Bedminster, New Jersey-based autism-focused technology company NeuroSpectrum Insights Inc. has filed for an initial public offering.
The company seeks to commercialize software that claims a more than 90% accuracy rate in diagnosing autism rates and can be used to diagnose children as young as nine months old. While the company estimates that its IPO will raise a modest $6.9 million, the potential for an accurate, objective and accessible autism diagnosis is hard to understate.
Documents released by NeuroSpectrum Insights, formerly known as Autism Diagnostic Technologies, on Friday state that the company has developed software to analyze the structural MRIs of the brain, which it has branded GyriCalc. That software analyzes parts of the cerebral cortex and their surface characteristics, looking for indications that suggest a neurological difference.
NeuroSpectrum has an exclusive license to commercialize the underlying technology with the University of Louisville Research Foundation. The algorithm used to diagnose autism was developed by the University of Louisville’s bioimaging laboratory and department of psychiatry and behavioral science. They partnered with the Norton Children’s Autism Center as well.
In late 2024, the company released a report detailing its analysis of several databases of structural MRIs, achieving an average accuracy rate of 92%. The highest and lowest accuracy rates for the analysis of datasets presented in the report were 97% and 88%, respectively.
The company is seeking approval from the U.S. Food and Drug Administration to market its software as a medical device. It applied for 510(k) approval in March 2025. Specifically, it refers to this software as a computer-aided diagnostic system.
Representatives of NeuroSpectrum Insights have not returned a request for comment. CEO Andrew Stewart and CFO Joerg Klaube lead the company. Two members of its board of directors are also part of Academic Technology Ventures, a venture capital firm that specializes in commercializing university-developed intellectual property.
The use of a medical device that analyzes patient anatomy with the scalability of software and existing imaging devices could address several aspects of the autism therapy industry that tend to be inefficient. The shortage of diagnosticians and the fundamentally subjective nature of the diagnostic process create a time-intensive process that delays potential treatment for patients in need and may inadvertently drive patients to treatment that may not be well suited for them, potentially harming the industry. It may also provide an alternative to the proliferation of telehealth services that aim to address the inefficiencies of legacy entities in the industry.
On top of that, parents welcome such a product. One study found that 97% of parents with experience caring for an autistic child and 68% of parents without experience said that they would be interested in biomarker-based predictive testing.
However, acceptance of new and innovative products is a difficult prospect in health care. Generally, it’s difficult to build interest and momentum behind a product in an industry that is defined by slow adoption of technology. That comes, in no small part, from the interests and demands of payers. NeuroSpectrum Insights acknowledges that in its IPO documents.
“The availability and extent of reimbursement by governmental and private payers is essential for most patients to be able to afford the clinical diagnostic tests that we currently or in the future plan to develop and sell,” the documents state.
While autism diagnosis is the first focus for the company, NeuroSpectrum Insights says it may begin developing diagnostics for dyslexia, ADHD and schizophrenia.
The company has yet to generate revenue. It will introduce its product to the market later in 2025. It projects it will generate over $40 million in annual revenue in 2029.
