CoMotion Innovation Gap Fund Awardees – Spring 2023

Automatic speech recognition (ASR) systems struggle with dialect variation making ASR less accurate for certain racial and ethnic groups due to historical, representation, and aggregation bias, resulting in lower use of ASR systems by those groups and even potential dangerous and deadly outcomes. SocioLx Solutions links people, technology and Sociolinguistic science to undo these 3 biases and improve ASR for all by acquiring user experience data, building dialect-rich datasets, and creating novel software (SocioLinx) with pronunciation modeling to tune algorithms to make ASR work for everyone.

Each year in the U.S. more than 100,000 patients suffer cardiac arrest caused by ventricular fibrillation, a lethal arrhythmia that causes death to occur without urgent resuscitation. Unfortunately, only about one out of five patients are saved, in part because resuscitation treatments are often administered too late or mismatched to a patient’s specific conditions following a one-size-fits-all protocol. We are developing a suite of next-generation defibrillator algorithms that evaluate a patient’s time-dependent physiology in order to guide rescuers in providing the right treatments at the right time, with the goal of tripling ventricular fibrillation cardiac arrest survival.

Gastrointestinal leaks are tears along the digestive tract that allow bacteria to enter the rest of the body and occurs in 1.3 million patients per year. With no FDA approved devices on the market, clinicians piece together their own devices to treat these leaks using endoscopic vacuum therapy (EVT or E-Vac), which takes between 30 and 90 minutes to build the device and place in the leak site. The team is developing a novel device to reduce the preparation and deployment of EVT treatment to 2 easy, quick steps saving valuable clinician time and improving outcomes. With Simpl-EVac, EVT will no longer be DIY.

Distal pancreatectomy surgery, where a portion of cancerous pancreas is removed, has complication rates of up to 40% due to standard surgical staples not adequately sealing the pancreas allowing fluids to leak into the body. To address this need the team has designed a new clip to seal the pancreas composed of a 3D printable protein-based bioplastics providing high strength, biocompatibility and biodegradable properties making it the perfect material for this and many other surgical applications.

Multi-organ failure (MOF) is a critical condition in ICUs and causes half million deaths annually. Current organ support devices do not provide efficient treatment outcomes and lead to high mortality. To improve the situation, the team provides the AMOR device, a modular system providing multiple organ functions support for kidney, liver, and lung. By enhancing toxin removal and excess body fluid control, MOF patients can survive until organ function recovery or transplantation happened. The concept has been verified by clinical trial MAHD and significantly improved 28-day survival rate from 10% to 80% among 10 liver failure patients. Compared with existing organ support devices, the AMOR system will be more effective, less risky, and simpler to use.

Liver cirrhosis causes over 1.3 million deaths annually, but this disease can be effectively treated with a carefully placed shunt. However, even in the US only 7% of patients receive this life-saving intervention due to the extreme difficulty and risk of the procedure. The team is developing a new device for performing this procedure that greatly increases the success rate and reduces the training needed making this life-saving procedure widely available.

Chest pain is a common reason for Emergency Department (ED) visits, but clinicians lack quality clinical decision tools appropriate for the busy ED setting resulting in wasteful unneeded tests and incorrect diagnosis with potentially dangerous outcomes. The Seattle Clinical Pathway Platform (SCiPP) is developing a comprehensive, easy to use evaluation tool based on the latest clinical guidelines for chest pain designed for the ED in the form of a user-friendly smartphone app to help clinicians make the best decisions to treat chest pain every time.

Modern manufacturing technologies, including 3D printing, enable the fabrication of devices with shapes and functions that exceed the complexity levels supported by existing software for computer-aided design (CAD) and engineering analysis. The limitations of existing software leads to time-consuming workflows that delay product development and inhibit bringing products to market early to enhance revenue generation. To meet the need for efficient design and engineering validation of new devices, the team is developing a new geometric modeling kernel and a suite of state-of-the-art analysis tools that accelerate the design/validation process by up to 1000x.

Blood in stool can be an indicator of serious gastrointestinal diseases including colorectal cancer or precancerous polyps, as well as inflammatory bowel diseases such as Crohn’s disease but many people do not follow recommended screenings resulting in missed disease diagnoses. The team is developing HemoLens, a novel, non-contact, optical method for detection of occult blood in the stool that consists of a smartphone app with small LED light attachment. HemoLens will allow patients to monitor their stool for blood inexpensively, at home, and without any direct contact with stool.

Preeclampsia, a pregnancy-associated disorder characterized by high blood pressure and multi-organ dysfunction, poses a significant threat to maternal and fetal health worldwide, but may be prevented if treated before symptoms start. Unfortunately, there are no diagnostic tests on the market that can accurately diagnose before symptom onset. The team is developing a blood test using maternal blood capable of diagnosing preeclampsia prior to its onset, to enable the timely administration of preventative treatments.

Studies have shown two common post-delivery issues cause longer hospital stays and negative patient experiences: 1) excessive, unnecessary wound pain after cesarean sections; 2) and maternal anxieties related to poor information about the recovery process, specifically regarding breastfeeding. Juno is an app that disrupts the classic care model where the patient is a silent recipient of care instructions from doctors and nurses by providing new mothers with personalized recommendations and reminders to get the care every new mother deserves.