitoring” app and a sensor-equipped bracelet and monitored over twelve weeks.
Of these 30 participants, 83% completed the whole study period. The bracelet was worn 53% of the time, and the smartphone was used 85% of the study days. The completion rate of daily questionnaires was 73%, and 28 patients were able to handle the wearables and operate the app without major problems.
Hospital-Level Care at Home for Acutely Ill Adults
A home care model was designed for acutely ill patients and proposed physician home visits with 24-hour physician coverage, twice-daily nurse visits, and home-based treatments. This model was compared with traditional hospitalizations. Home hospitalization has numerous advantages, including reduced cost, and it can provide care in a patient-centered manner as they can be around their family, in the comfort of their own home, and sleep without interruptions.
Results showed that home patients became more active (median percentage of the day, 12% vs. 23%) and spent less of the day lying down (median percentage of the day, 18% vs. 55%).
Discharge planning at home could have a positive impact as it can be personalized to the home environment.
Arrhythmia Detection Using Deep Learning and Multidimensional Representation
The implementation of a fully automated ECG analysis using AI has the potential to improve the accuracy and cost associated with ECG interpretation significantly. For this study, researchers proposed a new deep learning technique to detect arrhythmias in ECG. The system collects ECG and converts it into wavelets and short-time Fourier transform (STFT) to better represent beat-to-beat morphologies. Those segments are then imported into a densely connected convolutional neural network for arrhythmia classification.
In this study, they collected 121,346 single lead episodic ECG records from patients with a history of cardiac arrhythmias as the main data source. The original dataset was split into two parts, 80% for training and 20% for validation. For testing and performance evaluation, they applied a dataset of 600 ECG records and covered over 30 types of arrhythmias.
The results showed a 90% sensitivity, a 98% specificity, and an F1 score of 0.83. The algorithm outperformed the average of cardiologists and the models generated according to previous studies, showing how an automated diagnosis tool can help clinicians reduce time and increase affordability while analyzing an ECG.
Remote Optimization of medical therapy in patients with heart failure with a reduced ejection fraction
A cohort study included patients with heart failure and ejection fraction ≤ 40%. Medical therapy was adapted to a sequential titration algorithm modeled on the current ACC/AHA HF Guidelines The proportion of patients receiving guideline-directed medical therapy (GDMT) in the intervention and reference groups at three months was the following: out of 1028 eligible patients, 197 (19%) consented to participation in the medication optimization program, and 831 (81%) continued with traditional care. Seven hundred fifty-nine (73.8%) participants were medicated with angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and angiotensin-receptor neprilysin inhibitors. Seven hundred forty-six patients used beta-blockers, and 303 participants were treated with mineralocorticoid receptor antagonists.
A cohort study included patients with heart failure and ejection fraction ≤ 40%. Medical therapy was adapted to a sequential titration algorithm modeled on the current ACC/AHA HF Guidelines The proportion of patients receiving guideline-directed medical therapy (GDMT) in the intervention and reference groups at three months was the following: out of 1028 eligible patients, 197 (19%) consented to participation in the medication optimization program, and 831 (81%) continued with traditional care. Seven hundred fifty-nine (73.8%) participants were medicated with angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and angiotensin-receptor neprilysin inhibitors. Seven hundred forty-six patients used beta-blockers, and 303 participants were treated with mineralocorticoid receptor antagonists.
Results after three months showed that the remote intervention group experience was better rated compared to the baseline in the utilization of all categories of GDMT than those in the usual care group.
References