​Cognoa raises $11.6m to advance diagnosis of autism through machine learning

Healthtech startup Cognoa has raised $11.6m to help advance its machine learning-powered technology that targets the early diagnosis of autism and other developmental delays in children.
Written by Tas Bindi, Contributor on

Healthtech startup Cognoa has raised $11.6 million from Morningside, a venture capital group founded in Boston by the Chan family of Hong Kong. The latest round brings the total amount raised by the Palo Alto, California-based startup, since its founding in 2013, to $20.4 million.

Originally developed at Harvard and Stanford's medical schools, Cognoa's assessment platform claims to use information and videos provided by parents to detect autism and other developmental delays in children up to 13 months earlier than the regular diagnosis age. This in turn provides an important window for medical intervention, the startup said.

Over time, thanks to the machine learning algorithms developed over five years, Cognoa is able "learn" how to better identify autism and other developmental and behavioural disorders, the startup said.

Brent Vaughan, CEO of Cognoa, believes Cognoa's machine learning approach to diagnosing children with developmental delays could change the standard of care for all children.

Cognoa's assessment platform is currently sold to employers, who provide it to working parents as part of their company's broader healthcare benefits schemes, Vaughan said.

"By working with Cognoa, employers get clinically-validated developmental assessment, customised activity plans for each child, library of expert Q&A on child development, and care navigation for families that need it," he added.

The startup plans to deliver Cognoa through paediatricians in the future, but will focus on growing Cognoa for Employers in the interim.

To date, Cognoa's assessment platform has been used by 300,000 families, more than 20,000 of which are based outside the US.

The new funding will be used to advance Cognoa's machine learning-powered technology; support additional clinical validation and the US Food and Drug Administration (FDA) submissions and approvals process; and expand commercialisation with employers, payers, and clinicians.

Machine learning is beginning to be used more broadly in healthcare to assist with diagnoses and to identify optimal treatments for patients in many medical areas.

In February, IBM announced its researchers in Australia had trained a research version of Watson to recognise abnormalities in retina images that could assist doctors in the early detection of eye diseases such as glaucoma and diabetic retinopathy.

Earlier in March, IBM Watson Health and Tel Aviv-based MedyMatch announced a licensing agreement that will see the former's imaging group distribute the latter's AI-powered intracranial bleed assessment application to imaging experts working in hospital emergency rooms and other acute care settings around the world.

In August last year, Google's AI research unit DeepMind announced that it was partnering with the UK's National Health Service to explore how machine learning could help doctors treat head and neck cancers -- particularly whether it can help reduce planning times from four hours to one before radiotherapy treatment is started.

Microsoft has previously revealed a number of cancer-fighting projects it has under way that apply machine learning. One instance, Project Hanover, is seeking to make personalised, precision cancer therapy available to all cancer patients by helping oncologists sift through reams of biomedical research papers faster. Microsoft is targeting the so-called "molecular tumor board", the group that convenes to devise personal therapy to suit a patient's specific cancer.

Editorial standards