Success Story

Aloha Health Network matching Patients to Clinical Trials for best-results

At A Glance

We helped Aloha Health Network build their service to match Patients with the most advantageous of 50,000 current clinical trials in the United States by comparing a patient’s Clinical Phenotype, which is a patient’s precisely defined medical profile, to clinical trial inclusion criteria using NLP, AI and semantic web technologies.

What we did

  • Natural Language Processing libraries for clinical text analysis.
  • Google Cloud Platform native product application development using  Dialogflow supporting patient ‘chat’ sessions, Firebase for the patient’s mobile UI, and, Cloud Functions to create clinical trial protocol objects.
  • Google Cloud Platform engineering for non-native components including AllegroGraph to coorelate patient Clinical Phenotypes to clinical trial protocols.
  • Google Cloud Platform Kubernetes container engineering and security support.

 

Introducing Aloha Health Network

Aloha knows firsthand how challenging it can be to face difficult medical news. For some patients a clinical trial can be a valuable addition to their treatment strategy. But, with over 50,000 concurrent clinical trials in the U. S. discovering the most advantageous trial for a patient requires more research than physicians, much less patients, have time for.

Rob Wynden PhD, Aloha Health Network’s CEO, saw this as an opportunity to apply AI to a critical medical challenge. “Aloha’s founding mission is to deliver personalized medical insights directly to patients and their care teams.”, said Wynden.

Rob’s PhD in Biomedical Clinical Informatics from UCSF, and previous experience as Chief Scientist at UCSF’s IHPS Informatics Lab gave him unique insight into the AI strategy to approach this challenge. The key to the solution would be the Clinical Phenotype, a data object that gives precision and depth to otherwise imprecise and cursory medical records by applying Natural Language Processing and mapping terms to extensive medical ontologies containing millions of relations across terms.

The earliest adopters of Aloha’s clinical trial matching service are patient advocate clinics and charities. To make the Clinical Phenotype and the entire solution a reality, Aloha needed a technology partner with advanced Software-as-a-Service solution skills.

“We are on the cutting edge of applying AI to medical challenges. Telegraph Hill Software gave Aloha the engineering team to put our concept into production.”

Rob Wynden

Founder and CEO, Aloha Health Network

“As a global patient advocate platform,  we are constantly digging through extensive clinical trial listings to help patients find the trials which offer the best possible outcomes.  The Aloha Health Network automates that process. We are excited about  offering the Aloha service to all our patients!” Eric Drew

Founder, WeHeal

Advantages of experience

Natural Language Processing is a technology decades in the making. Recent investment in NLP by Google has accelerated it’s evolution. In-depth understanding of the underlying principles of NLP help apply new advances to greatest effect.   Telegraph Hill Software’s engineering team included senior technologists with decades of experience creating NLP solutions for  healthcare and financial services applications.

Semantic web processing matches terms in medical records and words used by patients in AI guided ‘chat’ sessions with terms in medical ontologies which contain millions of relations among terms resulting in a Clinical Phenotype with unprecedented breadth and depth. The engineering challenge is to make these enormous ontologies efficiently machine accessible and searchable.

Matching a patient’s clinical phenotype to a clinical trial inclusion criteria is far more complex than a simple SQL statement can achieve. Telegraph Hill Software’s engineers used the graph database AllegroGraph to connect patients with clinical trials and present the most promising trial alternatives to the patient via a mobile UI.

Because Telegraph’s core team was composed of senior technologists Aloha was able to extend their work with offshore, junior and even intern developers to make maximum progress on a minimum funding budget.

The Google Cloud Platform Solution

Aloha took advantage of native GCP tools to insure security of highly sensitive medical information and built custom applications within Kubernetes containers. 

Firebase for mobile application development and microsites for individual disease conditions.

Cloud Functions and Cloud SQL Postgres to process and store clinical trial protocols retrieved from clinicaltrials.gov.

Dialogflow to interact with patients in order to gather data about their condition in a natural, conversational mode.

AllegroGraph for custom development of semantic graphing from patient Clinical Phenotypes to Clinical Trial protocols.

 

What can we do for you?

7 + 13 =

Related Insights

CONTACT US

info@thpii.com

MAILING ADDRESS
268 Bush Street, #1500
San Francisco, CA 94104

OFFICE LOCATION
535 Mission Street
San Francisco, CA 94105